Principles And Practice Of Endocrinology And Metabolism

Principles and Practice of Endocrinology and Metabolism (December 2002): by Kenneth L. Becker (Editor), C. Ronald Kahn (Editor), Robert W. Rebar (Editor) By Lippincott Williams & Wilkins Publishers

By OkDoKeY

Principles and Practice of Endocrinology and Metabolism CONTENTS Editors Contributing Authors Preface Preface to the First Edition Part I General Principles of Endocrinology Kenneth L. Becker, Editor Part II The Endocrine Brain and Pituitary Gland Gary L. Robertson, Editor Part III The Thyroid Gland Leonard Wartofsky, Editor Part IV Calcium and Bone Metabolism John P. Bilezikian, Editor Part V The Adrenal Glands D. Lynn Loriaux, Editor Part VI Sex Determination and Development Robert W. Rebar and William J. Bremner, Editors Part VII Endocrinology of the Female Robert W. Rebar, Editor Part VIII Endocrinology of the Male William J. Bremner, Editor Part IX Disorders of Fuel Metabolism C. Ronald Kahn, Editor Part X Diffuse Hormonal Secretion Eric S. Nylén, Editor Part XI Heritable Abnormalities of Endocrinology and Metabolism Kenneth L. Becker, Editor Part XII Immunologic Basis of Endocrine Disorders Leonard Wartofsky, Editor Part XIII Endocrine and Metabolic Dysfunction in the Growing Child and in the Aged Wellington Hung, Editor Part XIV Interrelationships Between Hormones and the Body Kenneth L. Becker, Editor Part XV Hormones and Cancer Kenneth L. Becker, Editor Part XVI Endocrinology of Critical Illness Eric S. Nylén, Editor Part XVII Endocrine and Metabolic Effects of Toxic Agents Kenneth L. Becker, Editor Part XVIII Endocrine Drugs and Values Kenneth L. Becker, Editor

PART I GENERAL PRINCIPLES OF ENDOCRINOLOGY Kenneth L. Becker, Editor Chapter 1 Endocrinology and the Endocrine Patient KENNETH L. BECKER, ERIC S. NYLÉN, and RICHARD H. SNIDER, JR. Chapter 2 Molecular Biology: Present and Future MEHBOOB A. HUSSAIN and JOEL F. HABENER Chapter 3 Biosynthesis and Secretion of Peptide Hormones WILLIAM W. CHIN Chapter 4 Hormonal Action DARYL K. GRANNER Chapter 5 Feedback Control in Endocrine Systems DANIEL N. DARLINGTON and MARY F. DALLMAN Chapter 6 Endocrine Rhythms EVE VAN CAUTER Chapter 7 Growth and Development in the Normal Infant and Child GILBERT P. AUGUST

PART II THE ENDOCRINE BRAIN AND PITUITARY GLAND Gary L. Robertson, Editor Chapter 8 Morphology of the Endocrine Brain, Hypothalamus, and Neurohypophysis JOHN R. SLADEK, JR., and CELIA D. SLADEK Chapter 9 Physiology and Pathophysiology of the Endocrine Brain and Hypothalamus PAUL E. COOPER Chapter 10 Pineal Gland RUSSEL J. REITER Chapter 11 Morphology of the Pituitary In Health and Disease KAMAL THAPAR, KALMAN KOVACS, and EVA HORVATH

SECTION A ADENOHYPOPHYSIS Chapter 12 Growth Hormone and Its Disorders GERHARD BAUMANN Chapter 13 Prolactin and Its Disorders LAURENCE KATZNELSON and ANNE KLIBANSKI Chapter 14 Adrenocorticotropin: Physiology and Clinical Aspects DAVID J. TORPY and RICHARD V. JACKSON Chapter 15 Thyroid-Stimulating Hormone and Its Disorders JOSHUA L. COHEN Chapter 16 Pituitary Gonadotropins and Their Disorders WILLIAM J. BREMNER, ILPO HUHTANIEMI, and JOHN K. AMORY Chapter 17 Hypopituitarism JOSEPH J. PINZONE Chapter 18 Hypothalamic and Pituitary Disorders in Infancy and Childhood ALAN D. ROGOL Chapter 19 The Optic Chiasm in Endocrinologic Disorders R. MICHAEL SIATKOWSKI and JOEL S. GLASER Chapter 20 Diagnostic Imaging of the Sellar Region ERIC BOUREKAS, MARY OEHLER, and DONALD CHAKERES Chapter 21 Medical Treatment of Pituitary Tumors and Hypersecretory States DAVID H. SARNE Chapter 22 Radiotherapy of Pituitary-Hypothalamic Tumors MINESH P. MEHTA Chapter 23 Neurosurgical Management of Pituitary-Hypothalamic Neoplasms DAVID S. BASKIN Chapter 24 Pituitary Tumors: Overview of Therapeutic Options PHILIPPE CHANSON

SECTION B NEUROHYPOPHYSIAL SYSTEM Chapter 25 Physiology of Vasopressin, Oxytocin, and Thirst GARY L. ROBERTSON Chapter 26 Diabetes Insipidus and Hyperosmolar Syndromes PETER H. BAYLIS and CHRISTOPHER J. THOMPSON Chapter 27 Inappropriate Antidiuresis and Other Hypoosmolar States JOSEPH G. VERBALIS

PART III THE THYROID GLAND Leonard Wartofsky, Editor Chapter 28 Approach to the Patient with Thyroid Disease LEONARD WARTOFSKY Chapter 29 Morphology of the Thyroid Gland VIRGINIA A. LIVOLSI Chapter 30 Thyroid Physiology: Synthesis and Release, Iodine Metabolism, Binding and Transport H. LESTER REED Chapter 31 Thyroid Physiology: Hormone Action, Receptors, and Postreceptor Events PAUL M. YEN Chapter 32 Thyroid Hormone Resistance Syndromes STEPHEN JON USALA Chapter 33 Thyroid Function Tests ROBERT C. SMALLRIDGE Chapter 34 Thyroid Uptake and Imaging SALIL D. SARKAR and DAVID V. BECKER Chapter 35 Thyroid Sonography, Computed Tomography, and Magnetic Resonance Imaging MANFRED BLUM Chapter 36 Abnormal Thyroid Function Test Results in Euthyroid Persons HENRY B. BURCH Chapter 37 Adverse Effects of Iodide JENNIFER A. NUOVO and LEONARD WARTOFSKY Chapter 38 Nontoxic Goiter PAUL J. DAVIS and FAITH B. DAVIS Chapter 39 The Thyroid Nodule LEONARD WARTOFSKY and ANDREW J. AHMANN Chapter 40 Thyroid Cancer ERNEST L. MAZZAFERRI Chapter 41 Unusual Thyroid Cancers MATTHEW D. RINGEL Chapter 42 Hyperthyroidism KENNETH D. BURMAN Chapter 43 Endocrine Ophthalmopathy MELVIN G. ALPER and LEONARD WARTOFSKY Chapter 44 Surgery of the Thyroid Gland EDWIN L. KAPLAN Chapter 45 Hypothyroidism LAWRENCE E. SHAPIRO and MARTIN I. SURKS Chapter 46 Thyroiditis IVOR M. D. JACKSON and JAMES V. HENNESSEY Chapter 47 Thyroid Disorders of Infancy and Childhood WELLINGTON HUNG

PART IV CALCIUM AND BONE METABOLISM John P. Bilezikian, Editor Chapter 48 Morphology of the Parathyroid Glands VIRGINIA A. LIVOLSI Chapter 49 Physiology of Calcium Metabolism EDWARD M. BROWN Chapter 50 Physiology of Bone LAWRENCE G. RAISZ Chapter 51 Parathyroid Hormone DAVID GOLTZMAN and GEOFFREY N. HENDY Chapter 52 Parathyroid Hormone–Related Protein GORDON J. STREWLER Chapter 53 Calcitonin Gene Family of Peptides KENNETH L. BECKER, BEAT MÜLLER, ERIC S. NYLÉN, RÉGIS COHEN, OMEGA L. SILVA, JON C. WHITE, and RICHARD H. SNIDER, JR. Chapter 54 Vitamin D THOMAS L. CLEMENS and JEFFREY L. H. O’RIORDAN Chapter 55 Bone Quantification and Dynamics of Turnover DAVID W. DEMPSTER and ELIZABETH SHANE Chapter 56 Markers of Bone Metabolism MARKUS J. SEIBEL, SIMON P. ROBINS, and JOHN P. BILEZIKIAN Chapter 57 Clinical Application of Bone Mineral Density Measurements PAUL D. MILLER, ABBY ERICKSON, and CAROL ZAPALOWSKI Chapter 58 Primary Hyperparathyroidism SHONNI J. SILVERBERG and JOHN P. BILEZIKIAN Chapter 59 Nonparathyroid Hypercalcemia ANDREW F. STEWART Chapter 60 Hypoparathyroidism and Other Causes of Hypocalcemia SUZANNE M. JAN DE BEUR, ELIZABETH A. STREETEN, and MICHAEL A. LEVINE Chapter 61 Renal Osteodystrophy KEVIN J. MARTIN, ESTHER A. GONZALEZ, and EDUARDO SLATOPOLSKY Chapter 62 Surgery of the Parathyroid Glands GERARD M. DOHERTY and SAMUEL A. WELLS, JR. Chapter 63 Osteomalacia and Rickets NORMAN H. BELL Chapter 64 Osteoporosis ROBERT LINDSAY and FELICIA COSMAN Chapter 65 Paget Disease of Bone ETHEL S. SIRIS Chapter 66 Rare Disorders of Skeletal Formation and Homeostasis MICHAEL P. WHYTE Chapter 67 Diseases of Abnormal Phosphate Metabolism MARC K. DREZNER Chapter 68 Magnesium Metabolism ROBERT K. RUDE Chapter 69 Nephrolithiasis MURRAY J. FAVUS and FREDRIC L. COE Chapter 70 Disorders of Calcium and Bone Metabolism in Infancy and Childhood THOMAS O. CARPENTER

PART V THE ADRENAL GLANDS D. Lynn Loriaux, Editor Chapter 71 Morphology of the Adrenal Cortex and Medulla DONNA M. ARAB O’BRIEN Chapter 72 Synthesis and Metabolism of Corticosteroids PERRIN C. WHITE Chapter 73 Corticosteroid Action PERRIN C. WHITE Chapter 74 Tests of Adrenocortical Function D. LYNN LORIAUX Chapter 75 Cushing Syndrome DAVID E. SCHTEINGART Chapter 76 Adrenocortical Insufficiency D. LYNN LORIAUX Chapter 77 Congenital Adrenal Hyperplasia PHYLLIS W. SPEISER Chapter 78 Corticosteroid Therapy LLOYD AXELROD Chapter 79 Renin-Angiotensin System and Aldosterone DALILA B. CORRY and MICHAEL L. TUCK Chapter 80 Hyperaldosteronism JOHN R. GILL, JR. Chapter 81 Hypoaldosteronism JAMES C. MELBY Chapter 82 Endocrine Aspects of Hypertension DALILA B. CORRY and MICHAEL L. TUCK Chapter 83 Adrenocortical Disorders in Infancy and Childhood ROBERT L. ROSENFIELD and KE-NAN QIN Chapter 84 The Incidental Adrenal Mass D. LYNN LORIAUX Chapter 85 Physiology of the Adrenal Medulla and the Sympathetic Nervous System DAVID S. GOLDSTEIN Chapter 86 Pheochromocytoma and Other Diseases of the Sympathetic Nervous System HARRY R. KEISER Chapter 87 Adrenomedullary Disorders of Infancy and Childhood WELLINGTON HUNG Chapter 88 Diagnostic Imaging of the Adrenal Glands DONALD L. MILLER Chapter 89 Surgery of the Adrenal Glands GARY R. PEPLINSKI and JEFFREY A. NORTON

PART VI SEX DETERMINATION AND DEVELOPMENT Robert W. Rebar and William J. Bremner, Editors Chapter 90 Normal and Abnormal Sexual Differentiation and Development JOE LEIGH SIMPSON and ROBERT W. REBAR Chapter 91 Physiology of Puberty PETER A. LEE Chapter 92 Precocious and Delayed Puberty EMILY C. WALVOORD, STEVEN G. WAGUESPACK, and ORA HIRSCH PESCOVITZ Chapter 93 Micropenis, Hypospadias, and Cryptorchidism in Infancy and Childhood WELLINGTON HUNG

PART VII ENDOCRINOLOGY OF THE FEMALE Robert W. Rebar, Editor Chapter 94 Morphology and Physiology of the Ovary GREGORY F. ERICKSON and JAMES R. SCHREIBER Chapter 95 The Normal Menstrual Cycle and the Control of Ovulation ROBERT W. REBAR, GARY D. HODGEN, and MICHAEL ZINGER Chapter 96 Disorders of Menstruation, Ovulation, and Sexual Response ROBERT W. REBAR Chapter 97 Ovulation Induction MICHAEL A. THOMAS Chapter 98 Endometriosis ROBERT L. BARBIERI Chapter 99 Premenstrual Syndrome ROBERT L. REID and RUTH C. FRETTS Chapter 100 Menopause BRIAN WALSH and ISAAC SCHIFF Chapter 101 Hirsutism, Alopecia, and Acne ENRICO CARMINA and ROGERIO A. LOBO Chapter 102 Functioning Tumors and Tumor-Like Conditions of the Ovary I-TIEN YEH, CHARLES ZALOUDEK, and ROBERT J. KURMAN Chapter 103 The Differential Diagnosis of Female Infertility STEVEN J. ORY and MARCELO J. BARRIONUEVO Chapter 104 Female Contraception ALISA B. GOLDBERG and PHILIP DARNEY Chapter 105 Complications and Side Effects of Steroidal Contraception ALISA B. GOLDBERG and PHILIP DARNEY Chapter 106 Morphology of the Normal Breast, Its Hormonal Control, and Pathophysiology RICHARD E. BLACKWELL Chapter 107 Conception, Implantation, and Early Development PHILIP M. IANNACCONE, DAVID O. WALTERHOUSE, and KRISTINA C. PFENDLER Chapter 108 The Maternal-Fetal-Placental Unit BRUCE R. CARR Chapter 109 Endocrinology of Parturition JOHN R. G. CHALLIS Chapter 110 Endocrine Disease in Pregnancy MARK E. MOLITCH Chapter 111 Trophoblastic Tissue and Its Abnormalities CYNTHIA G. KAPLAN Chapter 112 Endocrinology of Trophoblastic Tissue Z. M. LEI and CH. V. RAO

PART VIII ENDOCRINOLOGY OF THE MALE William J. Bremner, Editor Chapter 113 Morphology and Physiology of the Testis DAVID M. DE KRETSER Chapter 114 Evaluation of Testicular Function STEPHEN J. WINTERS Chapter 115 Male Hypogonadism STEPHEN R. PLYMATE Chapter 116 Testicular Dysfunction in Systemic Disease H. W. GORDON BAKER Chapter 117 Erectile Dysfunction GLENN R. CUNNINGHAM and MAX HIRSHKOWITZ Chapter 118 Male Infertility RICHARD V. CLARK Chapter 119 Clinical Use and Abuse of Androgens and Antiandrogens ALVIN M. MATSUMOTO Chapter 120 Gynecomastia ALLAN R. GLASS Chapter 121 Endocrine Aspects of Benign Prostatic Hyperplasia ELIZABETH A. MILLER and WILLIAM J. ELLIS Chapter 122 Testicular Tumors NIELS E. SKAKKEBAEK and MIKAEL RØRTH Chapter 123 Male Contraception JOHN K. AMORY and WILLIAM J. BREMNER

PART IX DISORDERS OF FUEL METABOLISM C. Ronald Kahn, Editor SECTION A FOOD AND ENERGY Chapter 124 Principles of Nutritional Management ROBERTA P. DURSCHLAG and ROBERT J. SMITH Chapter 125 Appetite ANGELICA LINDÉN HIRSCHBERG Chapter 126 Obesity JULES HIRSCH, LESTER B. SALANS, and LOUIS J. ARONNE Chapter 127 Starvation RUTH S. MACDONALD and ROBERT J. SMITH Chapter 128 Anorexia Nervosa and Other Eating Disorders MICHELLE P. WARREN and REBECCA J. LOCKE Chapter 129 Fuel Homeostasis and Intermediary Metabolism of Carbohydrate, Fat, and Protein NEIL B. RUDERMAN, KEITH TORNHEIM, and MICHAEL N. GOODMAN Chapter 130 Vitamins: Hormonal and Metabolic Interrelationships ALAA ABOU-SAIF and TIMOTHY O. LIPMAN Chapter 131 Trace Minerals: Hormonal and Metabolic Interrelationships ROBERT D. LINDEMAN Chapter 132 Exercise: Endocrine and Metabolic Effects JACQUES LEBLANC

SECTION B DIABETES MELLITUS Chapter 133 Morphology of the Endocrine Pancreas SUSAN BONNER-WEIR Chapter 134 Islet Cell Hormones: Production and Degradation GORDON C. WEIR and PHILIPPE A. HALBAN Chapter 135 Glucose Homeostasis and Insulin Action C. RONALD KAHN Chapter 136 Classification, Diagnostic Tests, and Pathogenesis of Type 1 Diabetes Mellitus GEORGE S. EISENBARTH Chapter 137 Etiology and Pathogenesis of Type 2 Diabetes Mellitus and Related Disorders C. RONALD KAHN Chapter 138 Natural History of Diabetes Mellitus ANDRZEJ S. KROLEWSKI and JAMES H. WARRAM Chapter 139 Secondary Forms of Diabetes Mellitus VERONICA M. CATANESE and C. RONALD KAHN Chapter 140 Evaluation of Metabolic Control in Diabetes ALLISON B. GOLDFINE Chapter 141 Diet and Exercise in Diabetes OM P. GANDA Chapter 142 Oral Agents for the Treatment of Type 2 Diabetes Mellitus ALLISON B. GOLDFINE and ELEFTHERIA MARATOS-FLIER Chapter 143 Insulin Therapy and Its Complications GORDON C. WEIR Chapter 144 Pancreas and Islet Transplantation GORDON C. WEIR Chapter 145 Syndrome X GERALD M. REAVEN Chapter 146 Syndromes of Extreme Insulin Resistance JEFFREY S. FLIER and CHRISTOS S. MANTZOROS Chapter 147 Cardiovascular Complications of Diabetes Mellitus KARIN HEHENBERGER and GEORGE L. KING Chapter 148 Diabetic Neuropathy EVA L. FELDMAN, MARTIN J. STEVENS, JAMES W. RUSSELL, and DOUGLAS A. GREENE Chapter 149 Gastrointestinal Complications of Diabetes FREDERIC D. GORDON and KENNETH R. FALCHUK Chapter 150 Diabetic Nephropathy RALPH A. DEFRONZO Chapter 151 Diabetes and the Eye LAWRENCE I. RAND Chapter 152 Diabetes and Infection GEORGE M. ELIOPOULOS Chapter 153 Diabetes and the Skin ROBERT J. TANENBERG and RICHARD C. EASTMAN Chapter 154 The Diabetic Foot GARY W. GIBBONS Chapter 155 Diabetic Acidosis, Hyperosmolar Coma, and Lactic Acidosis K. GEORGE M. M. ALBERTI Chapter 156 Diabetes Mellitus and Pregnancy LOIS JOVANOVIC Chapter 157 Diabetes Mellitus in the Infant and Child DOROTHY J. BECKER and ALLAN L. DRASH

SECTION C HYPOGLYCEMIA Chapter 158 Hypoglycemic Disorders in the Adult

RICHARD J. COMI and PHILLIP GORDEN Chapter 159 Localization of Islet Cell Tumors DONALD L. MILLER Chapter 160 Surgery of the Endocrine Pancreas JON C. WHITE Chapter 161 Hypoglycemia of Infancy and Childhood JOSEPH I. WOLFSDORF and MARK KORSON

SECTION D LIPID METABOLISM Chapter 162 Biochemistry and Physiology of Lipid and Lipoprotein Metabolism ROBERT W. MAHLEY Chapter 163 Lipoprotein Disorders ERNST J. SCHAEFER Chapter 164 Treatment of the Hyperlipoproteinemias JOHN C. LAROSA Chapter 165 Endocrine Effects on Lipids HENRY N. GINSBERG, IRA J. GOLDBERG, and CATHERINE TUCK Chapter 166 Lipid Abnormalities in Diabetes Mellitus ROBERT E. RATNER, BARBARA V. HOWARD, and WILLIAM JAMES HOWARD

PART X DIFFUSE HORMONAL SECRETION Eric S. Nylén, Editor Chapter 167 General Characteristics of Diffuse Peptide Hormone Systems JENS F. REHFELD Chapter 168 Endogenous Opioid Peptides BRIAN M. COX and GREGORY P. MUELLER Chapter 169 Somatostatin YOGESH C. PATEL Chapter 170 Kinins DOMENICO C. REGOLI Chapter 171 Substance P and the Tachykinins NEIL ARONIN Chapter 172 Prostaglandins, Thromboxanes, and Leukotrienes R. PAUL ROBERTSON Chapter 173 Growth Factors and Cytokines DEREK LEROITH and VICKY A. BLAKESLEY Chapter 174 Compendium of Growth Factors and Cytokines BHARAT B. AGGARWAL Chapter 175 The Diffuse Neuroendocrine System ERIC S. NYLÉN and KENNETH L. BECKER Chapter 176 The Endocrine Brain ABBA J. KASTIN, WEIHONG PAN, JAMES E. ZADINA, and WILLIAM A. BANKS Chapter 177 The Endocrine Lung KENNETH L. BECKER Chapter 178 The Endocrine Heart MIRIAM T. RADEMAKER and ERIC A. ESPINER Chapter 179 The Endocrine Endothelium FRANCESCO COSENTINO and THOMAS F. LÜSCHER Chapter 180 The Endocrine Blood Cells HARISH P. G. DAVE and BEAT MÜLLER Chapter 181 The Endocrine Mast Cell STEPHEN I. WASSERMAN Chapter 182 The Endocrine Enteric System JENS J. HOLST Chapter 183 The Endocrine Kidney ALAN DUBROW and LUCA DESIMONE Chapter 184 The Endocrine Genitourinary Tract JAN FAHRENKRUG and SØREN GRÄS Chapter 185 The Endocrine Skin MARK R. PITTELKOW Chapter 186 The Endocrine Adipocyte REXFORD S. AHIMA and JEFFREY S. FLIER

PART XI HERITABLE ABNORMALITIES OF ENDOCRINOLOGY AND METABOLISM Kenneth L. Becker, Editor Chapter 187 Inheritance Patterns of Endocrinologic and Metabolic Disorders R. NEIL SCHIMKE Chapter 188 Multiple Endocrine Neoplasia GLEN W. SIZEMORE Chapter 189 Heritable Disorders of Collagen and Fibrillin PETER H. BYERS Chapter 190 Heritable Diseases of Lysosomal Storage WARREN E. COHEN Chapter 191 Heritable Diseases of Amino-Acid Metabolism HARVEY J. STERN and JAMES D. FINKELSTEIN Chapter 192 Heritable Diseases of Purine Metabolism EDWARD W. HOLMES and DAVID J. NASHEL

PART XII IMMUNOLOGIC BASIS OF ENDOCRINE DISORDERS Leonard Wartofsky, Editor Chapter 193 The Endocrine Thymus ALLAN L. GOLDSTEIN and NICHOLAS R. S. HALL Chapter 194 Immunogenetics, the Human Leukocyte Antigen System, and Endocrine Disease JAMES R. BAKER, JR. Chapter 195 T Cells in Endocrine Disease ANTHONY PETER WEETMAN Chapter 196 B Cells and Autoantibodies in Endocrine Disease ALAN M. MCGREGOR Chapter 197 The Immune System and Its Role in Endocrine Function ROBERT VOLPÉ

PART XIII ENDOCRINE AND METABOLIC DYSFUNCTION IN THE GROWING CHILD AND IN THE AGED Wellington Hung, Editor Chapter 198 Short Stature and Slow Growth in the Young THOMAS ACETO, JR.,DAVID P. DEMPSHER, LUIGI GARIBALDI, SUSAN E. MYERS, NANCI BOBROW, and COLLEEN WEBER Chapter 199 Endocrinology and Aging DAVID A. GRUENEWALD and ALVIN M. MATSUMOTO

PART XIV INTERRELATIONSHIPS BETWEEN HORMONES AND THE BODY Kenneth L. Becker, Editor Chapter 200 Cerebral Effects of Endocrine Disease HOYLE LEIGH Chapter 201 Psychiatric-Hormonal Interrelationships MITCHEL A. KLING,MARIANNE HATLE, RAMESH K. THAPAR, and PHILIP W. GOLD Chapter 202 Respiration and Endocrinology PRASHANT K. ROHATGI and KENNETH L. BECKER Chapter 203 The Cardiovascular System and Endocrine Disease ELLEN W. SEELY and GORDON H. WILLIAMS Chapter 204 Gastrointestinal Manifestations of Endocrine Disease ALLAN G. HALLINE Chapter 205 The Liver and Endocrine Function NICOLA DE MARIA, ALESSANDRA COLANTONI, and DAVID H. VAN THIEL Chapter 206 Effects of Nonrenal Hormones on the Normal Kidney PAUL L. KIMMEL ANTONIO RIVERA, and PARVEZ KHATRI Chapter 207 Renal Metabolism of Hormones RALPH RABKIN and MICHAEL J. HAUSMANN Chapter 208 Effects of Endocrine Disease on the Kidney ELLIE KELEPOURIS and ZALMAN S. AGUS Chapter 209 Endocrine Dysfunction due to Renal Disease ARSHAG D. MOORADIAN Chapter 210 Neuromuscular Manifestations of Endocrine Disease ROBERT B. LAYZER and GARY M. ABRAMS Chapter 211 Rheumatic Manifestations of Endocrine Disease DAVID J. NASHEL Chapter 212 Hematologic Endocrinology HARVEY S. LUKSENBURG, STUART L. GOLDBERG, and CRAIG M. KESSLER Chapter 213 Infectious Diseases and Endocrinology CARMELITA U. TUAZON and STEPHEN A. MIGUELES Chapter 214 Endocrine Disorders in Human Immunodeficiency Virus Infection STEPHEN A. MIGUELES and CARMELITA U. TUAZON Chapter 215 The Eye in Endocrinology ROBERT A. OPPENHEIM and WILLIAM D. MATHERS Chapter 216 Otolaryngology and Endocrine Disease STEPHEN G. HARNER Chapter 217 Dental Aspects of Endocrinology ROBERT S. REDMAN Chapter 218 The Skin and Endocrine Disorders JO-DAVID FINE, ADNAN NASIR, and KENNETH L. BECKER

PART XV HORMONES AND CANCER Kenneth L. Becker, Editor Chapter 219 Paraneoplastic Endocrine Syndromes KENNETH L. BECKER and OMEGA L. SILVA Chapter 220 Endocrine Tumors of the Gastrointestinal Tract SHAHRAD TAHERI, KARIM MEERAN, and STEPHEN BLOOM Chapter 221 Carcinoid Tumor and the Carcinoid Syndrome PAUL N. MATON Chapter 222 Hormones and Carcinogenesis: Laboratory Studies JONATHAN J. LI and SARA ANTONIA LI Chapter 223 Sex Hormones and Human Carcinogenesis: Epidemiology ROBERT N. HOOVER Chapter 224 Endocrine Treatment of Breast Cancer GABRIEL N. HORTOBAGYI Chapter 225 Endocrine Aspects of Prostate Cancer CHULSO MOON and CHRISTOPHER J. LOGETHETIS Chapter 226 Endocrine Consequences of Cancer Therapy DAIVA R. BAJORUNAS

PART XVI ENDOCRINOLOGY OF CRITICAL ILLNESS Eric S. Nylén, Editor Chapter 227 Critical Illness and Systemic Inflammation GARY P. ZALOGA BANKIM BHATT, and PAUL MARIK Chapter 228 Endocrine Markers and Mediators in Critical Illness ABDULLAH A. ALARIFI, GREET H. VAN DEN BERGHE, RICHARD H. SNIDER, JR., KENNETH L. BECKER, BEAT MÜLLER, and ERIC S. NYLÉN Chapter 229 The Hypothalamic–Pituitary–Adrenal Axis in Stress and Critical Illness STEFAN R. BORNSTEIN and GEORGE P. CHROUSOS Chapter 230 Neuroendocrine Response to Acute Versus Prolonged Critical Illness GREET H. VAN DEN BERGHE Chapter 231 Fuel Metabolism and Nutrient Delivery in Critical Illness THOMAS R. ZIEGLER Chapter 232 Endocrine Therapeutics in Critical Illness ERIC S. NYLÉN, GARY P. ZALOGA, KENNETH L. BECKER, KENNETH D. BURMAN, LEONARD WARTOFSKY, BEAT MÜLLER, JON C. WHITE, and ABDULLAH A. ALARIFI

PART XVII ENDOCRINE AND METABOLIC EFFECTS OF TOXIC AGENTS Kenneth L. Becker, Editor Chapter 233 Endocrine-Metabolic Effects of Alcohol ROBERT H. NOTH and ARTHUR L. M. SWISLOCKI Chapter 234 Metabolic Effects of Tobacco, Cannabis, and Cocaine OMEGA L. SILVA Chapter 235 Environmental Factors and Toxins and Endocrine Function LAURA S. WELCH

PART XVIII ENDOCRINE DRUGS AND VALUES Kenneth L. Becker, Editor Chapter 236 Compendium of Endocrine-Related Drugs DOLLY MISRA, MICHELLE FISCHMANN MAGEE, and ERIC S. NYLÉN Chapter 237 Reference Values in Endocrinology D. ROBERT DUFOUR Chapter 238 Techniques of Laboratory Testing D. ROBERT DUFOUR Chapter 239 Effects of Drugs on Endocrine Function and Values MEETA SHARMA Chapter 240 DNA Diagnosis of Endocrine Disease J. FIELDING HEJTMANCIK and HARRY OSTRER Chapter 241 Dynamic Procedures in Endocrinology D. ROBERT DUFOUR and WILLIAM A. JUBIZ

PREFACE This third edition of Principles and Practice of Endocrinology and Metabolism has been substantially and systematically revised. All of the chapters have been updated, many have been entirely rewritten, and many deal with completely new topics. Furthermore, additional important information and references have been added up until the very date of printing. The new chapters covering topics that did not appear in depth in the prior edition include: Molecular Biology: Present and Future; Pituitary Tumors: Overview of Therapeutic Options; The Incidental Adrenal Mass; Appetite; Pancreas and Islet Transplantation; Syndrome X; Endocrine Effects on Lipids; Compendium of Growth Factors and Cytokines; The Endocrine Blood Cells; The Endocrine Adipocyte; and Endocrine Disorders in Human Immunodeficiency Virus Infection. We would like to welcome the authors of these chapters, and also the new authors who have updated, extensively revised, or have entirely rewritten chapters on topics that appeared in the last edition. A new section has been added to this textbook: Endocrinology of Critical Illness. The six chapters comprising this section address the multiple aspects of this condition in a manner that is unique. Critical illness, which to some extent afflicts the great bulk of humankind at some time in their lives, has enormous hormonal and metabolic dimensions that relate directly to the diagnosis of the illness, influence the response of the host and the consequent evolution of the condition, and play a role in its outcome. The specific chapters include Critical Illness and Systemic Inflammation, Endocrine Markers and Mediators in Critical Illness, The Hypothalamic–Pituitary–Adrenal Axis in Stress and Critical Illness, Neuroendocrine Response to Acute versus Prolonged Critical Illness, Fuel Metabolism and Nutrient Delivery in Critical Illness, and Endocrine Therapeutics in Critical Illness. These subjects are of great importance to every endocrine clinician as well as many who are involved in fundamental endocrine research. Overall, the goal of this textbook is to continue to provide, in a readable, understandable, and well-illustrated format, the clinical and basic information on endocrinology and metabolism that will be useful to both clinicians and basic scientists. We also wish this book to be a useful source of information for internists, house staff, and medical students. We have attempted to cover the field thoroughly and broadly, to include most of the known endocrine and metabolic disorders and hormonal messenger molecules, to furnish appropriate and current references, and to be of practical benefit to our readers. A complete CD version of this entire textbook is available. It contains approximately 4000 self-assessment questions that have been assembled and edited by Dr. Meeta Sharma. I wish to acknowledge the very helpful library assistance of Joanne Bennett. I am very grateful for the indispensable editorial and pharmaceutical aid of my Editorial Assistant, Roberta L. Brown, Pharm. D. Kenneth L. Becker, MD , PhD

PREFACE TO THE FIRST EDITION Although there are several excellent large textbooks of endocrinology, we have felt the need for a book which would aim at encompassing all aspects of the field, a book which would be disease-oriented, would have practical applicability to the care of the adult and pediatric patient, and could be consulted to obtain a broad range of pathophysiologic, diagnostic, and therapeutic information. To fulfill this goal we called upon not only eminent specialists in endocrinology but also upon experts in many fields of medicine and science. The first part of the book surveys general aspects of endocrinology. The eight succeeding parts deal with specific fields of endocrinology: The Endocrine Brain and Pituitary Gland, The Thyroid Gland, Calcium and Bone Metabolism, The Adrenal Glands, Sex Determination and Development, Endocrinology of the Female, Endocrinology of the Male, and Disorders of Fuel Metabolism. Each of these parts contains relevant anatomic, physiologic, diagnostic, and therapeutic information and, when indicated, pediatric coverage of the topic. Diffuse Hormonal Secretion expounds upon the fact that endocrine function is not confined to anatomically discrete endocrine glands but is also intrinsic to all tissues and organs. This part is divided in two; it first presents a discussion of hormones which have a diffuse distribution and are not reviewed elsewhere in the book, and subsequently it deals with body constituents which are important sites of hormonal secretion. Heritable Abnormalities of Endocrinology and Metabolism underlines the importance of genetics in the causation of many endocrine and metabolic abnormalities. Endocrine and metabolic dysfunction in the young and in the aged is the subject of a separate part, because in both of these age groups hormonal function as well as endocrine disorders differ profoundly from those of individuals in their middle decades. Interrelationships Between Hormones and the Body discusses the impact of hormones on the soma and addresses clinical aspects of the disorders they may engender. Hormones and Cancer examines the phenomenon of hormone-induced neoplasms, elaborating on the fact that all neoplasms secrete hormones, that several of these hormones can cause additional clinical disorders, and that some neoplasms respond therapeutically to hormonal manipulation. The ensuing part, entitled Endocrine and Metabolic Effects of Toxic Agents deals with the sometimes subtle, sometimes profound influence of four nearly omnipresent agents: medication, alcohol, tobacco, and cannabis; it also addresses the consequences of environmental toxins on the endocrine system. The last part deals with the therapeutic use of drugs in endocrinology and the proper interpretation of laboratory values. It offers an extensive table on the clinical use of endocrine-related drugs, a table on reference values, and an outline of the dynamic procedures used in endocrinology. The goal of these tabular chapters is to facilitate the day-to-day evaluation and therapy of the endocrine patient. As a rule, the emphasis of this textbook is on the endocrinology of the human being. Animal data are presented only when contributing to a better understanding of human physiology and pathology. To maximize current relevance, historical information is kept to a minimum. While efforts were made to avoid repetition, the coverage of certain topics may recur when viewed from different standpoints. It is hoped that this will provide a wider dimension of the understanding of endocrine and metabolic function and dysfunction. In order not to interrupt continuity, bibliographic references are grouped at the end of each part. Finally, with the interest of the reader in mind, particular attention was given to composing an index as detailed as possible. I wish to thank the associate editors of this text for their skill, their enthusiasm, and their hard work. We all are very grateful for the expertise of our many eminent contributors. During the preparation of the manuscripts, there was considerable inter-communication between these contributors and their respective editors concerning both content and presentation. I wish to acknowledge the participation of Richard H. Snider, PhD, and Eric S. Nylén, MD, who have provided outstanding editorial assistance throughout the preparation of the textbook. The field of endocrinology and metabolism is evolving rapidly. New data are being developed continuously, and with this in mind, all contributors were encouraged to add up-to-date information until nearly the date of publication. There are numerous matters upon which there is no current common agreement, and logical arguments can be marshaled to buttress diametrically different viewpoints. This textbook is written by many authors; though most of the beliefs and conclusions of the contributors tend to reflect those of the editors, no attempt was made to impose a uniformity of pathophysiologic, diagnostic, or therapeutic viewpoints, and the book does not lack for differences of opinion. We hope that the Principles and Practice of Endocrinology and Metabolism will be a relevant sourcebook for those interested in the science and the practice of this fascinating discipline, whether they be clinicians, basic scientists, allied health personnel, or students. Kenneth L. Becker, MD , PhD

CONTRIBUTING AUTHORS Alaa Abou-Saif, MD Gastroenterology Fellow Department of Medicine Division of Gastroenterology Georgetown University School of Medicine Washington, DC Gary M. Abrams, MD Associate Professor of Clinical Neurology Department of Neurology University of California, San Francisco, School of Medicine San Francisco, California Thomas Aceto, Jr., MD Professor of Pediatrics Chairman Emeritus of Pediatrics Saint Louis University School of Medicine Cardinal Glennon Children’s Hospital St. Louis, Missouri Bharat B. Aggarwal, PhD Professor of Medicine and Biochemistry Department of Bioimmunotherapy Chief, Cytokine Research Section University of Texas–Houston Medical School M. D. Anderson Cancer Center Houston, Texas Zalman S. Agus, MD Emeritus Professor of Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Rexford S. Ahima, MD , PhD Assistant Professor of Medicine Division of Endocrinology, Diabetes and Metabolism University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Andrew J. Ahmann, MD Assistant Professor of Medicine Director of Adult Diabetes Services Oregon Health Sciences University School of Medicine Portland, Oregon Abdullah A. Alarifi, MD Consultant Endocrinologist Department of Medicine King Faisal Specialist Hospital and Research Centre Riyadh, Kingdom of Saudi Arabia K. George M. M. Alberti, MD , DPhil, PRCP, FRCP Professor of Medicine Department of Diabetes and Metabolism University of Newcastle upon Tyne Faculty of Medicine Newcastle upon Tyne, England Melvin G. Alper, MD Private Practice, Ophthalmology Chevy Chase, Maryland John K. Amory, MD Assistant Professor Department of Medicine University of Washington School of Medicine Veterans Affairs Puget Sound Health Care System Seattle, Washington Neil Aronin, MD Professor of Medicine and Cell Biology Director, Division of Endocrinology and Metabolism University of Massachusetts Medical School Worcester, Massachusetts Louis J. Aronne, MD Clinical Associate Professor of Medicine Weill Medical College of Cornell University New York, New York Gilbert P. August, MD Professor of Pediatrics Department of Endocrinology George Washington University School of Medicine and Health Sciences Children’s National Medical Center Washington, DC Lloyd Axelrod, MD Associate Professor of Medicine Harvard Medical School Physician and Chief of the James Howard Means Firm Massachusetts General Hospital

Boston, Massachusetts Daiva R. Bajorunas, MD Senior Director, Clinical Research Global Project Team Leader, Metabolism Aventis Pharmaceuticals Bridgewater, New Jersey H. W. Gordon Baker, MD , PhD, FRACP Associate Professor Department of Obstetrics and Gynaecology University of Melbourne School of Medicine Royal Women’s Hospital Victoria, Australia James R. Baker, Jr., MD Professor of Medicine Department of Internal Medicine–Allergy and Immunology Chief, Division of Allergy University of Michigan Medical School Ann Arbor, Michigan William A. Banks, MD Professor of Internal Medicine Division of Geriatrics Saint Louis University School of Medicine St. Louis, Missouri Robert L. Barbieri, MD Kate Macy Ladd Professor of Obstetrics, Gynecology and Reproductive Biology Harvard Medical School Boston, Massachusetts Marcelo J. Barrionuevo, MD Assistant Professor Department of Obstetrics and Gynecology University of Miami School of Medicine Margate, Florida David S. Baskin, MD , FACS Professor of Neurosurgery and Anesthesiology Baylor College of Medicine Houston, Texas Gerhard Baumann, MD Professor of Medicine Northwestern University Medical Center Chicago, Illinois Peter H. Baylis, MD , FRCP, FAMS Professor of Experimental Medicine Dean, Department of Medicine The Medical School University of Newcastle upon Tyne Faculty of Medicine Newcastle upon Tyne, England David V. Becker, MD Professor of Radiology and Medicine Division of Nuclear Medicine and Endocrinology Weill Medical College of Cornell University New York Presbyterian Hospital New York, New York Dorothy J. Becker, MB, Bch Professor of Pediatrics University of Pittsburgh School of Medicine Children’s Hospital of Pittsburgh Pittsburgh, Pennsylvania Kenneth L. Becker, MD , PhD Professor of Medicine Professor of Physiology and Experimental Medicine Director of Endocrinology and Metabolism George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Norman H. Bell, MD Distinguished University Professor of Medicine Medical University of South Carolina College of Medicine Charleston, South Carolina Bankim Bhatt, MD Medical Resident Department of Medicine Georgetown University School of Medicine Washington Hospital Center Washington, DC John P. Bilezikian, MD Professor of Medicine and Pharmacology Department of Medicine Columbia University College of Physicians and Surgeons New York, New York Richard E. Blackwell, MD , PhD

Professor of Obstetrics and Gynecology University of Alabama School of Medicine Birmingham, Alabama Vicky A. Blakesley, MD , PhD Director, Department of New Product Evaluation International Division Abbott Laboratories Abbott Park, Illinois Stephen Bloom, MA, MD , DSc, FRCPath, FRCP, FMedSci Professor of Medicine Department of Metabolic Medicine Division of Investigative Science University of London Imperial College School of Medicine London, England Manfred Blum, MD Professor of Clinical Medicine and Radiology Director, Nuclear Endocrine Laboratory New York University School of Medicine New York, New York Nanci Bobrow, PhD Assistant Clinical Professor of Pediatrics Cardinal Glennon Children’s Hospital Saint Louis University School of Medicine St. Louis, Missouri Susan Bonner-Weir, PhD Associate Professor of Medicine Harvard Medical School Senior Investigator Joslin Diabetes Center Boston, Massachusetts Stefan R. Bornstein, MD , PhD Assistant Professor and Research Scholar Pediatric and Reproductive Endocrinology Branch National Institute of Child Health and Human Development National Institutes of Health Bethesda, Maryland Eric Bourekas, MD Assistant Professor of Radiology Section of Diagnostic and Interventional Neuroradiology Ohio State University College of Medicine and Public Health Columbus, Ohio William J. Bremner, MD , PhD Robert G. Petersdorf Professor and Chairman Department of Medicine University of Washington School of Medicine Seattle, Washington Edward M. Brown, MD Professor of Medicine Endocrine–Hypertension Division Harvard Medical School Brigham and Women’s Hospital Boston, Massachusetts Henry B. Burch, MD Associate Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Department of Endocrine-Metabolic Service Walter Reed Army Medical Center Washington, DC Kenneth D. Burman, MD Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Clinical Professor Department of Medicine George Washington University School of Medicine and Health Sciences Professor of Medicine Georgetown University School of Medicine Chief, Endocrine Section Washington Hospital Center Washington, DC Peter H. Byers, MD Professor of Pathology and Medicine University of Washington School of Medicine Seattle, Washington Enrico Carmina, MD Professor Department of Endocrinology University of Palermo Palermo, Italy Visiting Professor Department of Obstetrics and Gynecology Columbia University College of Physicians and Surgeons

New York, New York Thomas O. Carpenter, MD Professor of Pediatrics Yale University School of Medicine Yale–New Haven Hospital New Haven, Connecticut Bruce R. Carr, MD Professor Paul C. Macdonald Distinguished Chair in Obstetrics and Gynecology Director, Division of Reproductive Endocrinology University of Texas Southwestern Medical Center at Dallas Southwestern Medical School Dallas, Texas Veronica M. Catanese, MD Assistant Professor Department of Medicine and Cell Biology New York University School of Medicine New York, New York Donald Chakeres, MD Professor of Radiology Ohio State University College of Medicine and Public Health Columbus, Ohio John R. G. Challis, PhD, Dsc, FIBiol , FRCOG, FRSC Department of Physiology Medical Sciences Building University of Toronto Faculty of Medicine Toronto, Ontario Canada Philippe Chanson, MD Professor of Medicine Department of Endocrinology University Paris XI Bicêtre University Hospital Le Kremlin-Bicêtre France William W. Chin, MD Professor of Medicine Harvard Medical School Boston, Massachusetts Vice President, Lilly Research Laboratories Eli Lilly & Co. Lilly Corporate Center Indianapolis, Indiana George P. Chrousos, MD Chief, Pediatric and Reproductive Endocrinology Branch National Institutes of Health Bethesda, Maryland Richard V. Clark, MD , PhD Principal Clinical Research Physician Clinical Pharmacology–Exploratory Department Glaxo Wellcome Research and Development Research Triangle Park, North Carolina Thomas L. Clemens, MD , PhD Professor of Medicine and Molecular and Cellular Physiology Department of Internal Medicine/Endocrinology University of Cincinnati College of Medicine Cincinnati, Ohio Fredric L. Coe, MD Professor Departments of Medicine and Physiology University of Chicago Pritzker School of Medicine Chicago, Illinois Joshua L. Cohen, MD Associate Professor of Medicine Department of Endocrinology George Washington University School of Medicine and Health Sciences Washington, DC Régis Cohen, MD , PhD Praticien Hospitalier Endocrine Staff Physician Avicenne Hospital Bobigny, France University of Leonardo Da Vinci Paris, France Warren E. Cohen, MD Associate Clinical Professor of Pediatrics and Neurology George Washington University School of Medicine and Health Sciences Washington, DC Medical Director, United Cerebral Palsy Nassau County, New York Alessandra Colantoni, MD Assistant Professor of Medicine

Department of Gastroenterology and Hepatology Loyola University of Chicago Stritch School of Medicine Loyola University Medical Center Maywood, Illinois Richard J. Comi, MD Associate Professor of Medicine Section of Endocrinology and Metabolism Dartmouth Medical School Dartmouth–Hitchcock Medical Center Hanover, New Hampshire Paul E. Cooper, MD , FRCPC Associate Professor of Neurology Departments of Clinical Neurological Sciences and Medicine University of Western Ontario Faculty of Medicine and Dentistry Health Sciences Addition London, Ontario Canada Dalila B. Corry, MD Associate Clinical Professor of Medicine Department of Medicine University of California, Los Angeles, UCLA School of Medicine Los Angeles, California Chief, Nephrology Olive View Medical Center Sylmar, California Francesco Cosentino, MD , PhD Assistant Professor Department of Experimental Medicine and Pathology University “La Sapienza” Rome, Italy Senior Research Associate Cardiovascular Research Department of Cardiology University Hospital Zurich, Switzerland Felicia Cosman, MD Associate Professor of Clinical Medicine Department of Medicine Columbia University College of Physicians and Surgeons New York, New York Helen Hayes Hospital West Haverstraw, New York Brian M. Cox, PhD Professor of Pharmacology and Neuroscience Department of Pharmacology Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Glenn R. Cunningham, MD Associate Chief of Staff Department of Medicine University of Texas–Houston Medical School Veterans Affairs Medical Center Houston, Texas Mary F. Dallman, PhD Professor of Physiology University of California, San Francisco, School of Medicine San Francisco, California Daniel N. Darlington, PhD Associate Professor of Surgery Departments of Surgery and Physiology University of Maryland School of Medicine Baltimore, Maryland Philip Darney, MD , MSc Professor of Obstetrics, Gynecology, and Reproductive Sciences University of California, San Francisco, School of Medicine San Francisco General Hospital San Francisco, California Harish P. G. Dave, MB, ChB, MRCP(UK) Associate Professor of Medicine Department of Hematology George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Faith B. Davis, MD Professor of Medicine and Cell Biology and Cancer Research Albany Medical College Staff Physician Stratton Veterans Affairs Medical Center Albany, New York Paul J. Davis, MD Professor of Medicine and Cell Biology and Cancer Research Senior Associate Dean for Clinical Research Albany Medical College Research Physician

Wadsworth Center, New York State Department of Health Staff Physician Stratton Veterans Affairs Medical Center Albany, New York Suzanne M. Jan De Beur, MD Assistant Professor of Medicine Johns Hopkins University School of Medicine Baltimore, Maryland Ralph A. DeFronzo, MD Professor of Medicine Chief, Diabetic Division Member, Nephrology Division University of Texas Medical School at San Antonio University Health Center San Antonio, Texas David M. De Kretser, MD , MBBS, FRACP Professor and Director Monash Institute of Reproduction and Development Monash University Monash Medical Centre, Clayton Clayton, Victoria Australia Nicola De Maria, MD Research Associate Liver Transplant Program Loyola University Medical Center Maywood, Illinois David P. Dempsher, MD , PhD Associate Professor of Pediatrics Cardinal Glennon Children’s Hospital Saint Louis University School of Medicine St. Louis, Missouri David W. Dempster, PhD Professor of Clinical Pathology Columbia University College of Physicians and Surgeons New York, New York Director, Regional Bone Center Helen Hayes Hospital West Haverstraw, New York Luca deSimone Nephrology Fellow Beth Israel Medical Center New York, New York Gerard M. Doherty, MD Associate Professor of Surgery Section of Surgical Oncology and Endocrinology Washington University School of Medicine St. Louis, Missouri Allan L. Drash, MD Emeritus Professor of Pediatrics University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Marc K. Drezner, MD Professor of Medicine Head, Section of Endocrinology, Diabetes, and Metabolism University of Wisconsin Medical School Madison, Wisconsin Alan Dubrow, MD Clinical Assistant Professor of Medicine Department of Nephrology Beth Israel Deaconess Medical Center New York, New York D. Robert Dufour, MD Clinical Professor of Pathology George Washington University School of Medicine and Health Sciences Washington, DC Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Chief, Pathology and Laboratory Medicine Service Veterans Affairs Medical Center Washington, DC Roberta P. Durschlag, PhD, RD Clinical Assistant Professor Department of Health Sciences Boston University School of Medicine Boston, Massachusetts Richard C. Eastman, MD Cygnus, Inc. Redwood City, California George S. Eisenbarth, MD , PhD Professor of Pediatrics, Immunology, and Medicine

University of Colorado Health Sciences Center Barbara Davis Center for Childhood Diabetes Denver, Colorado George M. Eliopoulos, MD Associate Professor of Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts William J. Ellis, MD Associate Professor and Clinic Director Department of Urology University of Washington School of Medicine Seattle, Washington Abby Erickson, BA Colorado Center for Bone Research Lakewood, Colorado Gregory F. Erickson, PhD Professor Department of Reproductive Medicine University of California, San Diego, School of Medicine La Jolla, California Eric A. Espiner, MD , FRACP, FRS(NZ) Professor Department of Endocrinology University of Otago Christchurch School of Medicine Christchurch Public Hospital Christchurch, New Zealand Jan Fahrenkrug, MD , DMSci Professor Department of Clinical Chemistry University of Copenhagen Faculty of Health Sciences Bispebjerg Hospital Copenhagen, Denmark Kenneth R. Falchuk, MD Associate Professor of Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Murray J. Favus, MD Professor of Medicine University of Chicago Pritzker School of Medicine Chicago, Illinois Eva L. Feldman, MD , PhD Professor of Neurology University of Michigan Medical School Ann Arbor, Michigan Jo-David Fine, MD , MPH Professor Department of Dermatology University of North Carolina at Chapel Hill School of Medicine Chapel Hill, North Carolina James D. Finkelstein, MD Senior Clinician Department of Medicine Veterans Affairs Medical Center Washington, DC Jeffrey S. Flier, MD George C. Reisman Professor of Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Ruth C. Fretts, MD , MPH Assistant Professor Department of Obstetrics and Gynecology Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Om P. Ganda, MD Associate Clinical Professor Department of Medicine Harvard Medical School Joslin Diabetes Center Beth Israel Deaconess Medical Center Boston, Massachusetts Luigi Garibaldi Beth Israel Medical Center Newark, New Jersey Gary W. Gibbons, MD Associate Clinical Professor of Surgery

Harvard Medical School Director, Quality Improvement Department of Surgery Beth Israel Deaconess Medical Center Boston, Massachusetts John R. Gill, Jr., MD Scientist, Emeritus Hypertension-Endocrine Branch National Heart, Lung, and Blood Institute National Institutes of Health Bethesda, Maryland Henry N. Ginsberg, MD Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Joel S. Glaser, MD Professor Departments of Ophthalmology and Neurology University of Miami School of Medicine Bascom Palmer Eye Institute Miami, Florida Department of Ophthalmology Cleveland Clinic of Florida Coral Gables, Florida Allan R. Glass, MD Adjunct Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Philip W. Gold, MD Branch Chief Department of Intramural Research Programs National Institute of Mental Health National Institutes of Health Bethesda, Maryland Alisa B. Goldberg, MD Assistant Adjunct Professor Department of Obstetrics, Gynecology and Reproductive Sciences University of California, San Francisco, School of Medicine San Francisco General Hospital San Francisco, California Ira J. Goldberg, MD Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Stuart L. Goldberg, MD Assistant Director, Bone Marrow Transplantation Program Temple University School of Medicine Philadelphia, Pennsylvania Allison B. Goldfine, MD Instructor of Medicine Department of Cellular and Molecular Physiology Harvard University Joslin Diabetes Center Boston, Massachusetts Allan L. Goldstein, PhD Chair, Department of Biochemistry and Molecular Biology George Washington University School of Medicine and Health Sciences Washington, DC David S. Goldstein, MD , PhD Chief, Clinical Neurocardiology Section National Institutes of Health Bethesda, Maryland David Goltzman, MD Professor of Medicine and Physiology McGill University Faculty of Medicine Royal Victoria Hospital Montreal, Quebec Canada Esther A. Gonzalez, MD Assistant Professor Division of Nephrology Saint Louis University School of Medicine St. Louis, Missouri Michael N. Goodman, PhD Professor of Medicine Department of Internal Medicine University of California, Davis, School of Medicine Sacramento, California Phillip Gorden, MD Director Emeritus National Institute of Diabetes and Digestive and Kidney Diseases

National Institutes of Health Bethesda, Maryland Frederic D. Gordon, MD Assistant Professor of Medicine Department of Hepatobiliary Surgery and Liver Transplantation Tufts University School of Medicine Lahey Clinic Medical Center Boston, Massachusetts Daryl K. Granner, MD Joe C. Davis Professor of Biomedical Science Professor of Molecular Physiology, Biophysics, and Internal Medicine Vanderbilt University School of Medicine Director, Vanderbilt Diabetes Center Staff Physician Veterans Affairs Hospital Nashville, Tennessee Søren Gräs, MD Senior Registrar Department of Obstetrics and Gynaecology Herlev University Hospital Herlev, Denmark Douglas A. Greene, MD Executive Vice President Department of Clinical Sciences and Product Development Merck & Co., Inc. Rahway, New Jersey David A. Gruenewald, MD , FACP Assistant Professor of Medicine University of Washington School of Medicine Veterans Affairs Puget Sound Health Care System Seattle, Washington Joel F. Habener, MD Professor of Medicine Laboratory of Molecular Endocrinology Harvard Medical School Massachusetts General Hospital Boston, Massachusetts Philippe A. Halban, PhD Professor of Medicine Louis-Jeantet Research Laboratories Geneva University Medical Center Geneva, Switzerland Nicholas R. S. Hall, PhD Health and Human Performance Orlando, Florida Allan G. Halline, MD Assistant Professor of Medicine Section of Digestive and Liver Diseases University of Illinois at Chicago College of Medicine Chicago, Illinois Stephen G. Harner, MD Professor of Otolaryngology Department of Otolaryngology Mayo Medical School Rochester, Minnesota Marianne Hatle, MD Resident University of Maryland School of Medicine Baltimore, Maryland Michael J. Hausmann, MD Professor Department of Nephrology Faculty of Health Sciences Ben Gurion University of the Negev Scroka Medical Center of Kupat Holim Beer Sheva, Israel Karin Hehenberger, MD , PhD Research Fellow Joslin Diabetes Center Harvard Medical School Boston, Massachusetts J. Fielding Hejtmancik, MD , PhD Medical Officer National Eye Institute National Institutes of Health Bethesda, Maryland Geoffrey N. Hendy, PhD Professor of Medicine McGill University Faculty of Medicine Royal Victoria Hospital Montreal, Quebec

Canada James V. Hennessey, MD Associate Professor of Medicine Division of Endocrinology Brown University School of Medicine Rhode Island Hospital Providence, Rhode Island Jules Hirsch, MD Professor Emeritus and Physician-in-Chief Emeritus Laboratory of Human Behavior and Metabolism Rockefeller University Rockefeller University Hospital New York, New York Angelica Lindén Hirschberg, MD , PhD Associate Professor of Obstetrics and Gynecology Karolinska Institute Karolinska Hospital Stockholm, Sweden Max Hirshkowitz, MD Associate Professor Department of Psychiatry Baylor College of Medicine Director, Sleep Center Houston Veterans Affairs Medical Center Houston, Texas Gary D. Hodgen, PhD Professor Department of Obstetrics and Gynecology Eastern Virginia Medical School Chair The Howard and Georgeanna Jones Institute for Reproductive Medicine Norfolk, Virginia Edward W. Holmes, MD Chairman, Department of Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Jens J. Holst, MD Department of Medical Physiology University of Copenhagen Faculty of Health Sciences The Panum Institute Copenhagen, Denmark Robert N. Hoover, MD , ScD Director, Epidemiology and Biostatistics Program National Cancer Institute National Institutes of Health Bethesda, Maryland Gabriel N. Hortobagyi, MD Professor of Medicine, Chairman Department of Breast and Gynecologic Medical Oncology University of Texas–Houston Medical School M. D. Anderson Cancer Center Houston, Texas Eva Horvath, PhD Associate Professor of Pathology Department of Laboratory Medicine Division of Pathology University of Toronto Faculty of Medicine St. Michael’s Hospital Toronto, Ontario Canada Barbara V. Howard, PhD President, MedStar Clinical Research Institute Washington, DC William James Howard, MD Professor of Medicine George Washington University School of Medicine Senior Vice President and Medical Director Washington Hospital Center Washington, DC Ilpo Huhtaniemi, MD , PhD Professor of Physiology University of Turku Faculty of Medicine Turku, Finland Wellington Hung, MD , PhD Professor Emeritus of Pediatrics Georgetown University School of Medicine Professorial Lecturer in Pediatrics George Washington University School of Medicine and Health Sciences Washington, DC Mehboob A. Hussain, MD Department of Medicine

New York University School of Medicine New York, New York Philip M. Iannaccone, MD , PhD George M. Eisenberg Professor Department of Pediatrics Northwestern University Medical School Children’s Memorial Institute of Education and Research Chicago, Illinois Ivor M. D. Jackson, MB, ChB Professor of Medicine Division of Endocrinology Brown University School of Medicine Rhode Island Hospital Providence, Rhode Island Richard V. Jackson, MBBS, FRACP Associate Professor of Medicine University of Queensland Faculty of Health Sciences Greenslopes Private Hospital Queensland, Australia Lois Jovanovic, MD Clinical Professor of Medicine University of Southern California School of Medicine Los Angeles, California Director and Chief Scientific Officer Sansum Medical Research Institute Santa Barbara, California William A. Jubiz, MD Director Endocrinology Center Cali, Colombia C. Ronald Kahn, MD Mary K. Iacocca Professor of Medicine Harvard Medical School President and Director, Research Division Joslin Diabetes Center Boston, Massachusetts Cynthia G. Kaplan, MD Associate Professor of Pathology SUNY at Stony Brook School of Medicine Health Sciences Center Stony Brook, New York Edwin L. Kaplan, MD , FACS Professor of Surgery University of Chicago Pritzker School of Medicine Chicago, Illinois Abba J. Kastin, MD Chief of Endocrinology Departments of Medicine and Neuroscience Tulane University School of Medicine Veterans Affairs Medical Center New Orleans, Louisiana Laurence Katznelson, MD Assistant Professor of Medicine Harvard Medical School Massachusetts General Hospital Boston, Massachusetts Harry R. Keiser, MD Scientist Emeritus National Heart, Lung, and Blood Institute Clinical Center National Institutes of Health Bethesda, Maryland Ellie Kelepouris, MD Professor of Medicine Temple University School of Medicine Philadelphia, Pennsylvania Craig M. Kessler, MD Professor of Medicine and Pathology Chief, Division of Hematology-Oncology Georgetown University School of Medicine Lombardy Cancer Center Washington, DC Parvez Khatri, MD Fellow, Department of Medicine/Nephrology George Washington University School of Medicine and Health Sciences Washington, DC Paul L. Kimmel, MD Professor of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Director, Diabetic Nephropathy Program Division of Kidney, Urologic, and Hematologic Diseases National Institute of Diabetes and Digestive and Kidney Diseases

National Institutes of Health Bethesda, Maryland George L. King, MD Professor of Medicine Acting Director of Research Joslin Diabetes Center Harvard Medical School Boston, Massachusetts Anne Klibanski, MD Professor of Medicine Harvard Medical School Chief, Neuroendocrine Unit Massachusetts General Hospital Boston, Massachusetts Mitchel A. Kling, MD Associate Professor of Psychiatry and Medicine University of Maryland School of Medicine Veterans Affairs Medical Center Baltimore, Maryland Mark Korson, MD Associate Professor of Pediatrics Division of Genetics Tufts University School of Medicine New England Medical Center Boston, Massachusetts Kalman Kovacs, MD , PhD Professor of Pathology Department of Laboratory Medicine Division of Pathology University of Toronto Faculty of Medicine Saint Michael’s Hospital Toronto, Ontario Canada Andrzej S. Krolewski, MD , PhD Associate Professor of Medicine Chief, Section of Genetics and Epidemiology Harvard Medical School Research Division Joslin Diabetes Center Boston, Massachusetts Robert J. Kurman, MD Richard W. TeLinde Distinguished Professor of Gynecologic Pathology Departments of Gynecology, Obstetrics, and Pathology Johns Hopkins University School of Medicine Baltimore, Maryland John C. LaRosa, MD , FACP President SUNY Downstate Medical Center College of Medicine University Hospital of Brooklyn Brooklyn, New York Robert B. Layzer, MD Professor Emeritus of Neurology University of California, San Francisco, School of Medicine San Francisco, California Jacques LeBlanc, MD Professor Emeritus of Physiology Université Laval Faculty of Medicine Quebec City, Canada Peter A. Lee, MD , PhD Professor of Pediatrics Pennsylvania State University College of Medicine The Milton S. Hershey Medical Center Hershey, Pennsylvania Z. M. Lei, MD , PhD Assistant Professor of Obstetrics and Gynecology University of Louisville School of Medicine Louisville, Kentucky Hoyle Leigh, MD Professor of Psychiatry University of California, San Francisco, School of Medicine San Francisco, California Derek LeRoith, MD , PhD Chief, Molecular and Cellular Endocrinology Branch National Institute of Diabetes and Digestive and Kidney Diseases National Institutes of Health Bethesda, Maryland Michael A. Levine, MD Professor of Pediatrics, Medicine, and Pathology Director, Pediatric Endocrinology Johns Hopkins University School of Medicine

Baltimore, Maryland Jonathan J. Li, PhD Director, Division of Etiology and Prevention of Hormone-Associated Cancers Professor of Pharmacology, Toxicology and Preventive Medicine University of Kansas School of Medicine Kansas Cancer Institute Kansas City, Kansas Sara Antonia Li, MD Associate Director Hormonal Carcinogenesis Laboratory University of Kansas School of Medicine Kansas Cancer Institute Kansas City, Kansas Robert D. Lindeman, MD Professor Emeritus of Medicine Department of Internal Medicine University of New Mexico School of Medicine University of New Mexico Hospital Albuquerque, New Mexico Robert Lindsay, MBChB, PhD, FRCP Professor of Clinical Medicine Columbia University College of Physicians and Surgeons New York, New York Chief of Internal Medicine Helen Hayes Hospital West Haverstraw, New York Timothy O. Lipman, MD Professor of Medicine Georgetown University School of Medicine Chief, Gastroenterology–Hepatology Nutrition Section Veterans Affairs Medical Center Washington, DC Virginia A. Livolsi, MD Professor of Pathology Department of Pathology and Laboratory Medicine University of Pennsylvania School of Medicine Philadelphia, Pennsylvania Rogerio A. Lobo, MD Willard C. Rappleye Professor of Obstetrics and Gynecology Chairman, Department of Obstetrics and Gynecology Columbia University College of Physicians and Surgeons Columbia Presbyterian Medical Center Director, Sloane Hospital for Women New York, New York Rebecca J. Locke Research Assistant Columbia University College of Physicians and Surgeons New York, New York Christopher J. Logethetis, MD Chairman and Professor Department of Genitourinary Medical Oncology University of Texas–Houston Medical School M. D. Anderson Cancer Center Houston, Texas D. Lynn Loriaux, MD , PhD Professor and Chair Department of Medicine Oregon Health Sciences University School of Medicine Portland, Oregon Harvey S. Luksenburg, MD Assistant Professor of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Thomas F. Lüscher, MD Professor and Head of Cardiology Hospital Universitaire de Zurich Zurich, Switzerland Ruth S. MacDonald, RD, PhD Professor of Nutrition Department of Food Science and Human Nutrition University of Missouri–Columbia School of Medicine Columbia, Missouri Michelle Fischmann Magee, MD , MB, BCh, BAO Medical Director, Diabetes Team MedStar Clinical Research Institute Washington Hospital Center Washington, DC Robert W. Mahley, MD , PhD Professor of Pathology and Medicine Director, Gladstone Institute of Cardiovascular Disease

University of California, San Francisco, School of Medicine San Francisco, California Christos S. Mantzoros, MD , Dsc Assistant Professor of Medicine Department of Internal Medicine Harvard Medical School Beth Israel Deaconess Medical Center Boston, Massachusetts Eleftheria Maratos-Flier, MD Associate Professor of Medicine Research Division Harvard Medical School Joslin Diabetes Center Boston, Massachusetts Paul Marik, MBBCh, FCP(SA), FRCP(C), FCCM, FCCP Department of Critical Care Mercy Hospital of Pittsburgh Pittsburgh, Pennsylvania Kevin J. Martin, MB, BCh, FACP Professor of Internal Medicine Department of Nephrology Director, Division of Nephrology Saint Louis University School of Medicine St. Louis, Missouri William D. Mathers, MD Professor of Ophthalmology Oregon Health Sciences University School of Medicine Casey Eye Institute Portland, Oregon Paul N. Maton, MD , FRCP, FACP, FACG Digestive Disease Specialists Incorporated Digestive Disease Research Institute Oklahoma City, Oklahoma Alvin M. Matsumoto, MD Professor Department of Medicine University of Washington School of Medicine Chief of Gerontology Veterans Affairs Puget Sound Health Care System Seattle, Washington Ernest L. Mazzaferri, MD , MACP Professor Emeritus and Chairman Department of Internal Medicine Ohio State University College of Medicine and Public Health Columbus, Ohio Alan M. McGregor, MA, MD , FRCP Professor of Medicine King’s College Guy’s, King’s and St. Thomas’ School of Medicine London, England Karim Meeran, MD , MRCP Senior Lecturer Division of Endocrinology and Metabolism University of London Imperial College School of Medicine Hammersmith Hospital London, England Minesh P. Mehta, MD , MB, ChB Associate Professor and Chairman Department of Human Oncology University of Wisconsin Medical School Madison, Wisconsin James C. Melby, MD Professor of Medicine and Physiology Boston University School of Medicine Boston Medical Center Boston, Massachusetts Stephen A. Migueles, MD Fellow, Infectious Diseases Laboratory of Immunoregulation National Institute of Allergy and Infectious Diseases National Institutes of Health Bethesda, Maryland Donald L. Miller, MD Professor of Radiology and Nuclear Medicine Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Elizabeth A. Miller Urology Resident University of Washington School of Medicine Seattle, Washington

Paul D. Miller, MD Clinical Professor Department of Medicine University of Colorado Health Sciences Center Denver, Colorado Dolly Misra, MD Assistant Clinical Professor of Medicine Division of Endocrinology and Metabolism George Washington University School of Medicine and Health Sciences Washington, DC Diabetes and Endocrine Consultants Waldorf, Maryland Mark E. Molitch, MD Professor of Medicine Center for Endocrinology, Metabolism, and Molecular Medicine Northwestern University Medical School Chicago, Illinois Chulso Moon, MD , PhD Clinical Fellow Department of Medicine University of Texas–Houston Medical School M. D. Anderson Cancer Center Houston, Texas Arshag D. Mooradian, MD Professor of Medicine Director of Endocrinology, Diabetes and Metabolism Saint Louis University School of Medicine St. Louis, Missouri Gregory P. Mueller, PhD Professor of Physiology Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Beat Müller, MD Department of Internal Medicine Division of Endocrinology University Hospitals Basel, Switzerland Susan E. Myers, MD Assistant Professor of Pediatrics Saint Louis University School of Medicine Cardinal Glenn Children’s Hospital St. Louis, Missouri David J. Nashel, MD Professor of Medicine Georgetown University School of Medicine Chief of Medical Service Veterans Affairs Medical Center Washington, DC Adnan Nasir, MD , PhD Department of Dermatology University of North Carolina at Chapel Hill School of Medicine Chapel Hill, North Carolina Jeffrey A. Norton, MD Professor of Surgery Vice Chairman, Department of Surgery University of California, San Francisco, School of Medicine San Francisco Veterans Affairs Medical Center San Francisco, California Robert H. Noth, MD Associate Professor of Medicine Department of Internal Medicine University of California, Davis, School of Medicine Davis, California Veterans Affairs Outpatient Clinic Martinez, California Jennifer A. Nuovo Endocrinologist MedClinic of Sacramento Sacramento, California Eric S. Nylén, MD Associate Professor of Medicine Department of Endocrinology George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Donna M. Arab O’Brien, MD Department of Medicine Division of Endocrinology St. Joseph’s Health Centre Toronto, Ontario

Canada Mary Oehler, MD Staff Radiologist Mount Carmel East Hospital New Albany, Ohio Robert A. Oppenheim, MD Naperville Eye Associates Naperville, Illinois Jeffrey L. H. O’Riordan Emeritus Professor of Metabolic Medicine University College London, United Kingdom Steven J. Ory, MD Clinical Associate Professor of Obstetrics and Gynecology University of Miami School of Medicine Miami, Florida Harry Ostrer, MD Associate Professor of Pediatrics and Pathology Human Genetics Program New York University School of Medicine New York, New York Weihong Pan, MD , PhD Assistant Professor of Medicine Tulane University School of Medicine New Orleans, Louisiana Yogesh C. Patel, MD , PhD, FACP, FRCP(C), FRACP, FRSC Professor of Medicine Director, Division of Endocrinology and Metabolism McGill University Faculty of Medicine Royal Victoria Hospital Montreal, Quebec Canada Gary R. Peplinski, MD Surgical Service San Francisco Veterans Affairs Medical Center San Francisco, California Ora Hirsch Pescovitz, MD Professor of Pediatrics, Physiology, and Biophysics Department of Pediatric Endocrinology Indiana University School of Medicine James Whitcomb Riley Hospital for Children Indianapolis, Indiana Kristina C. Pfendler, MD Postdoctoral Scholar Department of Obstetrics and Gynecology University of California, San Francisco, School of Medicine San Francisco, California Joseph J. Pinzone, MD Assistant Professor of Medicine Department of Internal Medicine George Washington University School of Medicine and Health Sciences Washington, DC Mark R. Pittelkow, MD Professor of Dermatology, Biochemistry, and Molecular Biology Mayo Medical School Consultant, Department of Dermatology Mayo Clinic Rochester, Minnesota Stephen R. Plymate, MD Research Professor of Medicine University of Washington School of Medicine Veterans Affairs Puget Sound Health Care System Seattle, Washington Ke-Nan Qin, MD Fellow of Pediatric Endocrinology Department of Pediatrics University of Chicago Pritzker School of Medicine University of Chicago Children’s Hospital Chicago, Illinois Ralph Rabkin, MB, Bch, MD Professor of Medicine and Nephrology Department of Medicine Stanford University School of Medicine Stanford, California Veterans Affairs Palo Alto Health Care System Palo Alto, California Miriam T. Rademaker, PhD Professor of Medicine University of Otago Christchurch School of Medicine

Christchurch, New Zealand Lawrence G. Raisz, MD Professor of Medicine Department of Endocrinology University of Connecticut School of Medicine University of Connecticut Health Center Farmington, Connecticut Lawrence I. Rand, MD Clinical Assistant Professor of Ophthalmology Harvard Medical School Boston, Massachusetts Ch. V. Rao, PhD Professor and Director Department of Obstetrics and Gynecology University of Louisville School of Medicine Louisville, Kentucky Robert E. Ratner, MD Associate Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Director, MedStar Clinical Research Institute Washington, DC Gerald M. Reaven, MD Professor of Medicine Stanford University School of Medicine Stanford, California Robert W. Rebar, MD Professor Department of Obstetrics and Gynecology University of Cincinnati College of Medicine Chief, Obstetrics and Gynecology University Hospital Cincinnati, Ohio Associate Executive Director American Society for Reproductive Medicine Birmingham, Alabama Robert S. Redman, DDS, MSD, PhD Chief, Oral Diagnosis Section, Dental Service Veterans Affairs Medical Center Washington, DC Clinical Associate Professor Department of Oral and Maxillofacial Pathology University of Maryland School of Medicine Baltimore College of Dental Surgery Baltimore, Maryland H. Lester Reed, MD Clinical Professor of Medicine University of Auckland Faculty of Medical and Health Sciences Middlemore Hospital Auckland, New Zealand Domenico C. Regoli, MD Professor Emeritus Department of Pharmacology Universite de Sherbrooke Faculte de Medecine Sherbrooke, Quebec Canada Jens F. Rehfeld, MD , DSc Professor of Clinical Biochemistry University of Copenhagen Faculty of Health Sciences Copenhagen University Hospital Copenhagen, Denmark Robert L. Reid, MD , FRCS(C) Professor Department of Obstetrics and Gynaecology Queen’s University School of Medicine Faculty of Health Sciences Kingston General Hospital Kingston, Ontario Canada Russel J. Reiter, PhD Professor of Neuroendocrinology Department of Cellular and Structural Biology University of Texas Medical School at San Antonio University Health Center San Antonio, Texas Matthew D. Ringel, MD Assistant Professor of Medicine Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Assistant Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Section of Endocrinology Washington Hospital Center

Washington, DC Antonio Rivera, MD Fellow, Department of Medicine Section of Renal Diseases and Hypertension George Washington University Medical Center Washington, DC Gary L. Robertson, MD Professor of Medicine and Neurology Department of Endocrinology Northwestern University Medical School Chicago, Illinois R. Paul Robertson, MD Professor of Medicine and Pharmacology Scientific Director, Pacific Northwest Research Institute Seattle, Washington Simon P. Robins, PhD, Dsc Head, Skeletal Research Unit Rowett Research Institute Aberdeen, Scotland Alan D. Rogol, MD , PhD Professor of Clinical Pediatrics Department of Pediatrics University of Virginia School of Medicine University of Virginia Medical Center Charlottesville, Virginia Clinical Professor of Internal Medicine Virginia Commonwealth University School of Medicine Richmond, Virginia Prashant K. Rohatgi, MB, MD Professor of Medicine George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Mikael Rørth, MD Professor of Clinical Oncology University of Copenhagen Faculty of Health Sciences Rigshospitalet Copenhagen, Denmark Robert L. Rosenfield, MD Professor of Pediatrics and Medicine Department of Pediatric Endocrinology University of Chicago Pritzker School of Medicine Chicago, Illinois Robert K. Rude, MD Professor of Medicine University of Southern California School of Medicine Los Angeles, California Neil B. Ruderman, MD , DPhil Professor Department of Medicine and Physiology Boston University School of Medicine Boston, Massachusetts James W. Russell, MD Assistant Professor Department of Neurology University of Michigan Medical School Ann Arbor GRECC Ann Arbor, Michigan Lester B. Salans, MD Adjunct Professor The Rockefeller University Clinical Professor of Medicine Mt. Sinai School of Medicine New York, New York Salil D. Sarkar Department of Radiology SUNY Health Sciences Center at Brooklyn College of Medicine Brooklyn, New York David H. Sarne, MD Associate Professor of Medicine Department of Internal Medicine University of Illinois at Chicago College of Medicine Chicago, Illinois Ernst J. Schaefer, MD Professor of Medicine Lipid Division Tufts University School of Medicine New England Medical Center Boston, Massachusetts Isaac Schiff, MD

Joe Vincent Meigs Professor of Gynecology Department of Obstetrics and Gynecology Harvard Medical School Massachusetts General Hospital Boston, Massachusetts R. Neil Schimke, MD Professor of Medicine and Pediatrics Chief, Division of Endocrinology and Genetics University of Kansas School of Medicine Kansas City, Kansas James R. Schreiber, MD Elaine and Mitchell Yanow Professor and Head Department of Obstetrics and Gynecology Washington University School of Medicine St. Louis, Missouri David E. Schteingart, MD Professor of Internal Medicine Division of Endocrinology and Metabolism University of Michigan Medical School Ann Arbor, Michigan Ellen W. Seely, MD Assistant Professor of Medicine Director of Clinical Research Endocrine-Hypertension Division Harvard Medical School Brigham and Women’s Hospital Boston, Massachusetts Markus J. Seibel, MD , PD Associate Professor of Medicine Division of Endocrinology and Metabolism University of Heidelberg Medical School Heidelberg, Germany Elizabeth Shane, MD Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Lawrence E. Shapiro, MD Clinical Professor of Medicine SUNY at Stony Brook School of Medicine Health Sciences Center Stony Brook, New York Director, Division of Endocrinology Winthrop University Hospital Mineola, New York Meeta Sharma, MBBS, MD Assistant Director, Diabetes Team Division of Endocrinology Georgetown University School of Medicine MedStar Diabetes Institute Washington Hospital Center Washington, DC R. Michael Siatkowski, MD Associate Professor of Ophthalmology Dean A. McGee Eye Institute Oklahoma City, Oklahoma Omega L. Silva, MD Professor Emeritus of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Shonni J. Silverberg, MD Associate Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Joe Leigh Simpson, MD Ernst W. Bertner Chairman and Professor Department of Obstetrics and Gynecology Baylor College of Medicine Houston, Texas Ethel S. Siris, MD Madeline C. Stabile Professor of Clinical Medicine Department of Medicine Columbia University College of Physicians and Surgeons New York, New York Glen W. Sizemore, MD Professor of Medicine Division of Endocrinology and Metabolism Loyola University of Chicago Stritch School of Medicine Maywood, Illinois Niels E. Skakkebaek, MD Professor of Growth and Reproduction University of Copenhagen Faculty of Health Sciences Rigshospitalet

Copenhagen, Denmark Celia D. Sladek, PhD Professor and Acting Chair Department of Physiology and Biophysics Finch University of Health Sciences Chicago Medical School North Chicago, Illinois John R. Sladek, Jr., PhD Professor and Chairman Department of Neuroscience Finch University of Health Sciences Chicago Medical School North Chicago, Illinois Eduardo Slatopolsky, MD Renal Division Washington University School of Medicine St. Louis, Missouri Robert C. Smallridge, MD Professor of Medicine Mayo Medical School Chair, Endocrine Division Mayo Clinic Jacksonville, Florida Robert J. Smith, MD Professor of Medicine Chief of Endocrinology Brown University School of Medicine Director, Hallett Diabetes Center Rhode Island Hospital Providence, Rhode Island Richard H. Snider, Jr., PhD Chief Chemist Endocrinology Research Laboratory Veterans Affairs Medical Center Washington, DC Phyllis W. Speiser, MD Professor of Clinical Pediatrics Department of Pediatrics New York University School of Medicine New York, New York North Shore University Hospital Manhasset, New York Harvey J. Stern, MD , PhD Genetics and IVF Institute Fairfax, Virginia Martin J. Stevens, MD Associate Professor of Internal Medicine University of Michigan Medical School Ann Arbor, Michigan Andrew F. Stewart, MD Professor of Medicine Chief, Division of Endocrinology University of Pittsburgh School of Medicine Pittsburgh, Pennsylvania Elizabeth A. Streeten, MD Clinical Assistant Professor of Medicine Department of Endocrinology, Diabetes, and Obesity University of Maryland School of Medicine Baltimore, Maryland Gordon J. Strewler, MD Professor of Medicine Department of Medical Service Harvard Medical School Boston, Massachusetts Veterans Affairs Boston Healthcare System West Roxbury, Massachusetts Martin I. Surks, MD Professor of Medicine and Pathology Department of Medicine Albert Einstein College of Medicine of Yeshiva University Montefiore Medical Center Bronx, New York Arthur L. M. Swislocki, MD Associate Professor of Medicine Department of Internal Medicine University of California, Davis, School of Medicine Davis, California Veterans Affairs Outpatient Clinic Martinez, California Shahrad Taheri, MSc, MB, MRCP Wellcome Trust Research Fellow Division of Endocrinology and Metabolism University of London Imperial College School of Medicine

Hammersmith Hospital London, England Robert J. Tanenberg, MD , FACP Professor of Medicine Section of Endocrinology and Metabolism Brody School of Medicine East Carolina University School of Medicine Greenville, North Carolina Kamal Thapar, MD Assistant Professor of Neurosurgery University of Toronto Faculty of Medicine Toronto Western Hospital, University Health Toronto, Ontario Canada Ramesh K. Thapar, MD Senior Resident Department of Psychiatry University of Maryland School of Medicine Baltimore, Maryland Michael A. Thomas, MD Associate Professor Department of Clinical Obstetrics and Gynecology University of Cincinnati College of Medicine Cincinnati, Ohio Christopher J. Thompson, MB, ChB, MD , FRCPI Consultant Physician and Endocrinologist Department of Endocrinology Royal College of Surgeons in Ireland Beaumont Hospital Dublin, Ireland Keith Tornheim, PhD Associate Professor of Biochemistry Boston University School of Medicine Boston, Massachusetts David J. Torpy, MBBS, PhD, FRACP Senior Lecturer Department of Medicine University of Queensland Faculty of Health Sciences Brisbane, Australia Carmelita U. Tuazon, MD , MPH Professor of Medicine George Washington University School of Medicine and Health Sciences Washington, DC Catherine Tuck, MD Assistant Professor of Medicine Columbia University College of Physicians and Surgeons New York, New York Michael L. Tuck, MD Professor of Medicine University of California, Los Angeles, UCLA School of Medicine Los Angeles, California Veterans Affairs Medical Center, Sepulveda Sepulveda, California Stephen Jon Usala, MD , PhD Clinical Associate Professor Department of Medicine Texas Tech University Health Sciences Center School of Medicine Amarillo, Texas Eve Van Cauter, PhD Professor of Medicine University of Chicago Pritzker School of Medicine Chicago, Illinois Greet H. Van Den Berghe, MD , PhD Associate Professor of Intensive Care Medicine Catholic University of Leuven Leuven, Belgium David H. Van Thiel, MD Director of Transplantation Loyola University of Chicago Stritch School of Medicine Loyola University Medical Center Liver Transplant Office Maywood, Illinois Joseph G. Verbalis, MD Professor of Medicine and Physiology Georgetown University School of Medicine Washington, DC Robert Volpé, MD , FRCP(C), MACP, FRCP (Edin & Lord) Professor Emeritus Department of Medicine

University of Toronto Faculty of Medicine Toronto, Ontario Canada Steven G. Waguespack, MD Fellow, Adult and Pediatric Endocrinology Departments of Medicine and Pediatrics Division of Endocrinology Indiana University School of Medicine Riley Children’s Hospital Indianapolis, Indiana Brian Walsh, MD Director, Menopause Center Department of Obstetrics and Gynecology Harvard Medical School Brigham and Women’s Hospital Boston, Massachusetts David O. Walterhouse, MD Assistant Professor of Pediatrics Northwestern University Medical School Children’s Memorial Hospital Chicago, Illinois Emily C. Walvoord, MD Senior Fellow Department of Pediatric Endocrinology and Diabetology Indiana University School of Medicine Riley Hospital for Children Indianapolis, Indiana James H. Warram, MD , ScD Senior Investigator Section on Genetics and Epidemiology Research Division Harvard Medical School Joslin Diabetes Center Boston, Massachusetts Michelle P. Warren, MD Professor of Obstetrics and Gynecology and Medicine Wyeth Ayerst Professor of Women’s Health Columbia University College of Physicians and Surgeons New York, New York Leonard Wartofsky, MD , MPH, MACP Clinical Professor of Medicine Georgetown University School of Medicine Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Chair, Department of Medicine Washington Hospital Center Clinical Professor of Medicine Howard University College of Medicine Washington, DC Professor of Medicine and Physiology Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland Stephen I. Wasserman, MD Helen M. Ranney Professor of Medicine Chair, Department of Medicine University of California, San Diego, School of Medicine La Jolla, California Colleen Weber, RN Pediatric Endocrine Nurse Cardinal Glennon Children’s Hospital St. Louis, Missouri Anthony Peter Weetman, MD , FRCP, DSc Professor of Medicine University Department of Clinical Sciences University of Sheffield School of Medicine Northern General Hospital Sheffield, England Gordon C. Weir, MD Professor of Medicine Research Division Harvard Medical School Joslin Diabetes Center Boston, Massachusetts Laura S. Welch, MD Director, Occupational and Environmental Medicine Georgetown University School of Medicine Washington Hospital Center Washington, DC Samuel A. Wells, Jr., MD Professor of Surgery Washington University School of Medicine St. Louis, Missouri

Jon C. White, MD , FACS Director of Surgical Intensive Care Department of Surgery Veterans Affairs Medical Center Associate Professor of Surgery George Washington University School of Medicine and Health Sciences Washington, DC Perrin C. White, MD Professor of Pediatrics University of Texas Southwestern Medical Center at Dallas Southwestern Medical School Dallas, Texas Michael P. Whyte, MD Medical-Scientific Director Department of Metabolic and Molecular Research Professor of Medicine, Pediatrics, and Genetics Division of Bone and Mineral Diseases Washington University School of Medicine Barnes–Jewish Hospital St. Louis, Missouri Gordon H. Williams, MD Professor of Medicine Harvard Medical School Chair, Endocrine-Hypertension Division Brigham and Women’s Hospital Boston, Massachusetts Stephen J. Winters, MD Professor of Medicine Chief, Division of Endocrinology and Metabolism University of Louisville School of Medicine Louisville, Kentucky Joseph I. Wolfsdorf, MB, BCh Associate Professor of Pediatrics Department of Medicine Division of Endocrinology Harvard Medical School Children’s Hospital National Medical Center Boston, Massachusetts I-Tien Yeh, MD Associate Professor Department of Pathology University of Texas Medical School at San Antonio University Health Center San Antonio, Texas Paul M. Yen, MD Chief, Molecular Regulation and Neuroendocrinology Clinical Endocrinology Branch National Institute of Diabetes and Digestive and Kidney Disease National Institutes of Health Bethesda, Maryland James E. Zadina, PhD Professor of Medicine Tulane University School of Medicine Director, Neuroscience Laboratory Department of Research Veterans Affairs Medical Center New Orleans, Louisiana Gary P. Zaloga, MA, MD Director of Critical Care Medicine Department of Medicine Georgetown University School of Medicine Washington Hospital Center Washington, DC Charles Zaloudek, MD Professor Department of Pathology University of California, San Francisco, School of Medicine San Francisco, California Carol Zapalowski, MD , PhD Colorado Center for Bone Research Lakewood, Colorado Thomas R. Ziegler, MD Associate Professor of Medicine Division of Endocrinology/Metabolism Emory University School of Medicine Atlanta, Georgia Michael Zinger, MD Clinical Instructor Department of Obstetrics and Gynecology Division of Reproductive Endocrinology University of Cincinnati College of Medicine Cincinnati, Ohio

EDITORS EDITOR Kenneth L. Becker, MD , PhD Professor of Medicine Professor of Physiology and Experimental Medicine Director of Endocrinology and Metabolism George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC ASSOCIATE EDITORS John P. Bilezikian, MD Professor of Medicine and Pharmacology Department of Medicine Columbia University College of Physicians and Surgeons New York, New York William J. Bremner, MD , PhD Robert G. Petersdorf Professor and Chairman Department of Medicine University of Washington School of Medicine Seattle, Washington Wellington Hung, MD , PhD Professor Emeritus of Pediatrics Georgetown University School of Medicine Professorial Lecturer in Pediatrics George Washington University School of Medicine and Health Sciences Washington, DC C. Ronald Kahn, MD Mary K. Iacocca Professor of Medicine Harvard Medical School President and Director, Research Division Joslin Diabetes Center Boston, Massachusetts D. Lynn Loriaux, MD , PhD Professor and Chair Department of Medicine Oregon Health Sciences University School of Medicine Portland, Oregon Eric S. Nylén, MD Associate Professor of Medicine Department of Endocrinology George Washington University School of Medicine and Health Sciences Veterans Affairs Medical Center Washington, DC Robert W. Rebar, MD Professor Department of Obstetrics and Gynecology University of Cincinnati College of Medicine Chief, Obstetrics and Gynecology University Hospital Cincinnati, Ohio Associate Executive Director American Society for Reproductive Medicine Birmingham, Alabama Gary L. Robertson, MD Professor of Medicine and Neurology Department of Endocrinology Northwestern University Medical School Chicago, Illinois Richard H. Snider, Jr., PhD Chief Chemist Endocrinology Research Laboratory Veterans Affairs Medical Center Washington, DC Leonard Wartofsky, MD , MPH, MACP Clinical Professor of Medicine Georgetown University School of Medicine Clinical Professor of Medicine George Washington University School of Medicine and Health Sciences Chair, Department of Medicine Washington Hospital Center Clinical Professor of Medicine Howard University College of Medicine Washington, DC Professor of Medicine and Physiology Uniformed Services University of the Health Sciences F. Edward Hébert School of Medicine Bethesda, Maryland

CHAPTER 1 ENDOCRINOLOGY AND THE ENDOCRINE PATIENT Principles and Practice of Endocrinology and Metabolism

CHAPTER 1 ENDOCRINOLOGY AND THE ENDOCRINE PATIENT KENNETH L. BECKER, ERIC S. NYLÉN, AND RICHARD H. SNIDER, J R. Definitions Role of the Endocrine System Hormones Chemical Classification Sources, Controls, and Functions Transport Types of Secretory Transport Overlap of Exocrine and Endocrine Types of Secretion Tyranny of Hormone Terminology Endocrine System Interaction with all Body Systems Genetics and Endocrinology Normal and Abnormal Expression or Modulation of the Hormonal Message and its Metabolic Effect The Endocrine Patient Frequency of Endocrine Disorders Cost of Endocrine Disorders Factors that Influence Test Results Reliability of the Laboratory Determination Determination of Abnormal Test Results Risks of Endocrine Testing Cost and Practicability of Endocrine Testing Conclusion Chapter References

DEFINITIONS Endocrinology is the study of communication and control within a living organism by means of chemical messengers that are synthesized in whole or in part by that organism. Metabolism, which is an integral part of the science of endocrinology, is the study of the biochemical control mechanisms that occur within living organisms. The term includes such diverse activities as gene expression; biosynthetic pathways and their enzymatic catalysis; the modification, transformation, and degradation of biologic substances; the biochemical mediation of the actions and interactions of such substances; and the means for obtaining, storing, and mobilizing energy. The chemical messengers of endocrinology are the hormones, endogenous informational molecules that are involved in both intracellular and extracellular communication.

ROLE OF THE ENDOCRINE SYSTEM The mammalian organism, including the human, is multicellular and highly specialized with regard to sustaining life and reproductive processes. Reproduction requires gametogenesis, fertilization, and implantation. Subsequently, the new intrauterine conception must undergo cell proliferation, organogenesis, and differentiation into a male or female. After parturition, the newborn must grow and mature sexually, so that the cycle may be repeated. To a considerable extent, the endocrine system influences or controls all of these processes. Hormones participate in all physiologic functions, such as muscular activity, respiration, digestion, hematopoiesis, sense organ function, thought, mood, and behavior. The overall purpose of the coordinating, regulating, integrating, stimulating, suppressing, and modulating effects of the many components of the endocrine system is homeostasis. The maintenance of a healthy optimal internal milieu in the presence of a continuously changing and sometimes threatening external environment is termed allostasis.

HORMONES CHEMICAL CLASSIFICATION Most hormones can be classified into one of several chemical categories: amino-acid derivatives (e.g., tryptophan ® serotonin and melatonin; tyrosine ® dopamine, norepinephrine, epinephrine, triiodothyronine, and thyroxine; L -glutamic acid ® g-aminobutyric acid; histidine ® histamine), peptides or polypeptides (e.g., thyrotropin-releasing hormone, insulin, growth hormone, nerve growth factor), steroids (e.g., progesterone, androgens, estrogens, corticosteroids, vitamin D and its metabolites), and fatty acid derivatives (e.g., prostaglandins, leukotrienes, thromboxanes). SOURCES, CONTROLS, AND FUNCTIONS Previously hormones were thought to be synthesized and secreted predominantly by anatomically discrete and circumscribed glandular structures, called ductless glands (e.g., pituitary, thyroid, adrenals, gonads). However, many microscopic organoid-like groups of cells and innumerable other cells of the body contain and secrete hormones (see Chap. 175). The classic “glands” of endocrinology have lost their exclusivity, and although they are important on physiologic and pathologic levels, the widespread secretion of hormones throughout the body by “nonglandular” tissues is of equal importance. Most hormones are known to have multiple sources. Moreover, the physiologic stimuli that release these hormones are often found to differ according to their locale. The response to a secreted hormone is not stereotyped but varies according to the nature and location of the target cells or tissues. TRANSPORT TYPES OF SECRETORY TRANSPORT Hormones have various means of reaching target cells. In the early decades of the development of the field of endocrinology, hormones were conceived to be substances that traveled to distal sites through the blood. This is accomplished by release into the extracellular spaces and subsequent entrance into blood vessels by way of capillary fenestrations. The most appropriate term for such blood-bone communication is hemocrine (Fig. 1-1).

FIGURE 1-1. Different types of hormonal communication detailed in the chapter. The darkened areas on the cell membrane represent receptors. (H, hormone.) See text for explanations.

Several alternative means of hormonal communication exist, however. Paracrine communication involves the extrusion of hormonal contents into the surrounding interstitial spaces; the hormone then interacts with receptors on nearby cells (see Fig. 1-1 and Chap. 4 and Chap. 175).1 Direct paracrine transfer of cytoplasmic messenger molecules into adjacent cells may occur through specialized gap junctions (i.e., intercrine secretion).2 Unlike hemocrine secretion, in which the hormonal secretion is diluted within the circulatory system, paracrine secretion delivers a very high concentration of hormone to its target site. Juxtacrine communication occurs when the messenger molecule does not traverse a fluid phase to reach another cell, but, instead, remains associated with the plasma membrane of the signaling cell while acting directly on an immediately adjacent receptor cell (e.g., intercellular signaling that is adhesion dependent and occurs between endothelial cells and leukocytes and transforming growth factor-a in human endometrium).3,4 Hormones may be secreted and subsequently interact with the same cell that released the substance; this process is autocrine secretion (see Fig. 1-1).5 The secreted hormone stimulates, suppresses, or otherwise modulates the activity of the secreting cell. Autocrine secretion is a form of self-regulation of a cell by its own product. When peptide hormones or other neurotransmitters or neuromodulators are produced by neurons, the term neurocrine secretion is used (see Fig. 1-1).6 This specialized form of paracrine release may be synaptic (i.e., the messenger traverses a structured synaptic space) or nonsynaptic (i.e., the messenger is carried to its local or distal site of action by way of the extracellular fluid or the blood). Nonsynaptic neurocrine secretion has also been called neurosecretion. An example of neurosecretion is the release of vasopressin and oxytocin into the circulatory system by nervous tissue of the pituitary (see Chap. 25). Several peptides and amines are secreted into the luminal aspect of the gut (e.g., gastrin, somatostatin, luteinizing hormone– releasing hormone, calcitonin, secretin, vasoactive intestinal peptide, serotonin, substance P).7 This process may be called solinocrine secretion (see Fig. 1-1), from the Greek word for a hollow tube. Solinocrine secretion also occurs into the bronchi, the urogenital tract, and other ductal structures.8 Commonly, the same hormone can be transported by more than one of these means.9 Extracellular transportation may not always be necessary for hormones to exert their effects. For example, some known hormonal secretions that are transported by one or more of these mechanisms are also found in extremely low concentrations within the cytoplasm of many cells. In such circumstances, these hormones do not appear to be localized to identifiable secretion granules and probably act primarily within the cell. This phenomenon may be called intracrine secretion. As shown in Figure 1-1, the process comprising uptake of a hormone precursor H1 and intracellular conversion into H2 (e.g., estrogens) or H3 (e.g., androgens) and subsequent binding and nuclear action is also a form of intracrine communication. OVERLAP OF EXOCRINE AND ENDOCRINE TYPES OF SECRETION Classically, an exocrine gland is a specialized structure that secretes its products at an external or internal surface (e.g., sweat glands, sebaceous glands, salivary glands, oxyntic or gastric glands, pancreatic exocrine glandular system, prostate gland). An exocrine gland may be unicellular (e.g., mucous or goblet cells of the epithelium of mucous membranes) or multicellular (e.g., salivary glands). Many multicellular exocrine glands possess a structured histologic organization that is suited to the production and delivery of secretions that are produced in relatively large quantities. A specialized excretory duct or system of ducts usually constitutes an intrinsic part of the gland. Some exocrine glandular cells secrete their substances by means of destruction of the cells themselves (i.e., holocrine secretion); an example is the sebaceous glands. Other exocrine glandular cells secrete their substances by way of the loss of a portion of the apical cytoplasm along with the material being secreted (i.e., apocrine secretion); an example is the apocrine sweat glands. Alternatively, in many forms of exocrine secretion, the secretory cells release their products through the cell membrane, and the cell remains intact (i.e., merocrine secretion); an example is the salivary glands. The constituents of some exocrine glands, particularly those opening on the external surface of the body, sometimes function as pheromones, which are chemical substances that act on other members of the species.10 Many exocrine glands contain cells of the diffuse neuroendocrine system (see Chap. 175) and neurons; both cell types secrete peptide hormones. Peptide hormones, steroids, and prostaglandins are found in all exocrine secretions (e.g., sweat, saliva, milk, bile, seminal fluid; see Chap. 106).11,12,13 and 14 Although they usually are not directly produced in such glands, thyroid and steroid hormones are found in exocrine secretions as well.15,16,17 and 18 The preferred approach is to view the term “exocrine” as a histologic-anatomic entity and not as a term that is meant to be antithetical to or to contrast with the term “endocrine.” Endocrinologists are concerned clinically and experimentally with all means of hormonal communication. The word “endocrine” is best used in a global sense, indicating any and all means of communication by messenger molecules.

TYRANNY OF HORMONE TERMINOLOGY Hormones usually are named at the time of their discovery. Sometimes, the names are based on the locations where they were first found or on their presumed effects. However, with time, other locations and other effects are discovered, and these new locations or effects often are more physiologically relevant than the initial findings. Hormonal names are often overly restrictive, confusing, or misleading. In many instances, such hormonal names have become inappropriate. For example, atrial natriuretic hormone is present in the brain, hypothalamus, pituitary, autonomic ganglia, and lungs as well as atrium, and it has effects other than natriuresis (see Chap. 178). Gastrin-releasing peptide is found in semen, far from the site of gastrin release. Somatostatin, which was found in the hypothalamus and named for its inhibitory effect on growth hormone, occurs in many other locations and has multiple other functions (see Chap. 169). Calcitonin, which initially was thought to play an important role in regulating serum calcium and was named accordingly, appears to exert many other effects, and its influence on serum calcium may be quite minor (see Chap. 53). Growth hormone–releasing hormone and arginine vasopressin are found in the testis, where effects on growth hormone release or on the renal tubular reabsorption of water are most unlikely. Vasoactive intestinal peptide is found in multiple tissues other than the intestines (see Chap. 182). Insulin, named for the pancreatic islets, is found in the brain and elsewhere.19 Prostaglandins have effects that are far more widespread than those exerted in the secretions of the prostate, from which their name derives (see Chap. 172). The endocrine lexicon also contains substances called hormones that are not hormones. In the human, melanocyte-stimulating hormone (MSH) is not a functional hormone, but it comprises amino-acid sequences within the proopiomelanocortin (POMC) molecule: a-MSH within the adrenocorticotropic hormone (ACTH) moiety, b-MSH within d-lipotropin, and d-MSH within the N-terminal fragment of POMC (see Chap. 14). Numerous peptide hormones exist that, because of their effects on DNA synthesis, cell growth, and cell proliferation, have been called growth factors and cytokines (see Chap. 173 and Chap. 174). These substances, which act locally and at a distance, often do not have the sharply delineated target cell selectivity that was attributed to them when they first were discovered. Their terminology also is confusing and often misleading. Aside from occasional readjustments of hormonal nomenclature, no facile solution appears to exist to the quandary of terminology, other than an awareness of the pitfalls into which the terms may lead us.

ENDOCRINE SYSTEM INTERACTION WITH ALL BODY SYSTEMS Although speaking in terms of the cardiovascular, respiratory, gastrointestinal, and nervous systems is convenient, the endocrine system anatomically and functionally overlaps with all body systems (see Part X). Extensive overlap is found between the endocrine system and the nervous system (see Chap. 175 and Chap. 176). Hormonal peptides are synthesized in the cell bodies of neurons, are transported along axons to nerve terminals, and are released at the nerve endings. Within these neurons, they coexist with classic neurotransmitters and often are coreleased with them. These substances play a role in neuromodulation or neurosecretion by means of the extracellular fluid. The nerves in which peptide hormones appear to play a role in the transmission of information are called peptidergic nerves.20 It is the ample similarity of ultrastructure, histochemistry, and hormonal contents of nerve cells and of many peptide-secreting endocrine cells that has led to the concept of the diffuse neuroendocrine system.

GENETICS AND ENDOCRINOLOGY The rapid application of new discoveries and new techniques in genetics has revolutionized medicine, including the field of endocrinology. DNA probes have been targeted to selected genes, and the chromosomal locations of genes related to many hormones and their receptors have been determined. A complete map of the human genome is gradually emerging.21 This approach has led to new knowledge about hormone biosynthesis and has provided important information concerning species differences and evolution. The elucidation of the chromosomal loci for genes controlling the biosynthesis of hormone receptors should provide insights into the

physiologic effects of hormones. Clinically, these techniques have potential significance as a diagnostic aid in evaluating afflicted patients, a means of identifying asymptomatic heterozygotes, and a method for identification of unborn individuals at risk (i.e., prenatal diagnosis; see Chap. 240). Delineation of processes of genetic expression is revealing the mechanisms of hormonal disease (e.g., obesity22) and also may lead to gene therapy for some forms of endocrine illness or humoral-mediated conditions.23

NORMAL AND ABNORMAL EXPRESSION OR MODULATION OF THE HORMONAL MESSAGE AND ITS METABOLIC EFFECT A sophisticated and faultless machinery is required for appropriate hormonal expression. The hormonal messenger is subject to modifications that may occur anywhere from its initial synthesis to its final arrival at its target site. Subsequently, the expression of the message at this site (i.e., its action) may also be modified (see Chap. 4). The modulations or alterations of the hormonal message or its final action may be physiologic or pathologic. Table 1-1 summarizes some of the normal and abnormal modulations of a hormonal message and its subsequent metabolic effects.

TABLE 1-1. Modulation of the Hormone Message and Its Subsequent Physiologic or Pathologic Metabolic Effects

On a physiologic level, the first steps in the genetic ordering of hormonal synthesis, the subsequent posttranslational processing of the hormone, the postsecretory extracellular transport, the receptor mediation of the hormone and subsequent transduction, and the inactivation and clearance of the hormone all contribute to expressing, diversifying, focalizing, and specifying the hormonal message and its ultimate action. On a pathologic level, all of these steps are subject to malfunction, causing disease. Our increased knowledge of endocrine systems has forced us to rethink many traditional concepts. To dispel some common misconceptions, listing several “nots” of endocrinology may be worthwhile (Table 1-2).

TABLE 1-2. Several “Nots” of Modern Endocrinology

THE ENDOCRINE PATIENT FREQUENCY OF ENDOCRINE DISORDERS In a survey of the subspecialty problems seen by endocrinologists, the six most common, in order of frequency, were found to be diabetes mellitus, thyrotoxicosis, hypothyroidism, nontoxic nodular goiter, diseases of the pituitary gland, and diseases of the adrenal gland. Some conditions seen by endocrinologists are infrequent or rare (e.g., congenital adrenal hyperplasia, pseudohypoparathyroidism), whereas others are relatively common (e.g., Graves disease, Hashimoto thyroiditis), and some are among the most prevalent diseases in general practice (e.g., diabetes mellitus, obesity, hyperlipoproteinemia, osteoporosis, Paget disease). The third most common medical problem encountered by general practitioners is diabetes mellitus, and the tenth most frequent problem is obesity.66 Of the total deaths in the United States (i.e., both sexes, all races, and all ages combined), diabetes mellitus is the seventh most common cause. The most common cause of death (heart disease) and the third most common (cerebrovascular accidents) are greatly influenced by metabolic conditions such as diabetes mellitus and hyperlipemia.67 COST OF ENDOCRINE DISORDERS The frequency and morbidity of endocrine diseases such as osteoporosis, obesity, hypothyroidism, and hyperthyroidism, and the grave consequences of other endocrine disorders such as Cushing syndrome and Addison disease demonstrate that the expense to society is considerable. In the case of diabetes, the health care expenditure is staggering. Approximately 10.3 million people have diabetes in the U.S., and an estimated 5.4 million have undiagnosed diabetes. Direct medical expenses attributed to diabetes total $44.1 billion. The total annual medical expenses of people with diabetes average $10,071 per capita, as compared to $2,669 for persons without diabetes.68 Interestingly, these expenses may be less if the appropriate specialties are involved in the care.69 FACTORS THAT INFLUENCE TEST RESULTS In clinical medicine, hormonal concentrations usually are ascertained from two of the most easily obtained sources: blood and urine. The diagnosis of an endocrinopathy often depends on the demonstration of increased or decreased levels of these blood or urine constituents. However, several factors must be kept in mind when interpreting a result that appears to be abnormal. These may include age, gender, time of day, exercise, posture, emotional state, hepatic and renal status, presence of other illness, and concomitant drug therapy (see Chap. 237 and Chap. 239). RELIABILITY OF THE LABORATORY DETERMINATION The practice of clinical endocrinology far from a large medical center was previously hindered by the difficulty in obtaining blood and urine tests essential for appropriate diagnosis and follow-up care. However, accurate and rapid analyses now are provided by commercial laboratories. Nevertheless, wherever performed, some tests are unreliable because of methodologic difficulties. Other tests may be difficult to interpret because of a particular susceptibility to alteration by physiologic or pharmacologic factors (e.g., plasma catecholamines; see Chap. 86). Although many tests are sensitive and specific, they all have innate interassay and intraassay variations that may be particularly misleading when a given value is close to the clinical “medical decision point” (see Chap. 237). Some laboratory differences are due to hormone heterogeneity (e.g., growth hormone has several isoforms, which bind differently to growth hormone–binding proteins).70 DETERMINATION OF ABNORMAL TEST RESULTS Not uncommonly, the intellectual or commercial enthusiasm engendered by a new diagnostic procedure of presumed importance is found to be unjustified, because the

“test” was based on an invalid premise, because too few of ill patients were studied, because normative data to establish reference values were insufficient, or because subsequent studies were not confirmatory (see Chap. 237 and Chap. 241). The increased sophistication of medical testing has made the physician and the patient aware of the presence of “abnormalities” that may be harmless: physiologic deviations from that which is most common, or pathologic entities that commonly remain asymptomatic. Such findings may cause considerable worry, lead to the expense and risk of further diagnostic procedures, and even cause needless therapeutic intervention. Some “abnormalities” are the result of methods of imaging. For example, sonography of the thyroid may demonstrate the presence of small nodules within the thyroid gland of a person without any palpable abnormality of that region of the gland; most such microlesions are benign or behave as if they were. Another “abnormality” revealed by imaging is the occasional heterogeneous appearance of a normal pituitary gland on a computed tomography (CT) examination. Intermingled CT-lucent and CT-dense areas are seen on the scan, and such nonhomogeneous areas may be confused with a microadenoma.71,72 The increasing use of magnetic resonance imaging (MRI) of the brain may reveal a bona fide asymptomatic microadenoma of the pituitary gland, but extensive endocrine workup often reveals many such lesions to be nonfunctional. They occur in as much as 10% of the normal population.73 Rathke cleft cysts of the anterior sella turcica or the anterior suprasellar cistern often are seen by MRI.74 Although an occasional patient may have a large and symptomatic lesion,75 most of these lesions are small and asymptomatic. During MRI or CT examination of the brain, the examiner often incidentally encounters a “primary empty sella,” an extension of the subarachnoid space into the sella turcica with a resultant flattening of the pituitary gland in a patient without any pituitary lesion or any prior surgery of that region (see Chap. 11). Although some of these patients may be symptomatic, most have no associated symptoms or hormonal deficit. Another, albeit rare, lesion of the pituitary region seen on MRI is a sellar spine. This asymptomatic anatomic variant is an osseous spine arising in the midline from the dorsum sella that protrudes into the pituitary fossa; it may be an ossified remnant of the cephalic tip of the notochord.76 MRI or CT scanning of the abdomen may reveal the presence of harmless morphologic variations of the adrenal gland (i.e., incidentalomas) that sometimes leads to unnecessary surgery.77 (See Chap. 84.) RISKS OF ENDOCRINE TESTING Endocrine testing is not always benign. Many procedures can cause mild to marked side effects.78,79,80 and 81 Other diagnostic maneuvers, particularly angiography, may result in severe illness.82 The expected benefit of any procedure that is contemplated for a patient clearly should be greater than the risk. COST AND PRACTICABILITY OF ENDOCRINE TESTING In addition to being aware of the many factors that influence hormonal values, the limitations of laboratory determinations, and the potential risks of some of these procedures, the endocrinologist must be aware of their expense, particularly because medical costs have increased at an annual rate that is almost twice the rate of overall inflation during the last several years. A hypertensive patient with hypokalemia who is taking neither diuretics nor laxatives should undergo studies of the renin-angiotensin-aldosterone system, and appropriate pharmacologic or dietary manipulations of sodium balance should be instituted (see Chap. 90). But what should be done with the hypertensive patient who is normokalemic? Occasionally, such a person may have an aldosteronoma.83 Should such normokalemic patients be studied? Similarly, should the approximately 25 million hypertensive patients in the United States undergo urinary collections for determinations of catecholamine metabolites to find the rare patient with pheochromocytoma? In the context of the individual physician-patient relationship, the answers to such questions may not be difficult, but they become more controversial when placed within the framework of fiscal guidelines.

CONCLUSION The complexity of the endocrine system presents a profound intellectual challenge. The macrosystem of endocrine glands secretes its hormones under the influence of other gland-based releasing factors or neural influences or both. The very act of secretion alters subsequent secretion by means of feedback controls (see Chap. 5). Superimposed on this already complex arrangement, the microsystem of dispersed, somewhat independent, but overlapping units throughout the body, as well as the continuous modulation of the receptors for the secreted hormones, allow general or focal actions that are coordinated with other body functions, tempered to the occasion, and appropriate to the needs of the individual. That such a complex system may go awry and that a dysfunction may have a considerable impact on the patient is not surprising. Because endocrinology and metabolism are broad subjects that incorporate much, if not all, of normal body functions and disease states, they defy easy categorization. However, these enormous complexities, rather than deterring the clinician, researcher, or student, should provide a stimulus to probe deeper into areas difficult to understand and should hasten the eventual application of new developments to patient care. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35.

Hofbauer LC, Khosla S, Dunstan CR, et al. The roles of osteoprotegerin and osteoprotegerin ligand in the paracrine regulation of bone resorption. J Bone Mineral Res 2000; 15:2. Usadel H, Bornstein SR, Ehrhart-Bornstein M, et al. Gap junctions in the adrenal cortex. Horm Metab Res 1993; 25:653. Zimmerman GA, Lorant DE, McIntyre TM, Prescott SM. Juxtacrine intercellular signaling: another way to do it. Am J Respir Cell Mol Biol 1993; 9:573. Bush MR, Mele JM, Couchman GM, Walmer DK. Evidence of juxtacrine signaling for transforming growth factor alpha in human endometrium. Biol Reprod 1998; 59:1522. Hashimoto K, Higashiyama S, Asada H, et al. Heparin-binding epidermal growth factor-like growth factor is an autocrine growth factor for human keratinocytes. J Biol Chem 1994; 269:20060. Gouin FJ. Morphology, histology and evolution of Myriapoda and insects. 3. The nervous system and the neurocrine systems. Fortschr Zool 1965; 17:189. Uvnas-Wallensten K. Luminal secretion of gut peptides. Clin Gastroenterol 1980; 9:545. Van Minnen J. Production and exocrine secretion of LHRH-like material by the male rat reproductive tract. Peptides 1988; 9:515. Becker KL. The coming of age of a bronchial epithelial cell. Am J Respir Crit Care Med 1994; 149:183. Cohn BA. In search of human skin pheromones. Arch Dermatol 1994; 130:1048. Corps AN, Brown KD, Rees LH, et al. The insulin-like growth factor I content in human milk increases between early and full lactation. J Clin Endocrinol Metab 1988; 67:25. Hammami MM, Haq A, Al-Sedairy S. The level of endothelin-like immunoreactivity in seminal fluid correlates positively with semen volume and negatively with plasma gonadotropin levels. Clin Endocrinol 1994; 40:361. Schmidt NA. Salivary cortisol testing in children. Issues Compr Pediatr Nurs 1998; 20:183. Voss HF. Saliva as a fluid for measurement of estriol levels. Am J Obstet Gynecol 1999; 180:S226. Nizankowska B, Abramowicz T, Korezowski R, Rusin J. Triiodothyronine and thyroxine in human, cow's and formula milk. Exp Clin Endocrinol 1988; 91:116. Langer P, Moravec R, Ohradka B, Foldes O. Iodothyronines in human bile. Endocrinol Exp 1988; 22:35. Shieh CC, Chang SC, Tzeng CR, et al. Measurement of testosterone in seminal plasma, saliva and serum by solid-phase enzyme immunoassay. Andrologia 1987; 19:614. O'Rorke A, Kane MM, Gosling JP, et al. Development and validation of a monoclonal antibody enzyme assay for measuring progesterone in saliva. Clin Chem 1994; 40:454. Hendricke SA, Roth J, Rishi S, Becker KL. Insulin in the nervous system. In: Krieger DT, Bronnstein MJ, Martin J, eds. Brain peptides. New York: John Wiley & Sons, 1983:403. Bean AJ, Zhang X, Hokfelt T. Peptide secretion: what do we know? FASEB J 1994; 8:630. Zabarovsky ER, Allikmets R, Kholodnyuk I, et al. Construction of representative NOTI linking libraries specific for the total human genome and chromosome 3. Genomics 1994; 20:312. Perusse L, Chagnon YC, Weisnagel J, Bouchard C. The human obesity gene map: the 1998 update. Obes Res 1999; 7:111. Moldawer LL, Edwards PD, Minter RM, et al. Application of gene therapy to acute inflammatory disease. Shock 1999; 12:83. Atkinson J, Martin R. Mutations to nonsense codons in human genetic disease: implications for gene therapy by nonsense suppressor tRNAs. Nucleic Acids Res 1994; 22:1327. Miller WL. Molecular biology of steroid hormone synthesis. Endocr Rev 1988; 9:295. Oohoshi H, Ohgawara H, Nanjo K, et al. Familial hyperproinsulinemia associated with NIDDM. Diabetes Care 1993; 16:1340. Abraham MR, Jahingir A, Alekseev AE, Terzic A. Channelopathies of inwardly rectifying potassium channels. FASEB J 1999; 13:1901. Goksel DL, Fischbach K, Duggirala R, et al. Variant in sulfonylurea receptor-1 gene is associated with high insulin concentrations in nondiabetic Mexican Americans: SUR-1 gene variant and hyperinsulinemia. Hum Genet 1998; 103:280. Duquesnoy P, Sobrier ML, Duriez B, et al. A single amino acid substitution in the exoplasmic domain of the human growth hormone GH receptor confers familial GH resistance (Laron syndrome) with positive GH-binding activity by abolishing receptor homodimerization. EMBO J 1994; 13:1386. Rosenfeld RG, Rosenbloom AL, Guevara-Aguirre J. Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocr Rev 1994; 15:369. Haas M, Forbush B III. The Na +-K+-Cl- cotransporters. J Bioenerg Biomembr 1998; 30:161. Torpy DJ, Chrousos GP. Hyper- and hypoaldosteronism. Vitam Horm 1999; 57:177. Glaser T, Lewis WH, Brung GH, et al. The beta subunit of follicle stimulating hormone is deleted in patients with aniridia and Wilms' tumor allowing a further definition of the WAGR locus. Nature 1986; 321:882. Sergi C, Serpi M, Müller-Navia J, et al. CATCH 22 syndrome: report of 7 infants with follow-up data and review of the recent advancements in the genetic knowledge of the locus 22q11. Pathologica 1999; 91:166. Stojdl DF, Bell JC. SR protein kinases: the splice of life. Biochem Cell Biol 1999; 77:293.

35a. Franklyn JA, Sheppard MC. Hormonal control of gene expression. Clin Endocrinol (Oxf) 1988; 29:337.

36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83.

Izumo S, Mahdavi V. Thyroid hormone receptor a isoforms generated by alternative splicing differentially activate myosin HC gene transcription. Nature 1988; 334:539. Koenig RJ, Lazar MA, Hodin RA, et al. Inhibition of thyroid hormone action by a non-hormone binding c-erbA protein generated by alternative mRNA splicing. Nature 1989; 337:659. Jung LJ, Kreiner T, Scheller RH. Prohormone structure governs proteolytic processing and sorting in the Golgi complex. Recent Prog Horm Res 1993; 48:415. Seidah NG, Day R, Marcinkiewicz M, Chrétien M. Mammalian paired basic amino acid convertases of prohormones and proproteins. Annals N Y Acad Sci 1993; 680:135. Belgorosky A, Rivarola MA. Progressive increase in non sex hormonebinding globulin-bound testosterone and estradiol from infancy to late puberty in girls. J Clin Endocrinol Metab 1988; 67:234. Tang KT, Yang HJ, Choo KB, et al. A point mutation in the albumin gene in a Chinese patient with familial dysalbuminemic hyperthyroxinemia. Eur J Endocrinol 1999; 141:374. Dahl KD, Bicsak TA, Hsueh AJW. Naturally occurring antihormones: secretion of FSH antagonists by women treated with a GnRH analog. Science 1988; 239:72. Wang C, Dahl KD, Leung A, et al. Serum bioactive follicle-stimulating hormone in men with idiopathic azoospermia and oligospermia. J Clin Endocrinol Metab 1987; 65:629. Bach JF. Antireceptor or antihormone autoimmunity and its relationship with the idiotype network. Adv Nephrol 1987; 16:251. Palmer JP, Asplin CM, Clemons P, et al. Insulin antibodies in insulindependent diabetics before insulin treatment. Science 1983; 222:1337. Yu L, Robles DT, Abiru N, et al. Early expression of antiinsulin autoantibodies of humans and the NOD mouse: evidence for early determination of subsequent diabetes. Proc Natl Acad Sci U S A 2000; 97:1701. Di Paolo S, Georgino R. Insulin resistance and hypoglycemia in a patient with systemic lupus erythematosus: description of antiinsulin receptor antibodies that enhance insulin binding and inhibit insulin action. J Clin Endocrinol Metab 1991; 73:650. Auclair M, Vigaroux C, Desbois-Mouthon C, et al. Antiinsulin receptor autoantibodies associate with insulin receptor substrate-1 and -2 and cause severe cell resistance to both insulin and insulin-like growth factor-I. J Clin Endocrinol Metab 1999; 84:3197. Chiovato L, Santini F, Vitti P, et al. Appearance of thyroid stimulating antibody and Graves disease after radioiodine therapy for toxic nodular goitre. Clin Endocrinol 1994; 40:803. Liberman UA, Eil C, Marx SJ. Clinical features of hereditary resistance to 1,25-dihydroxyvitamin D (hereditary hypocalcemic vitamin D resistant rickets type II). Adv Exp Med Biol 1986; 196:391. Chrousos GP, Detera-Wadleigh SD, Karl M. Syndromes of glucocorticoid resistance. Ann Intern Med 1993; 119:1113. Zennoro MC, Borensztein P, Seubrier F, et al. The enigma of pseudohypoaldosteronism. Steroids 1994; 59:96. Zoppi S, Wilson CM, Harbison MD, et al. Complete testicular feminization caused by an amino-terminal truncation of the androgen receptor with downstream initiation. J Clin Invest 1993; 91:1105. Talbert LM, Raj MH, Hammond MG, Greer T. Endocrine and immunologic ovary syndrome. Fertil Steril 1984; 42:7411. Fraser IS, Russell P, Greco S, Robertson DM. Resistant ovary syndrome and premature ovarian failure in young women with galactosemia. Clin Reprod Fertil 1986; 4:133. Suzuki Y, Hashimoto N, Shimada F, et al. Defects in insulin binding and receptor kinase in cells from a human with type A insulin resistance and from her family. Diabetologia 1991; 34:86. Bichet DG, Razi M, Lonergan M, et al. Hemodynamic and coagulation responses to 1-desamino (8-D-arginine) vasopressin in patients with congenital nephrogenic diabetes insipidus. N Engl J Med 1988; 318:881. Singer I, Forrest JN Jr. Drug-induced states of nephrogenic diabetes insipidus. Kidney Int 1976; 10:82. Jung Cy, Lee W. Glucose transporters and insulin action: some insights into diabetes management. Arch Pharm Res 1999; 22:329. Visser TJ, Kaptein E, Terpstra OT, Krenning EP. Deiodination of thyroid hormone by human liver. J Clin Endocrinol Metab 1988; 67:17. Bunnett NW. Postsecretory metabolism of peptides. Am Rev Respir Dis 1987; 136:S27. Roupas P, Herington AC. Receptor-mediated endocytosis and degradative processing of growth hormone by rat adipocytes in primary culture. Endocrinology 1987; 120:2158. Benzi L, Ceechetti P, Ciccarone A, et al. Insulin degradation in vitro and in vivo: a comparative study in men. Evidence that immunoprecipitable, partially rebindable degradation products are released from cells and circulate in blood. Diabetes 1994; 43:297. Yamaguchi T, Fukase M, Kido H, et al. Meprin is predominantly involved in parathyroid hormone degradation by the microvillar membranes of rat kidney. Life Sci 1994; 54:381. Ferrari P, Lovati E, Frey FJ. The role of the II beta-hydroxysteroid dehydrogenase type 2 in human hypertension. J Hypertens 2000; 18:241. Hum DW, Belanger A, Levesque E, et al. Characterization of UDP-glucuro-nyltransferases active on steroid hormones. J Steroid Biochem Mol Biol 1999; 69:413. National Center for Health Statistics. Monthly Vital Statistics Report 1999; 47(19):1. American Diabetes Association. Economic consequences of diabetes mellitus in the U.S. in 1977. Diabetes Care 1998; 21:296. Levetan CS, Passaro MD, Jablonski KA, Ratner RE. Effect of physician specialty on outcomes in diabetic ketoacidosis. Diabetes Care 1999; 22:1790. Baumann G. Growth hormone heterogeneity in human pituitary and plasma. Horm Res 1999; 51(Suppl 1):2. Roppolo HMN, Latchaw RE, Meyer JD, Curtin HD. Normal pituitary gland: 1. Macroscopic anatomy–CT correlation. Am J Neuroradiol 1983; 4:927. Tihansky DP, Crossen J, Markowitz H. Pseudotumor artifact of the dorsum sella in CT scanning. Comput Radiol 1987; 11:241. Hall WA, Luciano MG, Doppman JL, et al. Pituitary magnetic resonance imaging in normal human volunteers: occult adenomas in the general population. Ann Intern Med 1994; 120:817. Kucharczyk W, Peck WW, Kelly WM, et al. Rathke cleft cysts: CT, MR imaging, and pathologic features. Radiology 1987; 165:491. Hama S, Arita K, Tominago A. Symptomatic Rathke's cleft cyst coexisting with central diabetes insipidus and hypophysitis: case report. Endocr J 1999; 46:187. Fujisawa I, Asato R, Togashi K, et al. MR imaging of the sellar spine. J Comput Assist Tomogr 1988; 12:644. McGrath PC, Sloan DA, Schwartz RW, Kenady DE. Advances in the diagnosis and therapy of adrenal tumors. Curr Opin Oncol 1998; 10:52. Ratzmann GW, Zollner H. Hypomagnesemia and hypokalemia in the insulin hypoglycemia test. Z Gesamte Inn Med 1985; 40:567. Read RC, Doherty JE. Cardiovascular effects of induced insulin hypoglycemia in man during the Hollander test. Am J Surg 1972; 104:573. Sobel RJ, Ariad S. Adverse cardiovascular responses to thyrotropin-releasing hormone (200 micrograms) in cardiac patients. Isr J Med Sci 1987; 23:1107. Boice JD Jr. The danger of x-rays—real or apparent. N Engl J Med 1986; 315:828. Hash RB. Intravascular radiographic contrast media: issues for family physicians. J Am Board Fam Pract 1999; 12(1):32. Bravo EL, Tarazi RC, Dustan HP, et al. The changing clinical spectrum of primary aldosteronism. Am J Med 1983; 74:641.

CHAPTER 2 MOLECULAR BIOLOGY: PRESENT AND FUTURE Principles and Practice of Endocrinology and Metabolism

CHAPTER 2 MOLECULAR BIOLOGY: PRESENT AND FUTURE MEHBOOB A. HUSSAIN AND JOEL F. HABENER Cloning of Genes Genomic Libraries and Gene Isolation Gene Amplification by Polymerase Chain Reaction Variations of Polymerase Chain Reaction Approaches to the Quantitative Assessment of Gene Expression Transcription Assays Messenger RNA Assays Protein Expression Assays DNA-Protein Interaction Assays Genetic Manipulations in Animals in Vivo Transgenic Approaches Conditional (Developmental) Interruption of Gene Expression Prospects for the Future for Conditional Transgene Expression Expressed Sequence Tags DNA Arrays for the Profiling of Gene Expression Oligonucleotide Arrays (Genomic and Expressed Sequence Tags) Complementary DNA Arrays (Specific Tissues) Strategies for Mapping Genes on Chromosomes Genetic Linkage Maps and Quantitative Trait Loci Restriction Enzymes and Chromosomal Mapping Physical Maps Separating Chromosomes Somatic Cell Hybridization Chromosome Walking (Positional Cloning) Candidate Gene Approach Future Prospects Human Genome Project Stem Cells Somatic Cell Cloning in Vivo Gene Knock-Out Libraries Gene Therapy: Vectors and Problems Chapter References

The beginnings of molecular biology as a distinct discipline occurred in the late 1940s and early 1950s with the recognition that polynucleotides were the repository of genetic information in the form of DNA and the transmitters of genetic information in the form of messenger RNA (mRNA), and that transfer RNAs are fundamental for the assembly of amino acids into proteins. Detailed descriptions of the historical developments of this modern era of molecular biology are provided in several books.1,2,3 and 4 These were exciting times, as understanding progressed rapidly from the discovery by Avery and Brundage that DNA was a genetic substance; Chargaff established that DNA is composed of four different deoxyribonucleotides (dATP, dGRP, dTTP, dCTP); Watson and Crick elucidated the double-helical structure of DNA; Jacob and Monod identified mRNA as the intermediary in the transfer of information encoded in DNA to the assembly of amino acids into proteins; Holly discovered transfer RNAs; and Nirenberg et al. discovered the genetic code (i.e., each of the 21 amino acids is specified by a triplet of nucleotides, or codons, within the mRNA to be translated into a protein). In the 1960s, several major discoveries paved the way for the development of recombinant DNA technology and genetic engineering. Two of the major breakthroughs that made this possible were the discoveries of reverse transcriptase5 and restriction endonucleases,6,7 and 8 and techniques for determining the precise sequence of nucleotides in DNA.9,10 Reverse transcriptase, which is found encoded in the RNA of certain tumor viruses, is the means by which the virus makes DNA copies of its RNA templates. It allows molecular biologists to copy mRNA into complementary DNA (cDNA), which is an essential step in the preparation of recombinant DNA for purposes of cloning. Another fundamental discovery was that of restriction endonucleases, enzymes that cut DNA at specific sequences, typically of 4 to 10 base pairs. The application of specific restriction endonucleases allows for the cleavage of DNAs at precise locations, a property that is critical for the engineering of DNA segments. A most critical and important discovery was the technologic methodology to determine the sequential order of nucleotides in DNA. Both chemical and enzymatic approaches were developed. Currently, the nucleotide sequences of DNAs are determined by sophisticated automated instruments using random enzymatic cleavages of DNAs labeled with fluorescent markers. By fortunate coincidence, research into the mechanisms by which bacteria become resistant to certain antibiotics led to the discovery of bacterial plasmids, which are “viruses” that live within bacteria and lend genetic information to the bacteria to ensure their survival. Plasmids faithfully replicate within bacteria. Importantly, plasmid DNA is relatively simple in structure and is amenable to genetic engineering by excision of DNA sequences and insertion of foreign DNA sequences, which will replicate within bacteria without interference by the host bacterium. These plasmids have become useful vehicles in which to express and amplify foreign DNA sequences.

CLONING OF GENES Complementary DNA Libraries. The cloning of a particular expressed gene begins with the preparation and cloning of cDNAs from mRNAs of a particular cell (Fig. 2-1; Table 2-1) (for a more comprehensive description, see reference 11 and reference 12). The cDNAs are prepared by priming the reverse transcription of mRNAs, using reverse transcriptase and short oligonucleotide fragments of oligodeoxyribothymidine, which preferentially bind to the 3'-polyadenylate, or poly(A), tract that is characteristic of cellular mRNAs. Alternatively, random oligonucleotides of different base compositions may be used. Double-stranded DNA is then prepared from the single-stranded cDNA by using DNA polymerase, and the cDNAs are inserted into bacterial plasmids that have been cleaved at a single site with a restriction endonuclease. To ensure a reasonably high efficiency of insertion of the foreign DNA into the plasmids, cohesive, or “sticky,” ends are first prepared by adding short DNA sequences to the ends of the foreign DNA and to the plasmids. Vectors that are commonly used are derivatives of the plasmid pBR322, which was engineered specifically for the purposes of cloning DNA fragments (see Fig. 2-1). Foreign DNA is inserted into a unique site that is prepared by endonuclease cleaving of a desired site within a polylinker, multiple cloning site engineered into the plasmid. This site is often located within the gene that codes for bacterial b-galactosidase. The backbone plasmid also carries a gene for resistance to ampicillin or tetracycline. Thus, bacteria containing the plasmids can be selected by their resistance to ampicillin or tetracycline; those specifically containing DNA inserts can be selected by their inability to express b-galactosidase and to cleave b-galactopyranoside (blue-white screening).

FIGURE 2-1. An approach used in construction and molecular cloning of recombinant DNA. A, Preparation of double-stranded DNA from an mRNA template. The enzyme reverse transcriptase is used to reversetranscribe a single-stranded DNA copy complementary to the mRNA primed with an oligonucleotide of

polydeoxythymidylic acid hybridized to the poly(A) tract at the 3' end of mRNA. A complementary copy of the DNA strand is then prepared with DNA polymerase. Ends of double-stranded DNA are made flush by cleavage with the enzyme S1 nuclease, and homopolymer extensions of deoxycytidine are synthesized on 3' ends of DNA with the enzyme terminal transferase. Oligo(dC) homopolymer extensions form sticky ends for purposes of insertion of DNA into a linearized bacterial plasmid on which complementary oligo(dG) homopolymer extensions have been synthesized. B, Insertion of foreign DNA into a bacterial plasmid for molecular cloning. A bacterial plasmid, typically pBR322, that has been specifically engineered for purposes of cloning DNA is linearized by cleavage with restriction endonuclease Pst I. Poly(dG) homopolymer extensions are synthesized onto 3' ends of plasmid DNA. Foreign DNA with complementary poly(dC) homopolymer extensions is hybridized to and inserted into the plasmid. Recombinant plasmid DNA is transfected into susceptible host strains of bacteria, in which plasmid replicates apart from bacterial chromosomal DNA. Bacteria are then grown on a plate containing tetracycline. Colonies that are resistant to tetracycline are tested for sensitivity to ampicillin. Because native plasmids contain genes encoding resistance to both tetracycline and ampicillin and the gene encoding resistance to ampicillin is inactivated by insertion of a foreign DNA at the Pst I site, bacterial colonies harboring plasmids with DNA inserts are resistant to tetracycline and sensitive to ampicillin. Subsequent screening of tetracycline-resistant, ampicillin-sensitive clones containing specific DNA-inserted sequences is carried out by either DNA hybridization with labeled DNA probes or by other techniques such as hybridization arrest and cell-free translation.

TABLE 2-1. Approaches for the Selection of Cloned Complementary DNAs (cDNAs)

Hybridization Screening. The recombinant plasmids containing DNA sequences that are complementary to the specific mRNAs of interest are identified by hybridizing recombinant plasmids to the initial mRNA preparations used in the cloning. The hybrid-selected mRNA is subsequently eluted and translated in a cell-free system appropriate for the protein under study. Alternatively, specific inhibition of the translation of an mRNA can be used to identify the DNA of interest: DNA that is complementary to the mRNA being translated will bind the RNA, thus precluding translation and reducing the amount of the protein being synthesized. The initial techniques of hybridization selection and hybridization arrest, in which cell-free translation is used as the assay system, are now supplanted by hybridization of the bacterial colonies with synthetic oligonucleotide probes that are labeled with phosphorus-32 (32P). Mixtures of oligonucleotides in the range of 14 to 17 bases are prepared that are complementary to the nucleotide sequences predicted from the known amino-acid sequences of segments of the protein encoded by mRNA. Because of the degeneracy in the genetic code (there are 61 amino-acid codons and 20 amino acids), mixtures of from 24 to 48 oligonucleotides ordinarily represent all possible sequences complementary to a particular 14- to 17-base region of mRNA. Expression Screening. Later-generation cDNA libraries have been prepared in bacterial phages (l gt-11) or hybrids between plasmids and phages (phagemids), which have been engineered to allow the bacteria infected with the recombinant phages to translate mRNAs expressed from the cDNAs, and thereby to produce the protein products encoded by the cDNAs. The desired sequence of interest can be selected at the protein level by screening the library of bacterial clones with an antiserum directed to the protein. When the desired product is a DNA-binding protein, the library can be screened with a labeled DNA duplex containing copies of the target sequence to which the protein binds. Yeast Two-Site Interaction Trap. The cloning of cDNAs encoding proteins that interact with other known proteins can be accomplished using the yeast two-site interaction trap, which functions much as a bait and fish system. The bait is a cDNA encoding a known protein that is engineered to bind to an enhancer in the promoter of a gene that encodes a factor essential for the survival of a yeast cell. The sequences (fish) in the cDNA library are engineered with a strong transcriptional transactivation domain, such as that from the herpes simplex virus and yeast transcription factors VP16 or Gal-4, respectively. The occurrence of proteinprotein interactions between the bait and one of the fish activates the expression of the yeast survival gene, which thereby allows for the selection and cloning of the yeast cell that harbors the described cDNA sequence from the cDNA library. Rapid Amplification of Complementary DNA Ends. Most often cDNAs isolated by one or more of the approaches described above lack the complete sequence and are deficient in the 5' ends. The 5' sequences are determined by using the rapid amplification of cDNA ends (RACE) technique.

GENOMIC LIBRARIES AND GENE ISOLATION Southern Blots and Hybridization Screening. The techniques used in the cloning of genomic DNA are similar to those used for cloning cDNA, except that the genomic sequences are longer than the cDNA sequences and different cloning vectors are required. The common vectors are derivatives of the bacteriophage l that can accommodate DNA fragments of 10 to 20 kilobases (kb). Certain hybrids of bacteriophages and plasmids, called cosmids, can accommodate inserts of DNA of up to 40 to 50 kb. Even larger segments of DNA up to 1 to 2 megabases (Mb) can be cloned and propagated in yeast and are called yeast artificial chromosomes (YACs). In the cloning of genomic DNA, restriction fragments are prepared by partial digestion of unsheared DNA with a restriction endonuclease that cleaves the DNA into many fragments. DNA fragments of proper size are prepared by fractionation on agarose gels and are ligated to the bacteriophage DNA. The fragments of DNA containing the desired sequences can be detected by hybridization of a membrane blot prepared from the gel with a 32P-labeled cDNA, a Southern blot. The recombinant DNA is mixed with bacteriophage proteins, which results in the production of viable phage particles. The recombinant bacteriophages are grown on agar plates covered with growing bacteria. Then the bacteria are infected by a phage particle, which lyses the bacteria to form visible plaques. Specific phage colonies are transferred by nitrocellulose filters and are hybridized by cDNA probes labeled with 32P, similar to a Southern blot. Libraries of genomic DNA fragments and tissue-specific cDNAs from various animal species cloned in plasmids and bacteriophages are available from a number of commercial laboratories. The development of yeast chromosomal libraries that harbor large segments (several megabases) of chromosomal DNA has markedly accelerated the generation of gene linkage maps. Enhancer Traps. One approach to identifying novel genes imbedded in the genome is to randomly insert a transcriptional reporter gene into chromosomal DNA that has been cleaved into 1- to 2-kb fragments by digestion with a restriction endonuclease. The family of ligated hybrid fragments is then cloned into plasmids that are individually introduced (transfected) into host cell lines (e.g., NIH or BHK fibroblasts). After the transfected cell lines are incubated with the cloned DNA fragments for 1 to 2 days, extracts are prepared from the cells and assayed for expression of the transcriptional reporter gene. Typical transcriptional reporter genes used are firefly luciferase, bacterial chloramphenicol acetyl transferase, or bacterial alkaline phosphatase. When, by chance, a transcriptional enhancer is encountered, as determined by the activation of the reporter gene, the particular cloned DNA fragment is sequenced and searched for transcribed exonic and/or intronic sequences of genes, many of which typically reside 100 to 1000 base pairs from the enhancer sequence. The transcribed sequences of genes usually, but not always, reside 3' (downstream) from enhancer sequences. Rapid Amplification of Genomic DNA Ends. The principle of rapid amplification of genomic DNA ends (RAGE) is similar to that of RACE previously described and allows for the identification of unknown DNA sequences in genomic DNA. Oligonucleotide primers (amplimers) are annealed to the test genomic DNA sample and extended on the genomic DNA template with DNA polymerase, and a second set of oligonucleotide primers is ligated to the extended ends. The extended DNA fragments are then amplified by polymerase chain reaction (see next section), isolated by electrophoresis on agarose gels, and sequenced.

GENE AMPLIFICATION BY POLYMERASE CHAIN REACTION The development of the polymerase chain reaction PCR, a technique for the rapid amplification of specific DNA sequences, constituted a major technological breakthrough.13,14,15 and 16 This procedure relies on the unique properties of a thermally stable DNA polymerase (Taq polymerase) to allow for sequential annealing of small oligonucleotide primers that bracket a DNA sequence of interest; the result is successive synthesis of the DNA strands. Specific DNA sequences as short as 50 and as long as several thousand base pairs can be amplified over a million-fold in just a few hours by using an automated thermal cycler. The technique is so sensitive that DNA (genomic DNA or cDNA reverse-transcribed from RNA) from a single cell can be so amplified. Indeed, a sample containing only a single target DNA molecule

can be amplified. The applications of this technique are diverse. Not only is it possible to amplify and to clone rare sequences for detailed studies, but also the technique has applications in the fields of medical diagnosis and forensics. Scarce viruses can be detected in a drop of serum or urine or a single white blood cell. Genotyping can be done from a blood or semen stain, saliva, or a single hair. Paradoxically, a major drawback of PCR is its exquisite sensitivity, which leaves open the possibility of false-positive results because of minute contaminations of the samples being tested. Thus, extreme precautions must be taken to avoid the introduction of contaminants. PCR is carried out using DNA polymerase and oligonucleotide primers complementary to the two 3' borders of the duplex segment to be amplified. The objective of PCR is to copy the sequence of each strand between the regions at which the oligonucleotide primers anneal. Thus, after the primers are annealed to a denatured DNA containing the segment to be amplified, the primers are extended using DNA polymerase and the four deoxynucleotide triphosphates. Each primer is extended toward the other primer. The result is a double-stranded DNA (which itself is then denatured and annealed again with primer, and the DNA polymerase reaction is repeated). This cycle of steps (denaturation, annealing, and synthesis) may be repeated 60 times. At each cycle, the amount of duplex DNA segment doubles, because both new and old DNA molecules anneal to the primers and are copied. In principle (and virtually in practice), 2n copies (where n = number of cycles) of the duplex segment bordered by the primers are produced. The heat-stable polymerase isolated from thermophilic bacteria (Thermophilus aquaticus), Taq polymerase, allows multiple cycles to be carried out after a single addition of enzyme. The DNA, an excess of primer molecules, the deoxynucleotide triphosphates, and the polymerase are mixed together at the start. Cycle 1 is initiated by heating to a temperature adequate to assure DNA denaturation, followed by cooling to a temperature appropriate for primer annealing to the now-single strands of the template DNA. Thereafter, the temperature is adjusted for DNA synthesis (elongation) to occur. The subsequent cycles are initiated by again heating to the denaturation temperature. Thus, cycling can be automated by using a computer-controlled variable-temperature heating block. In addition to permitting automation, the use of the DNA polymerase of T. aquaticus has another advantage. The enzyme is most active between 70° and 75°C. Base pairing between the oligonucleotide primers and the DNA is more specific at this temperature than at 37°C, the optimal functioning temperature of Escherichia coli DNA polymerase. Consequently, the primers are less likely to anneal nonspecifically to unwanted DNA segments, especially when the entire genome is present in the target DNA. VARIATIONS OF POLYMERASE CHAIN REACTION Simple modifications of the PCR conditions can expand the opportunities of the PCR. For example, synthesizing oligonucleotide primers that recognize domains (motifs) shared by cDNAs and their respective protein products, and choosing less stringent annealing conditions for the primers, permit new sequences of yet unknown DNAs to be generated with PCR, ultimately resulting in the discovery of new cDNAs belonging to the same family. For example, the pancreatic B-cell transcription factor IDX-1 was identified by PCR using oligonucleotide primers that would anneal to sequences shared by the homeodomain transcription factor family. PCR primers can be modified in their sequence and thus are not completely complementary to the template DNA. The amplified PCR product then carries the sequence of the primer and not the original DNA sequence. This strategy can be used to insert mutations site-specifically into known DNA sequences.

APPROACHES TO THE QUANTITATIVE ASSESSMENT OF GENE EXPRESSION TRANSCRIPTION ASSAYS Nuclear Run-On Assays. Several assays are available that provide an index of relative rates of gene transcription (Fig. 2-2). A simple, straightforward assay is the nuclear run-on assay in which nuclei are isolated from tissue culture cells and nascent RNA chains are allowed to continue to polymerize in the presence of radiolabeled deoxyribonucleotides in vitro. This assay has the advantage that it surveys the density of nascent transcripts made from the endogenous genes of cells and, on average, is a good measure of gene transcription rates in response to the existing environmental conditions in which the cultured cells are maintained. Newly synthesized RNA is applied (hybridized) to a nylon membrane on which a cDNA target complementary to the desired RNA has been adsorbed. Radiolabeled RNA hybridized to the cDNA is determined in a radiation counter.

FIGURE 2-2. Approaches to the quantitative assessment of gene expression. Shown are the various types of assays that can be used to examine regulation of gene expression at various levels. (mRNA, messenger RNA; RNase, ribonuclease; RT-PCR, reverse transcription polymerase chain reaction.)

Cell-Free In Vitro Systems. Rates of RNA synthesis can also be determined in broken cell or cell-free lysates to assess the relative strengths of different promoters. To restrict the newly synthesized radiolabeled RNA to a single size and, thus, to enable more ready detection by electrophoresis, a DNA template is used that does not contain guanine bases, called a G-free cassette. RNA synthesis is carried out in the absence of the guanine nucleotide. After synthesis of a specified length of RNA at the end of which guanine bases are encountered, RNA synthesis is terminated. Transfection of Promoter-Reporters in In Vivo Cell Culture. Many of the currently used assays of gene transcription employ promoter sequences fused to genes encoding proteins that can be quantitated by bioassays (e.g., bacterial chloramphenicol acetyl transferase, firefly luciferase, alkaline phosphatase, or green fluorescent protein). The hybrid DNAs, so called promoter-reporter DNAs, are introduced into tissue culture cells by one of several chemical methods (i.e., DNA adsorbed to calcium phosphate precipitates, diethylaminoethyl (DEAE)-dextran incorporated into liposomes, or human artificial chromosomes [Table 2-2]); or physical methods (i.e., electroporation, direct microinjection of DNA, or ballistic injection using a gene gun [Table 2-3]). After introduction of the reporter DNA into the cells, the transfected cells are incubated for a specified time under the desired experimental conditions, the cells are harvested, and extracts are prepared for assays of the reporter-specific enzymatic activity. By these transfection methods, cell-type specificity for the expression of genepromoter sequences can be determined by comparing promoter-reporter efficiencies in cells of different phenotypes. In addition, important transcriptional control sequences in the promoter can be mapped by DNA mutagenesis studies.

TABLE 2-2. Chemical Methods for Introducing Genes into Mammalian Cells

TABLE 2-3. Physical Methods for Introducing Genes into Mammalian Cells

Transfection of Transcription Factor Expression Vectors. An extension of the promoter-reporter transfection approach is to cotransfect recombinant expression plasmids encoding transcription factors that bind to control sequences in the promoter DNA and activate transcription of the reporter. By this approach, critical functional components of transcription factors and critical bases in DNA control sequences can be examined experimentally. Transgenic In Vivo Mouse Models. A method developed for examining specificity of tissue expression and efficiency of expression of promoter-reporter genes is their introduction into mice in vivo, using transgenic technology (see the section Genetic Manipulations in Animals In Vivo). Recombinant promoter-reporter genes are injected into the pronucleus of fertilized mouse ova and implanted into surrogate females. The tissues of transgenic neonatal mice are examined for the tissue distribution and relative strength of the expression of the reporter function. Commonly used reporter functions are the genes encoding either b-galactosidase or green fluorescent protein. MESSENGER RNA ASSAYS Northern Blot Hybridization. RNA blotting (Northern blotting) is analogous to DNA blotting (Southern blotting). RNA is separated according to size by electrophoresis through agarose gels. Generally, the electrophoresis is performed under conditions that denature the RNA so that the effects of RNA secondary structure on the electrophoretic mobility of the RNAs can be minimized. Alkaline conditions are unsuitable; therefore, agents such as glyoxal, formaldehyde, or urea are used. The size-separated RNA is transferred by blotting to an immobilizing membrane without disturbing the RNA distribution along the gel. A labeled DNA is then used as a probe to find the position on the blot of RNA molecules corresponding to the probe. The immobilized RNA is incubated with DNA under conditions allowing annealing of the DNA to the RNA on the immobilized matrix. After washing away excess and unspecifically annealed DNA, the matrix is exposed to an x-ray film to detect the position of the probe. RNA blotting allows the estimation of the size of the RNA that is being detected. In addition, the intensity of the band on an x-ray film indicates the abundance of the RNA in the cell or tissue from which the RNA was extracted. Solution Hybridization Ribonuclease Protection. To obtain more precise information on the amount of a specific RNA species in a certain cell or tissue, a single-stranded radioactive probe is generated that is complementary to a portion of the RNA being studied. An excess amount of this single-stranded probe is then mixed in solution with the total RNA of the cells or tissue being investigated. Digestion with ribonuclease of all single-stranded nucleic acids present after hybridization leaves the double-stranded species, consisting of the labeled probe annealed to its complementary RNA, in the solution. The contents of the solution are then size-separated on an electrophoretic gel, which is exposed to an x-ray film. Knowing the amount of input labeled single-stranded probe allows a quantification of the specific RNA present in the total RNA of the cells or tissue. In Situ Hybridization. In situ hybridization with labeled single-stranded probes onto tissues is, in principle, similar to the ribonuclease protection assay. Detection and determination of the location of a certain species of RNA within a tissue is possible. Reverse Transcription Polymerase Chain Reaction. Reverse transcription polymerase chain reaction (RT-PCR) can be used to quantitate the abundance of a specific RNA. This method is particularly practical when small amounts of tissue or cells are available to be analyzed. The RNA is reverse-transcribed to DNA with reverse transcriptase. The cDNA population is then subjected to PCR amplification with specific primers that recognize the cDNA in question. By choosing the number of PCR cycles within the linear range of product generated after each cycle (i.e., enough primers, nucleotides, and DNA polymerase in the reaction mixture for none of them to be the limiting factor of the reaction) and adding to the PCR reaction a defined amount of an artificial DNA template that is also recognized by the primer oli gonucleotides but yields a differentsized product, one can detect differences in abundance of cDNA (and hence RNA in the original sample) among two or more samples. Newer methods allow for an on-line monitoring of each PCR reaction of the product generated. This is achieved by using primer oligonucleotides that can be monitored during the PCR reaction cycles by external optical devices. Such on-line continuous monitoring allows the performance of PCR reactions without prior determination of the number of cycles required to keep the PCR reaction within the linear range of amplification. Continuous PCR monitoring provides immediate information on abundance of a given cDNA species in PCR reactions. Knowledge of the absolute amount of labeled oligonucleotide primer added to the PCR reaction at the start can be used to determine the exact amount of the PCR product generated. PROTEIN EXPRESSION ASSAYS Cell-Free Translation. A commonly used method to analyze proteins encoded by mRNA is to translate the mRNA in cell-free translation systems in vitro. By this method, proteins can be radioactively labeled to a high specific activity. The cell-free translation also provides the primary protein product, such as a proprotein or prohormone, encoded by the mRNA. Pulse and Pulse-Chase Labeling. Studies of protein syntheses can also be carried out in vivo by incubation of cultured cells or tissues with radioactive amino acids (pulse labeling). Posttranslational processing (e.g., enzymatic cleavages of prohormones) can be assessed by first incubating the cells or tissues for a short time with radioactive amino acids and then incubating them for an additional period with unlabeled amino acids (pulse-chase labeling). Western Immunoblot. Another approach to the analyses of particular cellular proteins is the Western immunoblot technique. Proteins in cell extracts are separated by electrophoresis on polyacrylamide or agarose gels and transferred to a nylon or nitrocellulose membrane, which is then treated with a solution containing specific antibodies to the protein of interest. The antibodies that are bound to the protein fixed to the membrane are detected by any one of several methods, such as secondary antibodies tagged with radioisotopes, fluorophores, or enzymes. Immunocytochemistry. A refinement of the Western immunoblot technique is the detection of specific proteins within cells by immunocytochemistry (immunohistochemistry). Cultured cells or tissue sections are fixed on microscope slides and treated with solutions containing specific antibodies. The antibodies that are bound to the proteins within the cells are detected with fluorescently tagged secondary antibodies or by an avidin-biotin complex. Immunocytochemistry is a powerful technique when used for the simultaneous detection of two or even three different proteins with examination by confocal microscopy. DNA-PROTEIN INTERACTION ASSAYS ELECTROPHORETIC MOBILITY GEL SHIFT AND SOUTHWESTERN BLOTS The binding of proteins such as transcription factors to DNA sequences is commonly done by two approaches: electrophoretic mobility shift assay (EMSA) and Southwestern blotting. Typically, EMSA consists of incubation of protein extracts with a radiolabeled DNA sequence or probe. The mixture is then analyzed by electrophoresis on a nondenaturing polyacrylamide gel, followed by autoradiography or autofluorography to evaluate the distribution of the radioactivity or fluorescence in the gel. Interactions of specific proteins with the DNA probe are manifested by a retardation of the electrophoretic migration of the labeled probe, or band shift. The EMSA technique can be extended to include antibodies to specific proteins in the incubation mixture. The interaction of a specific antibody with a protein bound to the DNA probe causes a further retardation of migration of the DNA-protein complex, leading to a super shift. PROTEIN-PROTEIN INTERACTION ASSAYS

A number of different assays are used to determine and evaluate protein-protein interactions. Two in vitro assays are coimmunoprecipitation and polyhistidine-tagged glutathione sulfonyl transferase (GST) pull-down. Two in vivo assays are the yeast and mammalian two-site interaction assays. Coimmunoprecipitation. The commonly used coimmunoprecipitation assay makes use of antisera to specific proteins. In circumstances in which two different proteins, A and B, associate with each other, an antiserum to protein A will immunoprecipitate not only protein A, but also protein B. Likewise, an antiserum to protein B will coimmunoprecipitate proteins B and A. In practice, the proteins under investigation are radiolabeled by synthesis in the presence of radioactive amino acids, either in cell-free transcriptiontranslation systems in vitro, or in cell culture systems in vivo. Coimmunoprecipitated proteins are detected by gel electrophoresis and autoradiography. Alternatively, the proteins so immunoprecipitated or coimmunoprecipitated can be assayed by Western immunoblot techniques. Glutathione Sulfonyl Transferase Pull-Down. GST is an enzyme that has a high affinity for its substrate, glutathione. This property of high-affinity interactions has been exploited to develop a cloning vector plasmid encoding GST and containing a polylinker site that allows for the insertion of coding sequences for any protein of interest. Thus, if protein A is believed to interact with protein B, the coding sequence for either protein A or protein B can be inserted into the GST vector. The GST–protein A or B fusion protein is synthesized in large amounts by multiplication and expression of the plasmid vector in bacteria. The GST-fusion protein is then incubated with either labeled or unlabeled proteins in extracts of cells or nuclei. Proteins in the extracts bound to protein A or B in the GST-fusion protein are pulled down from the extracts by capturing the GST on glutathione-agarose beads. Proteins are released from the beads and analyzed by either gel electrophoresis and autoradiography (labeled proteins) or by Western immunoblot (unlabeled proteins). Similar methods using polyhistidine tag in place of GST are also used for pull-down experiments. Far Western Protein Blots. A variation on the Western blotting technique is the Far Western blot. In this technique, a radio-labeled or fluorescence-labeled known protein (instead of an antibody) is applied to a membrane to which proteins from an electrophoretic gel have been transferred. If the known protein binds to any one or more proteins on the membrane, it is detected as a labeled band by autoradiograph or autofluorography. Relatively strong protein-protein interactions are required for this approach to succeed. Yeast and Mammalian Cell Two-Site Interaction Traps. The two-site interaction traps are useful for demonstrating protein-protein interactions in vivo. The principle of the approach is that, when a specific protein-protein interaction occurs, it reconstitutes an active transcription factor which then activates the transcription of a reporter gene. The cells (yeast or mammalian) are programmed to constitutively express a strong DNA-binding domain, such as Gal-4, fused to the expression sequence of the selected protein, protein A (the bait). The cells are also programmed to express a transcriptional reporter (e.g., CAT or luciferase linked to a promoter) that has binding sites for Gal-4. Thus, protein A anchors to the DNA-binding site of the reporter promoter via the Gal-4 binding domain but does not activate transcription of the reporter gene, and no reporter function is expressed. Protein B, however, is expressed as a fusion protein with a strong transcriptional activator sequence (e.g., the transcriptional transactivation domain of Gal-4 or of VP16). This transcriptional activation domain–protein B fusion protein does not bind DNA, but when, or if, protein B physically interacts with (binds to) protein A, a fully active transcription factor is reconstituted, the promoter reporter gene is transcribed, and the reported function is expressed. The yeast two-hybrid system can be used to clone proteins that interact with a bait protein such as protein A. In this instance, a cDNA protein expression library is prepared or obtained that has all of the cDNAs of a given tissue fused to a coding sequence for a transactivation domain (e.g., VP16). Further, the reporter consists of a survival factor essential for the growth of the yeast cell. Thus, when a cDNA encodes a protein B (fish) that interacts with the bait protein A, the yeast cell expresses the survival protein and survives, whereas the other yeast cells die.

GENETIC MANIPULATIONS IN ANIMALS IN VIVO TRANSGENIC APPROACHES To create transgenic mice, DNA is injected into the male pronucleus of one-cell mammalian embryos (fertilized ova) that are then allowed to develop by insertion into the reproductive tract of pseudopregnant foster mothers (Fig. 2-3A). The transgenic animals that develop from this procedure contain the foreign DNA integrated into one or more of the host chromosomes at an early stage of embryo development. As a consequence, the foreign DNA is generally transmitted to the germline, and, in a number of instances, the foreign genes are expressed. Because the foreign DNA is injected at the one-cell stage, a good chance exists that the DNA will be distributed among all the progeny cells as development proceeds. This situation provides an opportunity to analyze and compare the qualitative and quantitative efficiencies of expression of the genes among various organs. The technique is quite efficient; >50% of postinjection ova produce viable offspring, and, of these, ~10% efficiently carry the foreign genes. In the transgenic animals, the foreign genes can be passed on and expressed at high levels in subsequent generations of progeny.

FIGURE 2-3. Approaches for (A) the integration of a foreign gene into the germline of mice, and (B) disruption or knock-out of a specific gene. A, DNA containing a specific foreign gene is microinjected into the male pronucleus of fertilized ova obtained from the oviduct of a mouse. Ova are then implanted into the uterus of pseudopregnant surrogate mothers. Progeny are analyzed for the presence of foreign genes by hybridization with 32P-labeled DNA probe and DNA prepared from a piece of tail from a mouse, which has been immobilized on a nitrocellulose filter (tail blots). B, To create a knock-out of a gene, pluripotential embryonic stem (Es) cells are used in vitro to introduce an engineered plasmid DNA sequence that will recombine with a homologous gene that is targeted. The recombination excises a portion of the gene in the ES cells, rendering it inactive (no longer expressible). ES cells in which the homologous recombination occurred successfully are selected by a combined positive-negative drug selection. The engineered ES cells are injected into the blastocoele of 3.5-day blastocysts that are then implanted into the uterus of pseudopregnant mice. The offspring are both chimeric and germline for expression of the knock-out gene and must be cross-bred to homozygosity for the genotype of a complete knock-out of the gene that is targeted for disruption.

Transgenic approaches can also be used to prevent the development of the lineage of a particular cell phenotype or to impair the expression of a selected gene. A cell lineage can be ablated by targeting a microinjected DNA containing a subunit of the diphtheria toxin to a particular cell type, using a promoter sequence specifically expressed in that cell type. The diphtheria toxin subunit inhibits protein synthesis when expressed in a cell, thereby killing the cell. The expression of a particular gene can be impaired by similar cell promoter– specific targeting of a DNA expression vector to a cell that produces an antisense mRNA to the mRNA expressed by the gene of interest. The antisense mRNA hybridizes to nuclear transcripts and processed mRNAs; this results in their degradation by double-stranded RNA–specific nucleases, thereby effectively attenuating the functional expression of the gene. The efficacy of the impairment of the mRNA can be enhanced by incorporating a ribozyme hammerhead sequence in the expressed antisense mRNA so as physically to cleave the mRNA to which it hybridizes. Another approach to producing a particular gene loss of function is to direct expression of a dominant negative protein (e.g., a receptor made deficient in intracellular signaling by an appropriate mutation, or a mutant transcription factor deficient in transactivation functions but sufficient for DNA binding). These dominant negative proteins compete for the essential functions of the wild-type proteins, resulting in a net loss of function. Another approach, termed targeted transgenesis, combines targeted homologous recombination in embryonic stem (ES) cells with gain-of-function transgenic approaches.11,12,17 This method allows for targeted integration of a single-copy transgene to a single desired locus in the genome and thereby avoids problems of random and multiple-copy integrations, which may compromise faithful expression of the transgene in the conventional approach. GENE ABLATION (KNOCK-OUTS) A major advance beyond the gain-of-function transgenic mouse technique has been the development of methods for producing loss of function by targeted disruption or replacement of genes. This approach uses the techniques of homologous recombination in cultured pluripotential ES cells, which are then injected into mouse

blastocysts and implanted into the uteri of pseudopregnant mice (Fig. 2-3B). The targeting vector contains a core replacement sequence consisting of an expressed-cell lethal-drug resistance gene (selectable marker) (e.g., neomycin [Pgk-neo]) flanked by sequences homologous to the targeted cellular gene, and a second selectable marker gene (e.g., thymidine kinase [pgk-tk]). The ES cells are transfected with the gene-specific targeting vector. Cells that take up vector DNA and in which homologous recombination occurs are selected by their resistance to neomycin (positive selection). To select against random integration, a susceptibility to killing by thymidine kinase (negative selection) is used; only homologous recombination in which the thymidine kinase gene has been lost will confer survival benefit. Because the ES cells are injected into multicellular 3.5-day blastocysts, many of the offspring are mosaics, but some are germline heterozygous for the recombined gene. F1 generation mice are then bred to homozygosity so as to manifest the phenotype of the gene knock-out. Using this approach of targeted gene disruption, literally thousands of knock-out mice have been created. Many of these knock-out mice are models for human genetic disorders (e.g., those of endocrine systems such as pancreatic agenesis [homeodomain protein IDX-1], familial hypocalciuric hypercalcemia [calcium receptor], intrauterine growth retardation [insulinlike growth factor-II receptor], salt-sensitive hypertension [atrial natriuretic peptide], and obesity [a3-adrenergic receptor]). CONDITIONAL (DEVELOPMENTAL) INTERRUPTION OF GENE EXPRESSION Although targeted transgenesis using chosen site integration and targeted disruption of genes has proven helpful in analyses of the functions of genes, conditionally to induce expression of transgenes or conditionally to inactivate a specific gene is useful. Early on, randomly integrated vectors for the expression of transgenes used the metallothionein promoter that is readily inducible by the administration of heavy metals to transgenic mice. Now techniques have been developed to conditionally inactivate targeted genes in a defined spatial and temporal pattern. Several approaches to achieve conditional gene inactivation have been developed. Two of these approaches are (a) the Cre recombinase–loxP system (Fig. 2-4)18 and (b) the tetracycline-inducible transactivator vector (tTA) system (Fig. 2-5).19 Occasionally, both of these systems have been used effectively to knock out (Cre-loxP) or to attenuate (reverse tTA) the expression of specific genes. Both the Cre-loxP and reverse tTA systems require the creation of two independent strains of transgenic mice, which are then crossed to produce double transgenic mice.

FIGURE 2-4. Schema of the Cre-loxP approach to conditionally knock out a specifically targeted gene in mice. A, The approach requires the creation of two separate strains of transgenic mice that are crossed to produce double transgenic mice to effect the conditional gene knock-outs. One mouse strain is created so as to replace the gene of interest by one that has been flanked by loxP recombination sequences (floxed), using targeted recombinational gene replacement in embryonic stem cells as illustrated in Figure 2-3B. The other mouse strain is a transgenic mouse in which the Cre recombinase enzyme expression vector is targeted to the tissue of interest using a tissue-specific promoter, such as the proinsulin gene promoter, to target and restrict expression to pancreatic B cells. B, A more detailed depiction of the strategy for preparation of the gene replacement by homologous recombination to generate mice with a floxed gene. This approach is similar to that described in Figure 2-3B to create knock-out mice.18

FIGURE 2-5. Diagram showing the approach to reversible conditional expression of a gene in mice, using a tetracycline-inducible gene expression system. A, As in the Cre-loxP system (see Fig. 2-4A), the tetracyclineinducible gene system requires the creation of two independent strains of transgenic mice. One strain of mice targets the expression of a specially engineered transcription factor (rtTA) to the tissue of interest, using a tissue-specific promoter (TSP). B, the rtTA transcription factor consists of a modification of the bacterial tetracycline-responsive repressor that has been genetically engineered so as to convert it into a transcriptional transactivator when exposed to tetracycline or one of its analogs. The other mouse strain is one in which a gene of interest is introduced, usually driven by a ubiquitous promoter such as a viral promoter (CMV, RSV) or an actin promoter. The gene of interest could be one encoding an antisense RNA to a messenger RNA of a protein that is to be knocked out. The creation of double transgenic mice then allows for the expression of the gene of interest in a specific tissue under the control of the induced tetracycline. (See text for more detailed description.57) (tet op, tetracycline resistance operon; P, promoter; As, antisense; TPE, tissue promoter element.)

CRE RECOMBINASE–LOXP SYSTEM The Cre-loxP approach is based on the Cre-loxP recombination system of bacteriophage P1 (see Fig. 2-4). This system is capable of mediating loxP sitespecific recombination in embryonal stem cells and in transgenic mice. Conditional targeting requires the generation of two mouse strains. One transgenic strain expresses the Cre recombinase under control of a promoter that is cell-type specific or developmental stage specific. The other strain is prepared by using ES cells to effect a replacement of the targeted gene with an exact copy that is flanked by loxP sequences required for recombination by the Cre recombinase. The recombined gene is said to be floxed. The presence of the loxP sites does not interfere with the functional expression of the gene and will be normally expressed in all of its usual tissues not coexpressing the Cre recombinase. In those tissues in which the Cre is expressed by virtue of its tissue-specific promoter, the target gene will be deleted by homologous recombination. Thus, the Cre-loxP system acts like a timer in which the events that are to take place are predetermined by the prior reprogramming of the genes: the target gene will be ablated during development where and when the promoter chosen to drive the expression of Cre is activated. Thus, a disadvantage of the Cre-loxP system is the lack of control over when the gene knock-out will take place, because it is preprogrammed in the system. Newer genetically engineered Cre derivatives allow for pharmacologic activation of the recombinant event. A potential advantage of the Cre-loxP system is that one can theoretically generate extensive collections of mice expressing the Cre recombinase specifically and individually in many different tissues so that these mice could be made commercially available to investigators. CONDITIONAL TETRACYCLINE-INDUCIBLE FORWARD AND REVERSE TRANSACTIVATOR VECTOR SYSTEMS The Cre-loxP system leads to the irreversible targeted disruption of a particular gene at the time that the promoter encoding the Cre recombinase is activated during development. Having available a system that can be reversibly activated at any time would be desirable. A system that holds promise in this regard is the tetracycline-inducible transactivator vector (forward or reverse tTA), which, in response to tetracycline, switches on a specific gene bearing a promoter containing the tetracycline-responsive operon (see Fig. 2-5). This system allows any recombinant gene marked by the presence of the tet operon to be turned on or off at will simply by the administration of a potent tetracycline analog to the transgenic mice. The vectors were engineered from the sequences of the E. coli bacterial tetracycline resistance operon (tet op), in which a repressor sits on the operon, keeping the resistance gene off. When tetracycline binds to the repressor, it is deactivated, falls off of the operon, and turns on the gene. First, the repressor was converted into an activator by fusing the DNA-binding domain to the potent activator sequence (VP16) of the herpes simplex virus. In this system, tetracycline turned off the activator (tet-off) and thereby caused failure of expression of target genes containing the tet operon binding sites for the repressor turned into an activator. This tTA system required the continued presence of tetracycline to keep the gene off and withdrawal of the tetracycline to turn on the gene, raising problems of long and variable clearance times for the drugs. Turning the gene on by administration of tetracycline (tet-on) would be preferable. Therefore, the tTA vector was reengineered to reverse the action of tetracycline: in the current vector system, the binding of tetracycline to the reverse

tTA enhances its binding to the tet operon. Theoretically, as the reverse tTA system now works, any gene can be reversibly turned on by the administration of tetracycline or one if its more potent analogs in the double transgenic mouse, which consists of a cross between a mouse that has the reverse tTA targeted to express in a specific tissue and a mouse that has a ubiquitously expressed transgene for any gene X under the control of the tet operon. The equivalent of gene knock-outs can be accomplished by constructing gene X in a context to express an antisense RNA containing a ribozyme sequence. When induced by tetracycline, antisense-ribozyme RNA binds to the mRNA expressed by gene X, cleaves it, and thereby functionally inactivates the gene. PROSPECTS FOR THE FUTURE FOR CONDITIONAL TRANSGENE EXPRESSION The availability of the Cre-loxP and the forward and reverse tTA systems now makes it feasible to combine their key features in the creation of triple transgenic mice so that a targeted recombinational disruption of a gene can be accomplished by the administration of tetracycline. The Cre recombinase could be placed under the control of a tissue-specific promoter containing the tet operon uniquely responsive to the presence of tTA and targeted to a specific tissue by standard pronuclear injection targeted transgenesis. A second transgenic mouse is created with a ubiquitously expressed promoter during the expression of the reverse tTA. In the third mouse, the gene desired to be deleted would be replaced with an appropriately floxed gene. The latter mouse would be prepared by implantation of recombinantly engineered ES cells into blastocysts. The administration of tetracycline to the triple transgenic mouse would induce the Cre recombinase in a tissue-specific manner, thus allowing temporal and spatial control of gene knock-outs. EXPRESSED SEQUENCE TAGS A very informative database of expressed sequence tags (ESTs) is being generated and placed in GenBank. Expressed sequence tags are prepared by random, single-pass sequencing of mRNAs from a repertoire of different tissues, mostly embryonic tissues (e.g., brain, eye, placenta, liver). Currently, the EST database contains ~50% of the estimated expressed genes in humans and mammals (70,000–80,000). The EST database will become extremely valuable when the sequences of the human, rat, and mouse genomes are completed.

DNA ARRAYS FOR THE PROFILING OF GENE EXPRESSION Two variants of DNA-array chip design exist.20,21 The first consists of cDNA (sequences unknown) immobilized to a solid surface such as glass and exposed to a set of labeled probes of known sequences, either separately or in a mixture of the probes. The second is an array of oligonucleotide probes (sequences known, based on either known genes in GenBank or ESTs) that are synthesized either in situ or by conventional synthesis followed by on-chip immobilization (Fig. 2-6). The array is exposed to labeled sample DNA (unknown sequence) and hybridized, and complementary sequences are determined.

FIGURE 2-6. Sample preparation and hybridization for oligonucleotide assay. A complementary DNA (cDNA) is transcribed in vitro to RNA, and then reverse-transcribed to cRNA. This material is fragmented and tagged with a fluorescent tag molecule (F). The fragments are hybridized to an array of oligonucleotides representing portions of DNA sequences of interest. After washing, hybridization of the cRNA probe is detected by localization of the fluorescent signals. (PCR, polymerase chain reaction.)

In cDNA chips, immobilized targets of single-stranded cDNAs prepared from a specific tissue are hybridized to single-stranded DNA fluorescent probes produced from total mRNAs to evaluate the expression levels of target genes. OLIGONUCLEOTIDE ARRAYS (GENOMIC AND EXPRESSED SEQUENCE TAGS) The oligonucleotide gene chip (1.28 × 1.28 cm2) consists of a solid-phase template (glass wafer) to which high-density arrays of oligonucleotides (distance between oligonucleotides of 100 Å) are attached, with each probe in a predefined position in the array. Each gene EST is represented by 20 pairs of 25 base oligonucleotides from different parts of the gene (5' end, middle, and 3' end). The specificity of the detection method is controlled by the presence of single-base mismatch probes. Pairs of perfect and single-base mismatch probes corresponding to each target gene are synthesized on adjacent areas on the arrays. This is done to identify and subtract nonspecific background signals. The gene chip is sensitive enough to detect one to five transcripts per cell and is much more sensitive than the Northern blot technique. COMPLEMENTARY DNA ARRAYS (SPECIFIC TISSUES) Poly (A) mRNA is isolated from cells or tissue of interest, and synthesis of double-stranded cDNA is accomplished by reverse transcription of cDNA, followed by synthesis of double-stranded cDNA using DNA polymerase I. In vitro transcription of double-stranded cDNA to cRNA is accomplished using biotin-16-UTP and biotin-11-CTP for labeling and a T7 RNA polymerase as enzyme. This cRNA is used for hybridization with the gene chip. The gene chip is stained with R-physoerythrin streptavidin to detect biotin-labeled nucleotides, and different wash cycles are performed. Thereafter the gene chip is scanned digitally and analyzed by special software. (A grid is automatically placed over the scanned image of the probe array chip to demarcate the probe cells.) After grid alignment, the average intensity of each probe cell is calculated by the software, which then analyzes the patterns and generates a report. The applications of the gene chip include: 1. 2. 3. 4.

Simultaneous analysis of temporal changes in gene expression of all known genes and ESTs. Sequencing of DNA. Large-scale detection of mutations and polymorphisms in specific genes (i.e., BRCA1, HIV-1, cystic fibrosis CFTR, b-thalassemia). Gene mapping by determining the order of overlapping clones.

Expensive equipment for generating and analyzing the data using genechips is required. When the cloning of all genes is completed (Human Genome Project), the gene chip will allow monitoring of the expression of all known genes in various situations.

STRATEGIES FOR MAPPING GENES ON CHROMOSOMES GENETIC LINKAGE MAPS AND QUANTITATIVE TRAIT LOCI A genetic linkage map shows the relative locations of specific DNA markers along the chromosome.22,23,24,25,26 and 27 Any inherited physical or molecular characteristic that differs among individuals and is easily detectable in the laboratory is a potential genetic marker. Markers can be expressed DNA regions or DNA segments that have no known coding function, but whose inheritance pattern can be followed. DNA sequence differences (polymorphisms; i.e., nucleotide differences) are especially useful markers because they are plentiful and easy to characterize precisely. Markers must be polymorphic to be useful in mapping. Alternative DNA polymorphisms exist among individuals, even among members of a single family, so that they are detectable among different families. Polymorphisms are variations in DNA sequence in the genome that occur every 300 to 500 base pairs. Variations within protein-encoding exon sequences can lead to observable phenotypic changes (e.g., differences in eye color, blood type, and disease susceptibility). Most variations occur within introns and have little or no effect on the phenotype (unless they alter exonic splicing patterns), yet these polymorphisms in DNA sequence are detectable and can be used as markers. Examples of these types of markers are: (a) restriction fragment

length polymorphisms (RFLPs), which reflect sequence variations in DNA sites that are cleaved by specific DNA restriction enzymes; and (b) variable number of tandem repeat sequences (VNTRs), which are short repeated sequences that vary in the number of repeated units and, therefore, in length. The human genetic linkage map is constructed by observing how frequently any two polymorphic markers are inherited together. Two genetic markers that are in close proximity tend to be passed together from mother to child. During gametogenesis, homologous recombination events take place in the metaphase of the first meiotic step (meiotic recombination crossing-over). This may result in the separation of two markers that originally resided on the same chromosome. The closer the markers are to each other, the more tightly linked they are and the less likely that a recombination event will fall between and separate them. Recombination frequency provides an estimate of the distance between two markers. On the genetic map, distances between markers are measured in terms of centimorgans (cM), named after the American geneticist Thomas Hunt Morgan. Two markers are said to be 1 cM apart if they are separated by recombination 1% of the time. A genetic distance of 1 cM is roughly equal to a physical distance of 1 million base pairs of DNA (1 Mb). The current resolution of most human genetic map regions is approximately 10 Mb. An inherited disease can be located on the map by following the inheritance of a DNA marker present in affected individuals but absent in unaffected individuals, although the molecular basis of a disease or a trait may be unknown. Linkage studies have been used to identify the exact chromosomal location of several important genes associated with diseases, including cystic fibrosis, sickle cell disease, Tay-Sachs disease, fragile X syndrome, and myotonic dystrophy. RESTRICTION ENZYMES AND CHROMOSOMAL MAPPING The restriction endonucleases, which have been isolated from various bacteria, recognize short DNA sequences and cut DNA molecules at those specific sites. A natural biofunction of restriction endonucleases is to protect bacteria from viral infection or foreign DNA by destroying the alien DNA. Some restriction enzymes cut DNA very infrequently, generating a small number of very large fragments, whereas other restriction enzymes cut DNA more frequently, yielding many smaller fragments. Because hundreds of different restriction enzymes have been characterized, DNA can be cut into many differentsized fragments. PHYSICAL MAPS Different types of physical maps vary in their degree of resolution. The lowest resolution physical map is the chromosomal (cytogenetic) map, which is based on the distinctive banding pattern observed by light microscopy of stained chromosomes. A cDNA map shows the locations of expressed DNA (exons) on the chromosomal map. The more detailed cosmid contiguous DNA block (contig) map depicts the order of overlapping DNA fragments spanning the genome (see the section High-Resolution Physical Mapping). A macrorestriction map describes the order and distance between restriction enzyme cleavage sites. The highest resolution physical map will be the complete elucidation of the DNA base-pair sequence of each chromosome in the human genome. LOW-RESOLUTION PHYSICAL MAPPING Chromosomal Map. In a chromosomal map, genes or other identifiable DNA fragments are assigned to their respective chromosomes, with distances measured in base pairs. These markers can be physically associated with particular bands (identified by cytogenetic staining) primarily by in situ hybridization, a technique that involves tagging the DNA marker with an observable label. The location of the labeled probe can be detected after it binds to its complementary DNA strand in an intact chromosome. As with genetic linkage mapping, chromosomal mapping can be used to locate genetic markers defined by traits observable only in whole organisms. Because chromosomal maps are based on estimates of physical distance, they are considered to be physical maps. The number of base pairs within a band can only be estimated. Fluorescence In Situ Hybridization.28,29 A fluorescently labeled DNA probe locates a DNA sequence detected on a specific chromosome. The fluorescence in situ hybridization (FISH) method allows for the orientation of DNA sequences that lie as close as 2 to 5 Mb. Modifications to the in situ hybridization methods, using chromosomes at a stage in cell division (interphase) when they are less compact, increase map resolution by an additional 100,000 base pairs. A cDNA map shows the positions of expressed DNA regions (exons) relative to particular chromosomal regions or bands. (Expressed DNA regions are those transcribed into mRNA.) The cDNA is synthesized in the laboratory using the mRNA molecule as the template. This cDNA can be used to map the genomic region of the respective molecule. A cDNA map can provide the chromosomal location for genes whose functions are currently unknown (ESTs). For hunters of disease genes, the map can also suggest a set of candidate genes to test when the approximate location of a disease gene has been mapped by genetic linkage analysis. HIGH-RESOLUTION PHYSICAL MAPPING Two current approaches to high-resolution mapping are termed top-down (producing a macrorestriction map) and bottom-up (resulting in a contig map). With either strategy, the maps represent ordered sets of DNA fragments that are generated by cutting genomic DNA with restriction enzymes. The fragments are then amplified by cloning or by PCR methods. Electrophoretic techniques are used to separate the fragments (according to size) into different bands, which are visualized by staining or by hybridization with DNA probes of interest. The use of purified chromosomes, separated either by fluorescence-activated flow sorting from human cell lines or in hybrid cell lines, allows a single chromosome to be mapped. A number of strategies can be used to reconstruct the original order of the DNA fragments in the genome. Many approaches make use of the ability of single strands of DNA and/or RNA to hybridize to form double-stranded segments. The extent of sequence homology between the two strands can be inferred from the length of the double-stranded segment. Fingerprinting uses restriction enzyme cleavage map data to determine which fragments have a specific sequence (finger-print) in common and, therefore, overlap. Another approach uses linking clones as probes for hybridization to chromosomal DNA cut with the same restriction enzyme. In top-down mapping, a single chromosome is cut (using rare-cutter restriction enzymes) into large pieces, which are ordered and subdivided; the smaller pieces are then mapped further. The resulting macrorestriction maps depict the order of and distance between locations at which rare-cutter restriction sites are found in the chromosome. This approach yields maps with more continuity and fewer gaps between fragments than contig maps, but map resolution is lower and the map may not be useful in finding particular genes. In addition, this strategy generally does not produce long stretches of mapped sites. Currently, this approach allows DNA pieces to be located in regions measuring ~100 kb to 1 Mb. The development of pulsed-field gel (PFG) electrophoretic methods has improved the mapping and cloning of large DNA molecules. Whereas conventional gel electrophoretic methods separate pieces of DNA 700-kDa) complex with at least eight proteins. One of these, the TATA-binding protein (TBP), allows the binding of TFIID to the TATA box or related Inr (initiator) sequences. The other components of TFIID have also been partially characterized and are known as coactivators or TBP-associated proteins (TAFs). They are essential for the communication of enhancer-binding protein signals to the basal transcriptional machinery and the subsequent regulation of gene expression. Basal transcription depends on the formation of a preinitiation

complex involving TFIID-TFIIA-TFIIB, followed by the rapid entry of RNA polymerase II to facilitate the establishment of the transcriptional machinery.14,15 and 16 The second DNA element is the upstream promoter element (UPE), which is located 60 to 110 nucleotides upstream from the cap site.17 It includes elements such as the CCAAT and GC-rich (GGGCGG) boxes that bind to CAAT-binding proteins and SP1 (a cellular DNA-binding protein that interacts with the SV40 genome), respectively. These elements also associate with DNA-binding proteins that augment the efficiency of transcription by RNA polymerase II. These UPEs may or may not require specific TATA boxes to perform their function most efficiently. Together, the TATA box and UPEs are components close to the structural region and are essential for maintenance of basal levels of gene transcription (Fig. 3-7).

FIGURE 3-7. Regulation of transcriptional rates by interactions of transacting factors. Various permutations of interactions of nuclear binding proteins with various DNA elements within regulatory regions determine rates of transcription at the basal and regulated levels. Such proteins include the TATA box, the upstream promoter element (UPE), and enhancer or hormone regulatory element (HRE)–binding proteins.

Enhancers are located in variable positions and may act independently of orientation. They may be located more distal than the promoter elements and are found up to several thousand nucleotides upstream or downstream of the transcriptional unit. These elements also bind proteins that enhance transcriptional rates or diminish them (i.e., silencers) in an ill-defined manner and constitute the foci of regulated transcription (see Fig. 3-7). Several proposed mechanisms include the cooperative interaction of a number of DNA-binding proteins to effect efficient formation of the transcription-initiation complex of RNA polymerase II with the regulatory or promoter region.18 Another hypothesis suggests that the interaction of proteins with these elements opens up the configuration of DNA, perhaps by “bending” to allow access of the gene to the transcription machinery. With recombinant DNA techniques, a reporter gene construct can be produced that may be transfected into foreign cells by gene transfer.19,20 This allows the expression of the reporter gene with enhanced production of an enzyme or polypeptide product that is not normally produced in eukary-otic cells. The synthesis of such products may be detected by sensitive enzyme assays or radioimmunoassays. DNA constructs in which a structural region corresponding to the enzyme alone is transfected into cells are not expressed in the absence of regulatory regions. However, if a promoter element is placed 5' to the reporter gene, then expression may occur. Using such approaches, structural analysis of various portions of the 5'-regulatory regions of genes, including enhancer and upstream regulatory elements, may be performed. After the RNA transcript is initiated, the RNA polymerase II continues the process of template reading by elongation of the transcript until termination occurs. The actual site of transcription termination is variable, located 50 to 200 or more nucleotides downstream from the 3' end of the last exon or polyadenylation site.21 Although potential weak consensus sequences have been discerned that may determine the site at which the initial RNA transcript is terminated, the polyadenylation site appears to be obtained only by virtue of endonucle-olytic cleavage of longer heterogeneous 3' ends of the hnRNA. The well-conserved consensus polyadenylation site sequence AAUAAA, which is located 15 to 20 nucleotides upstream from the polyadenylation site, and other proximal downstream sequences 10 to 12 nucleotides from the polyadenylation site, may serve as points of recognition for this processing event. Although the mechanisms of polyadenylation are not well known, the presence of the consensus sequences suggests a requirement for stem-loop formation and involvement of small nuclear ribonucleoproteins (snRNPs).

MESSENGER RNA PROCESSING The hnRNA product of gene transcription is rapidly processed in the nucleus with a half-time (t1/2) of 5 to 20 minutes (Fig.3-8).22 Three major events transform the large heterogeneous RNA precursors into the mature RNA. First, at the 5' end of hnRNA, a 7methylguanosine residue is added to the first nucleotide of the transcript by means of a 5'-5' triphosphate bond after 20 to 30 nucleotides have been polymerized. This reaction is rapid (t1/2 < 1 minute) and is catalyzed by the 5'-capping enzyme, including guanylyl and methyl transferases. The 5'-methyl cap associates with 5'-cap binding proteins, which favors the formation of a stable 40S translation-initiation complex and increases the stability and efficiency of translation of the eventual mature mRNA.23

FIGURE 3-8. Gene transcription and RNA processing. The initial RNA transcript is known as heterogeneous nuclear RNA (hnRNA I). It contains exons and introns of the structural region and rapidly undergoes 5' capping with 7methylguanosine (7meG) and 3' polyadenylation (An) (hnRNA II). Little heterogeneous nuclear RNA has been detected without 5' cap or 3' polyadenylated (poly[A]) tails. In a slower process, introns are removed by RNA splicing followed by religation of exon sequences. The mature messenger RNA (mRNA) is composed of fused exon sequences and contains a 5' cap and a 3' poly(A) tail.

The second modification occurs at the 3' end and involves the addition of a polyadenylate, or poly(A), tail. Polyadenylation includes the addition of 250 to 300 adenylate (A) residues at the polyadenylation site located at the 3' end of the RNA. This poly(A) tail, which is reduced to 30 to 250 residues during nuclear processing and export, may also be important for increased RNA stability. These two additions, capping and polyadenylation, occur within minutes after the synthesis of hnRNA and generally before RNA splicing; almost all isolated hnRNA contains both modifications. The third major processing step involved in mRNA maturation is the removal of introns during RNA splicing.24,25 and 26 This process includes endonucleolytic cleavage of introns and religation of exons. The 5' and 3' ends of introns have consensus sequences, as shown in Figure 3-9.

FIGURE 3-9. RNA splicing: consensus intron sequences and mechanisms for intron removal. A consensus sequence has been determined for the 5' and 3' ends of intron sequences. Data suggest the potential mechanism of intron removal by means of lariat formation preceded by interaction with nuclear RNAs. (nt, nucleotides.)

These consensus sequences may be necessary for the appropriate interaction of U1 snRNP species present in the nucleus to serve as a “splicing adapter” for the splicing process. Moreover, a polypyrimidine tract is located adjacent to the 3' AG residues and a critical adenylate residue in a branch sequence, 30 nucleotides upstream of the 3' end of the intron. The first step in the splicing process involves the formation of the spliceosome, which includes the hnRNA, U1 snRNP, and other factors. The initial event is endonucleolytic cleavage at the 5' splice site, followed by the formation of a 5'-2' phosphodiester bond between the 5' G and the downstream A located in a branch sequence. This “lariat” intermediate is then cleaved at the 3' end and degraded, and the exons are ligated. The removal of introns from hnRNA must be precise; errors can change the exon or mRNA-coding regions. The sequence of removal of multiple introns within a gene is generally nonrandom, although the mechanism is unknown. Variations in the splicing pattern in a given hnRNA transcript can occur, and tissue-specific interactions of RNA splicing-modification proteins may dictate alternate patterns of intron-RNA splicing, causing altered mRNA forms.27 In a complex transcriptional unit, an alternate exon choice, including alternative internal acceptor and donor site use, may yield different mRNA products. A complex transcriptional unit may also possess alternate transcriptional start sites in the same contiguous segment of DNA (i.e., in the same exon) or in multiple transcriptional start sites in different exons contributed by alternate exon choice. Another possible mechanism for diversity in the complex transcriptional unit is alternative final exon choice (i.e., differences in polyadenylation sites). The splicing process is another rate-limiting step and takes place over 5 to 30 minutes; it is much slower than the capping and polyadenylation reactions. What role RNA splicing plays in the informational flow is unclear. However, the potential contribution of RNA diversity by RNA splicing has been discussed. There are mRNAs that lack poly(A) tails (e.g., histone mRNAs), mRNAs that lack a 5' cap (e.g., poliovirus mRNAs), and eukaryotic genes that lack introns. Such modifications are not essential for RNA maturation.

RNA TRANSPORT The newly synthesized mature mRNA is actively transferred from the nuclear to the cytoplasmic compartments by way of the nuclear pore complex (NPC). The NPC is a large multiple-component structure that is located in the nuclear envelope and serves as a channel for the movement of macromolecules such as RNAs. The mRNA and other RNAs subject to transport are closely associated with proteins and exist as ribonucleoproteins. Each RNA likely possesses distinct protein-targeting sequences that permit its export and import. This shuttling of mRNA from the nucleus to the cytoplasm is mediated by a large family of transport factors known collectively as exportins and importins. However, the precise nature of the interactions of these shuttling proteins, mRNAs, and the NPC is not well understood.28

TRANSLATION The structure of mRNA is shown in Figure 3-10. The exons encode two major regions of the mRNA: translated and untranslated. The translated or coding region contains the open reading frame, beginning from the initiation methionine codon to the termination codon. The untranslated regions flank the coding region and are known as 5' or 3' untranslated regions. The functions of the untranslated regions are not well established, but data indicate that the 5' untranslated region may be important in determining the efficiency of translation of the mRNA.29,30 The 3' untranslated region may contain important RNA elements, especially several AU-rich sequences that determine the stability of mRNA in the cytoplasm.31 Each of these regions may mediate its effects by binding to specific RNA-binding proteins.32,33

FIGURE 3-10. Structure and translation of messenger RNA (mRNA). In most cases, the mature mRNA represents the fusion of multiple exons. These sequences encode two major regions: translated and untranslated. The translated or coding region is delimited by the translation initiator codon, AUG, at its 5' end and the termination codon, UGA, UAA, UAG, at its 3' end. This coding region represents a series of codons in an open-reading frame that determines the amino-acid sequence of its encoded polypeptide. The 5' and 3' untranslated regions are shown. The mRNA enters the cytoplasm to interact with the ribosome. There, protein synthesis is initiated, and by way of a series of several cotranslational events, secretory polypeptide hormone precursors are processed. The steps involve cleavage of the signal or leader peptide, followed by addition of asparagine-linked carbohydrate moieties in glycoprotein subunits or hormones, and intramolecular folding with the formation of disulfide linkages. These events occur within the lumen of the rough endoplasmic reticulum. These partially processed polypeptide hormones are then shuttled to the Golgi stack, where these molecules are transported, sorted, and further processed posttranslationally to yield the bioactive hormone located in secretory granules or vesicles.

The mRNA in the cytoplasm rapidly interacts with the ribosome (see Fig. 3-10). The ribosome is a complex ribonuclear particle that contains 28S, 18S, and 5S RNAs, along with a group of ribosomal proteins. Among these proteins are factors responsible for the initiation, elongation, and termination of mRNA translation. For the typical mRNA, 3 to 15 ribosomes may be attached at any given time. As the ribosome reads the mRNA in the process of translation, amino acids are brought to the translation complex by way of adapter tRNA molecules. These molecules are differentiated by the presence of anticodon structures (i.e., RNA sequences complementary to a particular codon) at one end and attachment sites for specific amino-acid residues at the other end of the L-shaped molecule. The reading of successive codons causes the alignment of the appropriate amino acids and polymerization to yield the polypeptide chain. Translation initiation occurs at the initiator codon or AUG, which represents the amino residue methionine. Translation generally begins at the first AUG codon located at the 5' end of the mRNA. This initiation methionine codon is normally followed by an open reading frame of codons encoding amino acids until a termination codon is reached. When a UAG, UAA, or UGA is encountered, protein synthesis stops, and the nascent polypeptide chain is released from the ribosome complex. The context of the methionine codon that is used for translation initiation has been characterized further to include a consensus sequence: 5'-CCACCAUGG-3'. This sequence nest presents the AUG as the most favorable initiation codon.34 However, examples have been found in which the AUG is located 5' of the authentic start site. In these instances, the context may not be ideal or may be quickly followed by a termination codon in frame. Whether peptides encoded by these short-reading frames are eventually expressed is unknown.35,36 All polypeptide hormones and almost all other proteins destined for membrane, lysosome, ER, and Golgi stack locations or for secretion are encoded by a larger

polypeptide precursor. All polypeptide hormones possess a signal or leader peptide that is a characteristic segment of protein located at the N-terminal end37,38 (Table 3-1). Although no consensus primary sequence has been obtained for this signal peptide, it generally possesses a hydrophobic core preceded by basic amino-acid residues in its 16- to 30-amino-acid residue extent.

TABLE 3-1. Polypeptide Hormones: Some of Their Precursor Proteins*

Several events occur before the entire polypeptide chain is synthesized (Fig. 3-11). After the synthesis of ~70 amino acids, the signal recognition particle (SRP), a group of six proteins and a small RNA (7S), interacts with the signal peptide to momentarily halt translation elongation in the RNA–ribosome-nascent protein complex.39,40 The 7S RNA contains a signal peptide recognition and an elongation arrest domain. This complex then interacts with the SRP receptor, an integral membrane protein located on the cytoplasmic face of the ER. In this process, poly-ribosomes are attached to membranous structures associated with the endoplasmic reticulum to form the rough ER (RER). After this interaction occurs, translational arrest is relieved, and translation proceeds as usual. At this point, the signal peptide is vectorially transported through the membrane into the cisternal aspect of the ER. The newly synthesized protein has been translocated from the inside to the outside of the cell in a topologic sense.

FIGURE 3-11. Details of translational and cotranslational processes. The messenger RNA (mRNA) interacts with the ribosome where protein synthesis is initiated. In the case of polypeptide hormones, the first segment of protein synthesized is the N-terminal signal or leader peptide. As soon as the signal peptide emerges from the ribosomal complex, a protein-RNA particle known as the signal recognition particle (SRP) associates with the signal peptide. This interaction allows the ribosomal-mRNA–nascent polypeptide complex to interact with the SRP receptor located on the cytoplasmic face of the endoplasmic reticulum (ER) membrane and brings the ribosome in close apposition to the ER to form the rough ER. The momentary translational arrest that occurs on interaction of the complex with SRP is released to allow further protein synthesis. Cleavage of the signal peptide from the apoprotein by signal peptidase and other modifications, including addition of asparagine-linked carbohydrates (CHO), intramolecular folding, and disulfide linkage formation, occurs coincidentally with release of ribosomes from the ER. In this manner, the partially processed protein, although initially synthesized in the cytoplasmic space, enters the luminal space.

As protein synthesis continues, the signal peptide is transiently immobilized in the membrane by virtue of its hydrophobic nature or its binding to a putative signal peptide receptor.41,42 Although the nascent protein chain is transferred to the cisterna by way of an unknown, energy-dependent translocation process, a luminal surface enzyme, signal peptidase, rapidly performs proteolytic cleavage to remove the signal peptide. The transmembrane transport of the protein does not require signal peptide cleavage and may take place by way of a protein channel or, less likely, through lipid. If the protein is to be N-glycosylated (i.e., to contain asparagine-linked carbohydrate moieties), other enzymes and the dolichol-lipid oligosaccharide carrier provide core glycosylation in this cotranslational process. Moreover, protein folding and oxidation of cysteine residues in disulfide formation occur. At the completion of protein synthesis and complete transfer of the protein to the luminal space, the SRP complex dissociates from its receptor and is recycled into the cytoplasm. Polyribosomes also are disaggregated to form free ribosomes and RNA. The translation of the polypeptide hormone causes the synthesis of a polypeptide core derived from the initial protein precursor, which is already modified, in some instances, by the addition of carbohydrate moieties and by folding and formation of intramolecular disulfide linkages. The precursor polypeptide encoded by mRNA is not found in vivo, because the signal peptide is removed before the completion of the polypeptide chain. The exit from the ER probably depends on appropriate protein assembly or conformation, but glycosylation is not required.

POSTTRANSLATION Up to the translational and cotranslational steps in the ER, all secretory, membrane, lysosome, endogenous ER, and Golgi proteins have traversed the same biosynthetic path. After this point, the major task of sorting and transferring the proteins to the correct intracellular destinations must be completed. This complex process occurs in the Golgi stack and requires sorting signals among the proteins and sorting mechanisms in this organelle. A polypeptide hormone destined for regulated secretion must exit the ER, traverse the Golgi stack, and arrive properly in the secretory granule (Fig. 3-12).

FIGURE 3-12. The polypeptide hormone highway. Protein hormone synthesis is initiated in the cytoplasm on polyribosomes. The partially processed hormone, with the signal peptide removed and N-linked carbohydrate moieties attached and with appropriate folding, enters the lumen of the rough endoplasmic reticulum (RER). By way of transport vesicle–transitional elements, these partially processed products are transferred to the Golgi stack on fusion and release. In a serial process of budding formation of secretory vesicles and fusion, processed products are transferred through the Golgi stack, from which they exit as secretory vesicles or granules after sorting in the trans and trans-Golgi network compartments of the Golgi. Materials are then released from granules by the fusion of vesicles or granules with the plasma membrane.

The Golgi stack comprises a series of flattened, saccular membranous compartments that encompass four histologically and functionally distinct regions: the cis, medial, and trans regions of the Golgi complex and the trans-Golgi network (TGN)43,44 (Fig. 3-13). The cis-Golgi region is most proximal to the transitional elements of the RER, and the TGN is most distal. The maintenance of distinct Golgi-specific antigens, unique enzyme markers, and different lectin-binding characteristics suggest that the compartments are not contiguous.

FIGURE 3-13. The Golgi stack. The Golgi stack consists of numerous membranous compartments, including cis, medial, and trans-Golgi elements. These compartments may be differentiated by the presence of specific enzymes. Partially processed protein hormones traverse this system by way of intermediate secretory vesicles in a budding-fusion reiterative process. In addition to transport, protein processing occurs. Sorting with routing to ultimate destinations in cellular sites is accomplished in the trans-Golgi network (TGN). Secretory peptides may be sorted to constitutive or regulated secretory pathways. Constitutive secretory pathways are equivalent to the pathways taken by membrane proteins, whereby non–clathrin-coated membrane segments are used. The regulated secretory-secretory granule pathway involves a clathrin-coated pit among membrane segments. This is similar to the pathway taken by lysosomal components. (Adapted from Griffiths G, Simons K. The trans Golgi network: sorting at the exit site of the Golgi complex. Science 1986; 234:438.)

A vesicle transfer model has been proposed to account for transport of materials from the RER to the TGN. In this model, membrane vesicles form from the upstream compartment by budding at the rims of the Golgi plates and rejoin the adjacent downstream compartment by vesicle fusion and the interaction of microfilaments. The reiterative process of budding and fusion of secretory or transport vesicles causes vectorial transfer of proteins from the RER to the TGN in a unidirectional and energy-dependent process. The newly synthesized polypeptide in the lumen of the RER is first translocated to the cis-Golgi region (see Fig. 3-12). From this point, the protein is transported and processed in the Golgi stack. This organelle may be appropriately considered an assembly line for posttranslational processing. It is here that N-linked carbohydrate cores are further modified among glycoproteins45 (Fig. 3-14). This process involves digestion of the high-mannose peripheral sugars in the N-linked carbohydrate cores by multiple glycosidases and subsequent addition of distal or terminal sugars by way of numerous glucosyltrans-ferases. The steps in this process of carbohydrate maturation occur in different Golgi compartments. Other processes also occur, including phosphorylation, acetylation, sulfation, acylation, -amidation of COOH termini, addition of ubiquitin, other modifications, and degradation.46

FIGURE 3-14. Proximal and distal glycosylation. The pathway of glycosylation in the rough endoplasmic reticulum (RER) and Golgi is shown. Core carbohydrate moieties are added cotranslationally by way of a dolichol-sugar intermediate (Dol-) to Asn residues in the protein backbone in the RER. Several glycosidases (steps 1–4) remove distal sugars in this compartment. Distal glycosylation occurs by the actions of mannosidases (steps 5–7) and glycosyl transferases (steps 6, 8–1) in the Golgi. Phosphorylation (I, II) of N-acetyl glucosamines in carbohydrate moieties in the cis Golgi occurs in proteins destined for lysosome localization. (From Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 1985; 54:631.)

Another important function of the Golgi stack is the delivery of nascent polypeptides to the appropriate targets within the cell, which occurs in the trans-Golgi region or TGN.47 The proteins destined for lysosomal sites are targeted to those organelles by way of the mannose-6-phosphate receptor.48 In a similar manner, receptor and secretory proteins are targeted to membrane and secretory granule sites, respectively.49,50,51,52 and 53 The nearly mature polypeptide emerges from the Golgi stack in the TGN, where transport organelles, known as secretory vesicles or granules, are formed. These vesicles allow the exit of the nearly mature protein hormone from the Golgi stack. Secretory proteins are released from a cell by way of two pathways: the constitutive pathway and the regulated pathway.54,55 The constitutive pathway is thought to be mediated by a passive aggregation sorting mechanism whereby peptide hormones form aggregates in the TGN, an action that is facilitated by acidic pH and high calcium concentrations in this compartment. The polarity of the secretory faces of epithelial cells enables proteins that are released in a nonregulated or constitutive manner to be released on the apical surface and regulated release to be performed at the basolateral surface. Whether such polarity of secretion exists in endocrine cells is unknown. Constitutive release generally involves the rapid exocytosis of newly synthesized peptides, but regulated secretion involves the classic secretory granule and signaled degranulation, causing hormone- or factor-regulated release of hormones. Secretory peptides must be segregated into one pathway or the other. Regulated secretion involves the formation of secretory residues and granules composed of clathrin-containing membrane segments, as found in lysosomes. Proteins destined for regulated secretion must end up in a reservoir known as the secretory granule, where the polypeptide hormones are concentrated and stored. This pathway is now considered to operate by active sorting via a signal ligand receptor. Proteins destined for secretion in this manner clearly contain sorting signals in their precursor molecules. For instance, the precursors to proopiomelanocortin (POMC) and proenkephalin have a stretch of aliphatic hydrophobic and acidic amino-acid residues at the N termini that are necessary and sufficient for efficient sorting into secretory granules. Further, carboxypeptidase E (Cpe) appears to serve as a sorting receptor for these peptide signals as determined by biochemical and genetic approaches. In particular, the Cpefat, which harbors a mutant and ineffective Cpe, is obese, diabetic, and infertile. It has elevated levels of proinsulin in pancreatic B cells and of POMC in the anterior pituitary, and decreased insulin and ACTH release.56 Three types of vesicles are formed in the TGN. One is the secretory vesicle, which is not clathrin coated and mediates non–receptor-dependent transport of membrane proteins and protein to be secreted in the constitutive pathway. The other two are the secretory granule, which is partially clathrin coated and mediates the receptor-dependent transfer of regulated secretory peptides, and the lysosome, which is predominantly clathrin coated and mediates transport of lysosomal enzymes and proteins.57 The secretory vesicle participates in the default, bulk-flow sorting system, but the others require the presence of “sorting patches” or sorting signals based on secondary and tertiary, but not primary, structures.55 Although the secretory granules are derived from immature granules with clathrin-coated pits, the precise nature of the receptor-mediated sorting of peptide hormones is unknown. Evidence exists for pH-regulated, receptor-dependent sorting in the trans-Golgi and TGN. The pH of the compartments decreases as the Golgi stack is traversed from cis to trans regions. Such gradients in pH may participate in the molecular aggregation of polypeptide hormones. Possibly, these aggregates formed in the process of hormone concentration may initiate the budding of secretory granules. Chloroquine, which prevents Golgi acidification, may inhibit granule formation by preventing aggregation in neutralized Golgi stacks. Proteolytic processing of protein precursors (i.e., proproteins or polyproteins) to yield smaller bioactive peptides (see Table 3-1) also occurs in acidic Golgi and

secretory vesicles.58 Such proteolysis, however, is not required for packaging.

SECRETORY GRANULE Much has been learned about the nature of polypeptide hormones and secretory granules.55,59,60 The hormones in this organelle are highly concentrated. In particular, a number of polypeptide hormones are condensed in a crystal lattice formation to increase the amount of hormone (up to 200-fold) in this organelle. Secretory granules allow cells to store enough hormone to be released on demand by extracellular signals at a level not possible by de novo synthesis. The t1/2 of stored hormones may be days, whereas the t1/2 of similar proteins in secretory vesicles may be minutes. The size of secretory granules varies greatly, depending on the nature of the stored hormone. The condensation of hormone is demonstrated by the presence of electron-opaque or “dense” cores. The granule core is quite stable and is often visible even after exocytosis or in vitro enzymatic digestion of the granule membrane. It is osmotically inert yet sensitive to pH levels higher than 7.0.55 The formation of the secretory granule proceeds in stages, beginning in the trans-Golgi, where the initial hormone concentration may be observed. This aggregation process60a is facilitated by changes in pH, calcium concentration, and possible presence of other proteins such as secretogranins, chromogranins, and sulfated proteoglycans. Aggregates may form in different regions of the secretory granule. The colocalization of two or more polypeptide hormones in a granule may be observed. Within a cell, the relative distribution of two hormones is constant from granule to granule; however, variability in overall distribution is achieved from cell to cell. The mechanism by which the gonadotrope, a cell that generally produces luteinizing hormone (LH) and follicle-stimulating hormone (FSH), may be regulated to release LH and FSH differentially remains unclear.61

SECRETION Secretory granules release their contents by cytoskeletal protein-mediated movement of the granule toward the cellular surface.61a There, secretory granule membranes fuse with the plasma membrane and allow eversion or exocytosis of stored hormone.62 This process of emiocytosis causes secretion of hormone. The mechanisms involved in stimulus-secretion coupling are not well known, although responses to cellular signals causing changes in intracellular calcium, ion currents, or intra-cellular pH may lead to these events. In the unstimulated cell, a web of actin-associated microfilaments on the cytoplasmic face of the plasma membrane may act as a physical barrier to secretory granule fusion. However, changes in intracellular calcium, ion currents, or intracellular pH may cause differences in actin-binding protein interactions and alterations in the “secretory barrier” and permit exocytosis to occur. Secretion and rapid membrane fusion of multiple secretory granules require an endocytotic pathway to retrieve the extra membranes resulting from exocytosis in the plasma membrane and to return them to the Golgi stack and lysosome.

REGULATION OF POLYPEPTIDE HORMONE SYNTHESIS Regulation of the biosynthesis of polypeptides may occur at any of the biosynthetic levels in the pathway (Table 3-2 and Fig. 3-4). Of major interest is the regulation of peptide hormone synthesis at the transcriptional level.

TABLE 3-2. Loci of Genetic Regulation of Polypeptide Hormone Synthesis*

Studies using gene transfer and structure-function analysis have established that specific DNA elements in the regulatory region of the transcriptional unit are critical for determining transcriptional rates of various structural regions.63,64 In particular, hormone-regulatory elements (HREs) have been characterized for glucocorticoid, estrogen, androgen progesterone, vitamin D, mineralocorticoid, retinoic acid, and thyroid hormone receptors. In each case, a DNA element 8 to 20 nucleotides long may be necessary and sufficient for conferring hormonal regulation to its associated structural region. Several factors, including the steroid and thyroid hormones, interact with nuclear receptor proteins, which interact with DNA elements directly to modulate gene transcription.65,66,67 and 68 For the glucocorticoid receptor, the glucocorticoid ligand binds to the inactive glucocorticoid receptor in the cytoplasm, present in a complex with heat shock proteins, hsp 90 and hsp 70, and others. The activated receptor-ligand complex interacts as a transacting factor to bind the DNA element corresponding to the glucocorticoid regulatory element (GRE). Studies have been performed on GREs in genes for mouse mammary tumor virus (MMTV) and murine sarcoma viruses, human metallothionein IIa, tyrosine aminotransferase, tryptophan oxygenase, and growth hormone, and in other genes. The long terminal repeat region of MMTV contains five GREs.64,69 A consensus sequence for the putative GRE is shown by the sequence 5'-GGTA-CANNNTGTTCT-3', inwhich N = A, C, G, or T. The structures of the steroid and thyroid hormone receptors are better known. These hormone receptors are encoded by genes related to a viral oncogene, v-erbA.70,71 The thyroid hormone receptor is encoded by the protooncogene c-erbA. Each receptor contains a stereotypic structure, including a protein that is ~45 to 60 kDa, with a central DNA-binding domain and a carboxyl-terminal ligand-binding domain. These and other regions mediate trans-activation, dimerization, and nuclear localization. The DNA-binding region consists of multiple cysteine and histidine residues that are critical for the formation of Zn2+ fingers first described in the DNA-binding protein TFIIIa, a transcription regulatory factor for the 5S ribosomal gene in Xenopus.72 This Zn2+finger interaction is a common motif for the binding of many eukaryotic proteins to DNA.73,74 and 75 The steroid–thyroid hormone receptors represent the first major examples of trans-acting factors well described in mammalian systems. The motif found in prokaryotic systems, particularly the interactions of cro and lambda repressor proteins with their target DNA elements in bacteriophage lambda, occurs with a homopolymeric dimer of subunits containing alpha helix–turn–alpha helix structure. The binding generally involves protein dimers; it requires a twofold axis of symmetry in the DNA sequence and involves the major groove of the target DNA over several helical turns. Data indicate that the thyroid hormone, retinoic acid, and vitamin D receptors are active only in the heterodimeric state, with other nuclear factors such as retinoid X receptors as their partners. Hormones that act by way of surface membrane receptors may induce the production of second messengers that may directly or indirectly interact with DNA elements within the gene.76,77,78 and 79 HREs may not be restricted to interactions observed with steroid–thyroid hormone receptor complexes, but they may involve other protein-DNA interactions. Advances in the isolation of such trans-acting factors and the identification of cis-acting HREs will probably speed an understanding of the molecular mechanisms of the hormonal regulation of gene expression at the transcriptional level.80 The presence of multiple enhancer elements or HREs in the regulatory regions of genes allows fine tuning of transcriptional efficiency and influences the rate of production of the initial RNA transcript81,82,83,84 and 85 (Fig. 3-15).

FIGURE 3-15. Thyroid hormone action. This diagram depicts the mechanism of action of thyroid hormones in the regulation of a thyroid hormone–responsive gene. Thyroxine (T4) or triiodothyronine (T3) enters the cell. T4 is converted to T3 intracellularly in many cells by means of 5'-deiodinase activity. T3 then enters the nucleus, where it binds to the nuclear thyroid hormone receptor, which is encoded by c-erbA. This hormone nuclear receptor complex then serves as a transacting factor for binding to a thyroid hormone regulatory element (TRE), which may then positively or negatively regulate gene expression, with resultant production of RNA and protein derived from the thyroid hormone–regulated gene. (mRNA, messenger RNA.)

Other loci for regulation in this biosynthetic pathway include elongation and termination of transcription86 (see Fig. 3-9). The various steps of RNA maturation, most notably RNA splicing, may also change mRNA levels encoding a particular polypeptide hormone, which ultimately determines the amount of polypeptide produced. The nuclear stability of the hnRNA and transport of the RNA from the nucleus to the cytoplasm also may be regulated. A major determinant of the steady-state levels of mRNA is cytoplasmic mRNA stability. Examples include the estrogen regulation of chicken liver vitellogenin mRNA, prolactin regulation of breast casein mRNA stabilities, and thyroid hormone control of the TSH b subunit.87,88 and 89 The interaction of mRNA with the protein synthetic machinery in the process of translation may be regulated. Several examples of translational control have been observed, including glucose regulation of the translational efficiency of insulin mRNA. Moreover, a number of the posttranslational processing events that occur in the RER and Golgi stack and the control of secretory granule formation and release may also be loci for regulation. Even after proteins are released from the secretory cell, the bioactive peptide may be further acted on by degradative processes and proteolytic events that may activate proteins in extracellular steps to determine the bioactivity of a particular polypeptide hormone. A major example involves the cascade of the extracellular enzymatic conversion of the precursors of angiotensin II (see Chap. 79). Another example of postsecretion proteolytic processing of precursor polypeptides involves the conversion of iodinated thyroglobulin to the iodinated thyronines, thyroxine and triiodothyronine, in the follicular cell of the thyroid. Plasma stability of a polypeptide is a major determinant of the activity of the hormone in its eventual interaction with target cells.

GENERATION OF DIVERSITY A major example of the generation of diversity is the calcitonin and calcitonin gene–related peptide (CGRP) system. In this system, the C cell of the thyroid expresses a calcitonin-CGRP transcript that initially contains six exons. In the C cell, tissue-specific factors determine the use of the polyadenylation site in the fourth exon, but in the brain, transcription through the sixth exon, which encodes CGRP and the alternative polyadenylation site present in that exon, provides the alternative splicing and deletion of the fourth exon, which encodes calcitonin. The C cells express mostly calcitonin and not much CGRP; conversely, the hypothalamus produces mostly CGRP but not much calcitonin (see Chap. 53 and Fig. 53-1). Other examples of alternative splicing yielding different polypeptides include the synthesis of the alternate human growth hormone form, substance P, substance K, and protooncogenes.90,91 Alternative processing of polypeptides in a posttranslational process is important for the generation of polypeptide diversity.92,93 A major example of this is the production of ACTH and b-lipotropin from the POMC precursor (Fig. 3-16). Using the same mRNA transcript, the anterior pituitary gland produces ACTH and b-lipotropin, and the intermediate lobe of the pituitary gland performs further alternate proteolytic processing and produces b-endorphin, corticotropin-like intermediate lobe peptide (CLIP), a-melanocyte-stimulating hormone, and other products (see Chap. 16).

FIGURE 3-16. Alternative protein processing of the preproopiomelano-cortin (POMC) precursor. In the anterior pituitary gland, the single POMC precursor is processed posttranslationally to produce adreno-corticotropic hormone (ACTH) and b-lipotropin (b-LPH). However, the intermediate lobe further processes these peptides to a-melanocyte-stimulating hormone (a-MSH), corticotropin-like intermediate lobe peptide (CLIP), g-lipotropin (g-LPH), and b-endorphin. (From Douglass J, Civielli O, Herbert E. Polyprotein gene expression: generation of diversity of neuroendocrine peptides. Annu Rev Biochem 1984; 53:665.)

CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Alberts B, Watson JD, Bray D, et al. Molecular biology of the cell. New York: Garland Publishing, 1994. Matsudaira P, Berk A, Zipursky L, et al. Molecular cell biology. New York: WH Freeman Co, 1999. Lewin BM. Genes VII. New York: Oxford University Press, 1999. Newport JW, Douglass JF. The nucleus: structure, function, and dynamics. Annu Rev Biochem 1987; 56:535. Chin WW. Organization and expression of glycoprotein hormone genes. In: Imura H, ed. The pituitary gland. New York: Raven Press, 1985:164. Eipper BA, Mains RE, Glembotski CC. Identification in pituitary tissue of a peptide alpha-amidation activity that acts on glycine-extended peptides and requires molecular oxygen, copper, and ascorbic acid. Proc Natl Acad Sci U S A 1983; 80:5144. Bradbury AF, Finnie MDA, Smyth DG. Mechanism of C-terminal amide formation by pituitary enzymes. Nature 1982; 298:686. Gilbert W. Why genes in pieces? Nature 1978; 271:501. Sudhoff TC, Russell DW, Goldstein JL, et al. Cassette of eight exons shared by genes for LDL receptor and EGF receptor. Science 1985; 228:893. Nevins JR. The pathway of eukaryotic mRNA formation. Annu Rev Biochem 1983; 52:441. Darnell JE Jr. Variety in the level of gene control in eukaryotic cells. Nature 1982; 297:365. Rowley A, Dowell SJ, Diffley FX. Recent developments in the initiation of chromosomal DNA replication: a complex picture emerges. Biochim Biophys Acta 1994; 1217:239. Albright SR, Tjian R. TAFs revisited: more data reveal new twists and confirm old ideas. Gene 2000; 242:1. Zawel L, Reinberg D. Advances in RNA polymerase II transcription. Curr Opin Cell Biol 1992; 4:488. Pugh BF, Tjian R. Diverse transcriptional functions of the multisubunit eukaryotic TFIID complex. J Biol Chem 1992; 267:679. Conaway RC, Conaway JW. General initiation factors for RNA polymerase II. Annu Rev Biochem 1993; 62:161. Busby S, Ebright RH. Promotor structure, promotor recognition, and transcription activation in prokaryocytes. Cell 1994; 79:743. Ptashne M. Gene regulation by proteins acting nearby and at a distance. Nature 1986; 322:697. Walker MD, Edlund T, Boulet AM, Rutter WJ. Cell-specific expression controlled by the 5'-flanking region of insulin and chymotrypsin genes. Nature 1983; 306:557. Edlund T, Walker MD, Barr PJ, Rutter WJ. Cell-specific expression of the rat insulin gene: evidence for role of two distinct 5'-flanking elements. Science 1985; 230:912. Platt T. Transcription termination and the regulation of gene expression. Annu Rev Biochem 1986; 55:339. Brawerman G. Determinants of messenger RNA stability. Cell 1987; 48:5. Shatkin AJ. mRNA cap binding proteins: essential factors for initiating translocation. Cell 1985; 40:223. Padgett RA, Grabowski PJ, Konarska MM, et al. Splicing of messenger RNA precursors. Annu Rev Biochem 1986; 55:1119. Sharp PA. Splicing of messenger RNA precursors. Science 1987; 253:766. Keller W. The RNA lariat: a new ring to the splicing of mRNA precursors. Cell 1984; 39:423.

27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Andreadis A, Gallego ME, Nadal-Ginard B. Generation of protein isoform diversity by alternative splicing: mechanistic and biological implications. Annu Rev Cell Biol 1987; 3:207. Weis K. Importins and exportins: how to get in and out of the nucleus. Trends Biol Sci 1998; 23:185. Pelletier J, Sonenberg N. Insertion mutagenesis to increase secondary structure within the 5'-noncoding region of a eucaryotic mRNA reduces translational efficiency. Cell 1985; 40:515. Darveau A, Pelletier J, Sonenberg A. Differential efficiencies of in vitro translation of mouse cmyc transcript differing in the 5'-untranslated region. Proc Natl Acad Sci U S A 1985; 82:2315. Shaw G, Kamen R. A conserved AU sequence from the 3'-untranslated region of GM-CSF mRNA mediates selective mRNA degradation. Cell 1986; 46:659. Nielsen DA, Shapiro DJ. Insights into hormonal control of messenger RNA stability. Mol Endocrinol 1990; 4:953. Atwater JA, Wisdom R, Verma IM. Regulated mRNA stability. Annu Rev Genet 1990; 24:519. Kozak M. Compilation and analysis of sequences upstream from the translational start site in eukaryotic mRNAs. Nucleic Acids Res 1984; 12:857. Kozak M. Selection of initiation sites by eucaryotic ribosomes: effect of inserting AUG triplets upstream from the coding sequence for preproinsulin. Nucleic Acids Res 1984; 12:3873. Kozak M. Bifunctional messenger RNAs in eukaryotes. Cell 1986; 47:481. Von Heijne G. A new method for predicting signal sequence cleavage sites. Nucleic Acids Res 1986; 14:4683. Gierasch LM. Signal sequences. Biochemistry 1989; 28:1. Walter P, Gilmore R, Blobel G. Protein translocation across the endoplasmic reticulum. Cell 1984; 38:5. Wickner WT, Lodish HF. Multiple mechanisms of protein insertion into and across membranes. Science 1985; 230:400. Wiedmann M, Kurzchalia TV, Hartmann E, Rapoport TA. A signal sequence receptor in the endoplasmic reticulum membrane. Nature 1987; 328:830. Dunphy WG, Rothman JE. Compartmental organization of the Golgi stack. Cell 1985; 42:13. Mellman I, Warren G. The road taken: past and future foundations of membrane traffic. Cell 2000; 100:99. Griffiths G, Simons K. The trans Golgi network: sorting at the exit site of the Golgi complex. Science 1986; 234:438. Kornfeld R, Kornfeld S. Assembly of asparagine-linked oligosaccharides. Annu Rev Biochem 1985; 54:631. Dice JF. Molecular determinants of protein half-lives in eukaryotic cells. FASEB J 1987; 1:349. Sanders SL, Schekman R. Polypeptide translocation across the endoplasmic reticulum membrane. J Biol Chem 1992; 267:13791. Kornfeld S. Trafficking of lysosomal enzymes in normal and disease states. J Clin Invest 1986; 77:1. Munro S. Pelham HRB. A C-terminal signal prevents secretion of lumenal ER proteins. Cell 1987; 48:899. Munro S, Pelham HRB. An HSP70-like protein in the ER: identity with the 78 kd glucose-regulated protein and immunoglobulin heavy chain binding protein. Cell 1986; 46:291. Johnson LM, Bankaitis VA, Emr SD. Distinct sequence determinants direct intracellular sorting and modification for a yeast vacuolar protease. Cell 1987; 48:875. Valls LA, Hunter CP, Rothman JH, Stevens TH. Protein sorting in yeast: the localization determinant of yeast vacuolar carboxypeptide Y residues in the propeptide. Cell 1987; 48:887. Moore HH, Kelly RB. Re-routing of a secretory protein by fusion with human growth hormone sequences. Nature 1986; 321:443. Kelly RB, Grote E. Protein targeting in the neuron. Annu Rev Neurosci 1993; 16:95. Burgess TL, Kelly RB. Constitutive and regulated secretion of proteins. Annu Rev Cell Biol 1987; 3:243. Loh YP, Snell CR, Cool DR. Receptor-mediated targeting of hormones to secretory granules. Role of carboxypeptidase E. Trends Endocrinol Metab 1997; 8:130. Tooze J, Tooze SA. Clathrin-coated vesicular transport of secretory proteins during the formation of ACTH-containing secretory granules in AtT-20 cells. J Cell Biol 1986; 103:839. Orci L, Ravazzola M, Storch M-J, et al. Proteolytic maturation of insulin is a post-Golgi event which occurs in acidifying clathrin-coated secretory vesicles. Cell 1987; 49:865. Mellman I, Fuchs R, Helenius A. Acidification of the endocytic and exocytic pathways. Annu Rev Biochem 1986; 55:663. Hong W, Tang BL. Protein trafficking along the exocytotic pathway. Bioas-says 1993; 15:231.

60a. Gerdes HH, Glombik MM. Signal-mediated sorting to the regulated pathway of protein secretion. Anat Anz 1999; 181:447. 61. Inoue K, Kurosumi K. Ultrastructural immunocytochemical localization of LH and FSH in the pituitary of the untreated male rat. Cell Tissue Res 1984; 235:77. 61a. Gullberg U, Bengtsson N, Bulow E, et al. Processing and targeting of granule proteins in human neutrophils. J Immunol Methods 1999; 232:201. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93.

DeLisle RC, Williams IA. Regulation of membrane fusion in secretory exocytosis. Annu Rev Physiol 1986; 48:225. Thomas G, Thorne BA, Hruby DE. Gene transfer technique to study neuropeptide processing. Annu Rev Physiol 1988; 50:323. Yamamoto KR. Steroid receptor regulated transcription of specific genes and gene networks. Annu Rev Genet 1985; 19:209. Shupnik MA, Chin WW, Habener JF, Ridgway EC. Transcriptional regulation of the thyrotropin subunit genes by thyroid hormone. J Biol Chem 1985; 260:2900. Larsen PR, Harney JW, Moore DD. Sequences required for cell type specific thyroid hormone regulation of rat growth hormone promoter activity. J Biol Chem 1986; 261:14373. Wright PA, Crew MD, Spindler SR. Discrete positive and negative thyroid hormone-responsive transcription regulatory elements of the rat growth hormone gene. J Biol Chem 1987; 262:5659. Flug F, Copp RP, Casanova J, et al. Cis-acting elements of the rat growth hormone gene which mediate basal and regulated expression by thyroid hormone. J Biol Chem 1987; 262:6373. Jantzen HM, Strahle U, Gloss B, et al. Cooperativity of glucocorticoid response elements located far upstream of the tyrosine aminotransferase gene. Nature 1987; 49:29. Weinberger C, Thompson CC, Ong ES, et al. The c-erb-A gene encodes a thyroid hormone receptor. Nature 1986; 324:64 1. Green S, Chambon P. A super family of potentially oncogenic hormone receptors. Nature 1986; 324:615. Brown DD. The role of stable complexes that repress and activate eucaryotic genes. Cell 1984; 37:359. von Hippel PH, Bear DG, Morgan WD, McSwiggen JA. Protein-nucleic acid interactions in transcription: a molecular analysis. Annu Rev Biochem 1984; 53:389. Harrison SC. A structural taxonomy of DNA-binding domains. Nature 1991; 353:715. Pabo CO. Transcription factors: structural families and principles of DNA recognition. Annu Rev Biochem 1992; 61:1053. Murdoch GH, Franco R, Evans RM, Rosenfeld RG. Polypeptide hormone regulation of gene expression. Thyrotropin-releasing hormone rapidly stimulates both transcription of the prolactin and the phosphorylation of a specific nuclear protein. J Biol Chem 1983; 258:15329. Montminy MR, Sevarino KA, Wagner JA, et al. Identification of a cyclic AMP responsive element within the rat somatostatin gene. Proc Natl Acad Sci U S A 1986; 83:6682. Hunter T, Karin M. The regulation of transcription by phosphorylation. Cell 1992; 70:375. Habener JF. Cyclic AMP response element binding proteins: a cornucopia of transcription factors. Mol Endocrinol 1990; 4:1087. Kadonaga JT, Tjian R. Affinity purification of sequence-specific DNA-binding proteins. Proc Natl Acad Sci U S A 1986; 83:5889. Brent R. Repression of transcription in yeast. Cell 1985; 42:3. Diamond MI, Miner JN, Yoshinaga SK, Yamamoto KR. Transcriptional factor interactions: selectors of positive and negative regulation from a single DNA element. Science 1990; 249:1266. Guarente L. Yeast promoters: positive and negative elements. Cell 1984; 36:799. Jones NC. Negative regulation of enhancers. Nature 1986; 321:202. Maniatis T, Goodbourn S, Fischer JA. Regulation of inducible and tissue-specific gene expression. Science 1987; 236:1237. Yanofsky C. Transcription attenuation. J Biol Chem 1988; 263:609. Brock ML, Shapiro DJ. Estrogen stabilizes vitellogenin mRNA against cytoplasmic degradation. Cell 1983; 34:207. Guyette WA, Matusik RJ, Rosen JM. Prolactin-mediated transcriptional and post-transcriptional control of casein gene expression. Cell 1979; 17:1013. Krane IM, Spindel ER, Chin WW. Thyroid hormone decreases the stability and the poly(A) tract length of rat thyrotropin -subunit messenger RNA. Mol Endocrinol 1991; 5:469. Koenig RJ, Lazar MA, Hodin RA, et al. Inhibition of thyroid hormone action by a non-hormone binding cerbA protein generated by alternative mRNA splicing. Nature 1989; 337:659. Chew SL. Alternative splicing of mRNA as a mode of endocrine regulation. Trends Endocrinol Metab 1997; 8:405. Douglass J, Civelli O, Herbert E. Polyprotein gene expression: generation of diversity of neuroendocrine peptides. Annu Rev Biochem 1984; 53:665. Wilson HE, White A. Prohormones: their clinical relevance. Trends Endocrinol Metab 1998; 9:396.

CHAPTER 4 HORMONAL ACTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 4 HORMONAL ACTION DARYL K. GRANNER General Features of Hormone Systems and Historical Perspective Target Cell Concept Hormone Receptors General Features Recognition and Coupling Domains of Receptors Receptor Occupancy and Bioeffect Agonist-Antagonist Concept Regulation of Receptors Structure of Receptors Classification of Hormones Mechanism of Action of Group I Hormones Mechanism of Action of Group II Hormones Chapter References

GENERAL FEATURES OF HORMONE SYSTEMS AND HISTORICAL PERSPECTIVE Multicellular organisms use intercellular communication mechanisms to ensure their survival by coordinating the responses necessary for adjusting to constantly changing external and internal environments. Two systems comprising several highly differentiated tissues have evolved to serve these functions. One is the nervous system, and the other is the endocrine system, which classically has been viewed as using mobile hormonal messages that are secreted from one gland or tissue to act on a distant tissue. There is an exquisite convergence of these regulatory systems. For example, neural regulation of the endocrine system is important; many neurotransmitters resemble hormones in their synthesis, release, transport, and mechanism of action; and many hormones are synthesized in the nervous system (see Chap. 175). The focus of this chapter is the endocrine system and how hormones work. The word hormone is derived from a Greek term that means to arouse to activity. Classically defined, a hormone is a substance that is synthesized in one organ and transported by the circulatory system to act on another tissue. However, this original description is too restrictive, because hormones can act on adjacent cells (i.e., paracrine action) and on the cell in which they were synthesized (i.e., autocrine action) without entering the circulation. Early studies concentrated on defining the endocrine action of hormones by removing or ablating an organ to localize the site of production. An extract of the tissue was then used to restore the function, and this served as a bioassay for subsequent purification of the hormone and the elucidation of physiologic and biochemical actions. This classic era of the study of hormonal action was descriptive. During this period, many hormones were discovered, and their major effects were defined. Because it was assumed that hormones had a unique source and a single or predominant action, they were named for the tissue of origin (e.g., thyroid hormone) or for the action (e.g., growth hormone). The next era of investigation of hormonal action was characterized by the discovery of many more hormones and by a more detailed analysis of how hormones work. The investigation of their functions was aided by methods and ideas previously exploited by endocrinologists, including the use of radioisotopes, the concept of turnover, improved means of purifying molecules, and the availability of sophisticated analytic machinery. Such studies changed the direction of research in hormonal action from a descriptive (i.e., organ or tissue) to a mechanistic (i.e., molecule or function) approach. Where a molecule worked was no longer as important as how it acted. A single hormone could have hemocrine (i.e., transportation through the blood), paracrine, or autocrine actions but affect the different target cells in a similar way, and some effects could be produced by a variety of hormones. For example, naming a single molecule the “growth hormone” was incorrect, because this hormone is but one of several—including the thyroid hormones, sex hormones, glucocorticoids, insulin, and various growth-promoting polypeptides—that are involved in growth, and growth promotion is only one of the actions of the so-called growth hormone. The principles of hormone synthesis, storage, secretion, transport, metabolism, and feedback control were established during this period. A major contribution was the elaboration of the concept of hormone receptors, and of the properties of specificity and selectivity of response, how target cells are defined, how responses are modulated, how signals are transduced from the outside of a cell to its interior, and how hormones can be classified according to their mechanism of action. The techniques of molecular biology and recombinant DNA have been applied to hormonal action with remarkable success. It is now possible to analyze hormonal effects on gene expression and to study which few nucleotides of the 3×109 in each haploid genome confer the response. Another exciting area is the overlapping spectrum of activity of components of hormonal action systems with nonhormonal proteins. Consider the similar features of the guanosine triphosphate (GTP)–binding proteins involved in the hormone-sensitive adenylate cyclase system with the transforming RAS oncogene family of proteins or with transducin, which is the protein that couples photoactivation to the visual response.1,2 The homology of platelet-derived growth factor (PDGF) gene and the v-sis transforming gene is remarkable, as is the similarity between the insulin and epidermal growth factor receptors, both of which have intrinsic tyrosine kinase activity.3,4,5 and 6 Researchers are exploring the molecular bases of endocrine diseases, such as pseudohypoparathyroidism, several types of dwarfism, Graves disease, certain types of extreme insulin resistance, testicular feminization, acromegaly, vitamin D resistance, and hereditary nephrogenic diabetes insipidus, to name a few.7,8,9,10,11,12,13 and 14 This knowledge has challenged many of the earlier concepts of hormonal action and endocrine disease.

TARGET CELL CONCEPT There are ~200 types of differentiated cells in humans. Only a few produce hormones, but virtually all of the 75 trillion cells in a human body are targets of one or more of the ~50 known hormones. The concept of target cells is undergoing redefinition. It was thought that hormones affected a single cell type, or only a few kinds of cells, and that a hormone elicited a unique biochemical or physiologic action. For example, it was presumed that thyroid-stimulating hormone (TSH) stimulated thyroid growth and thyroid hormonogenesis; adrenocorticotropic hormone (ACTH, also called corticotropin) enhanced growth and function of the adrenal cortex; glucagon increased hepatic glucose production; and luteinizing hormone (LH) stimulated gonadal steroidogenesis. However, these same hormones also stimulate lipolysis in adipose cells.15 Although the physiologic importance of this effect is unclear, the concept of unique sites of actions of these hormones is untenable. A more relevant example is that of insulin, which effects various responses in different cells and occasionally influences different processes within the same cell. It enhances glucose uptake and oxidation in muscle, lipogenesis in fat, amino acid transport in liver and lymphocytes, and protein synthesis in liver and muscle. These and other examples necessitated a reevaluation of the target cell concept. With the delineation of specific cell-surface and intracellular hormone receptors, the definition of a target has been expanded to include any cell in which the hormone binds to its receptor, whether or not a biochemical or physiologic response has been determined. This definition also is imperfect, but it has heuristic merit, because it presumes that not all actions of hormones have been elucidated. The response of a target cell is determined by the differentiated state of the cell, and a cell can have several responses to a single hormone. Cells can respond to a hormone in a hemocrine, paracrine, or autocrine manner. An example is the hormone gastrin-releasing peptide (also called mammalian bombesin). Gastrin-releasing peptide has hemocrine and paracrine actions in the gut but is produced by and stimulates the growth of small cell carcinoma cells of the lung.16 Several factors determine the overall response of a target cell to a hormone. The concentration of a hormone around the target cell depends on the rate of synthesis and secretion of the hormone, the proximity of target and source, the association-dissociation constants of the hormone with specific plasma carrier proteins, the rate of conversion of an inactive or suboptimally active form of the hormone into the active form, and the rate of clearance of the hormone from blood by other tissues or by degradation or excretion. The actual response to the hormone depends on the relative activity and state of occupancy, or both, of the specific hormone receptors on the plasma membrane or within the cytoplasm or nucleus; the metabolism of the hormone within the target cell; the presence of other factors within the target cell that are necessary for the hormone response; and postreceptor desensitization of the cell. Alterations of any of these processes can change the hormonal effect on a given target cell and must be considered in addition to the classic feedback loops.

HORMONE RECEPTORS GENERAL FEATURES

One of the major challenges in making the hormone-based communication system work is depicted in Figure 4-1. Hormone concentrations are very low in the extracellular fluid, generally in the range of 10-15 to 10-9 M. This is much lower than that of the many structurally similar molecules (e.g., sterols, amino acids, peptides) and other molecules that circulate at concentrations in the 10-5 to 10-3 M range. Target cells must identify the various hormones present in small amounts and differentiate a given hormone from the 106- to 109-fold excess of other, often closely related, molecules. This high degree of discrimination is provided by cell-associated recognition molecules called receptors. Hormones initiate their bioeffects by binding to specific receptors, and because any effective control system must provide a means of stopping a response, hormone-induced actions usually terminate after the effector dissociates from the receptor.

FIGURE 4-1. Specificity and selectivity of hormone receptors. Many different molecules circulate in the extracellular fluid (ECF), but only a few are recognized by hormone receptors. Receptors must select these molecules from among high concentrations of the other molecules. This simplified drawing shows that a cell may have no hormone receptors, have one receptor, have a receptor but no hormone in the vicinity, or have receptors for several hormones.

A target cell is defined by its ability to bind a given hormone selectively by means of a receptor, an interaction that is often quantitated using radioactive ligands that mimic hormone binding. Several features of this interaction are important. The radioactivity must not alter the bioactivity of the ligand. The binding should be specific, in which case the ligand is displaceable by unlabeled agonist or antagonist. Binding should be saturable. Binding should occur within the concentration range of the expected biologic response. RECOGNITION AND COUPLING DOMAINS OF RECEPTORS All receptors, whether for polypeptides or steroids, have at least two functional domains, and most have several more. A recognition domain binds the hormone, and a second region, the coupling domain, generates a signal that links hormone recognition to some intracellular function. The binding of hormone by receptor implies that some region of the hormone molecule has a conformation that is complementary to a region of the receptor molecule. The degree of similarity, or fit, determines the tightness of the association; this is measured as the affinity of binding. If the native hormone has a relative affinity of 1, other natural molecules range between 0 and 1. In absolute terms, this actually spans a binding affinity range of more than a trillion. Ligands with a relative affinity of more than 1 for some receptors have been synthesized and are used to study receptor biology. Coupling (i.e., signal transduction) occurs in two ways. Polypeptide and protein hormones, and the catecholamines, bind to receptors located in the plasma membrane, and thereby generate signals that regulate various intracellular functions. Steroids, thyroid hormones, retinoids, and other hormones of this class interact with intracellular receptors, and this complex provides the initial signal. The amino acid sequences of the recognition and coupling domains have been identified in many polypeptide hormone receptors. Hormone analogues with specific amino acid substitutions were used to change binding and alter the bioactivity of the hormone. Steroid hormone receptors also have these two functional domains; one site binds the hormone and the other binds to specific DNA regions. They also have other domains important for their function, which are described later. Several receptors have been characterized by recombinant DNA techniques, and structural analysis shows that these domains are highly homologous. This homology has been used to isolate cDNAs encoding several receptors that had not been obtained through classic protein purification procedures. The investigations have shown that these nuclear receptors are part of a large family of related proteins.17 This family of proteins is thought to regulate gene transcription, often in association with other transcription factors and coregulatory molecules. The ligands for these are called orphan receptors. The dual functions of binding and coupling ultimately define a receptor, and it is the coupling of hormone binding to signal transduction, called receptor-effector coupling, that provides the first step in the amplification of the hormonal response. This dual purpose also differentiates the target cell receptor from the plasma carrier proteins that bind hormone without generating a signal. It is important to differentiate the binding of hormones to receptors from the association that hormones have with various transport or carrier proteins. Table 4-1 lists several features of these functionally different classes of proteins.

TABLE 4-1. A Comparison of Hormone Receptors with Transport Proteins

RECEPTOR OCCUPANCY AND BIOEFFECT The concentrations of hormone required for occupancy of the receptor and for elicitation of a specific biologic response often are similar (Fig. 4-2A). This is especially true for steroid hormones, but some polypeptide hormones also exhibit this characteristic. This tight coupling is remarkable, considering the many steps that must occur between hormone binding and complex responses, such as transport, enzyme induction, cell lysis, or cell replication. When receptor occupancy and bioeffect are tightly coupled, significant changes in the latter occur when receptor occupancy changes. This happens when fewer receptors are available (Fig. 4-3A) or the affinity of the receptor changes but hormone concentration remains constant (see Fig. 4-3B). Otherwise, there is a marked dissociation of binding and effect, and a maximal bioeffect occurs when only a small percentage of the receptors are occupied (see effect 2 in Fig. 4-2B).

FIGURE 4-2. Hormone binding and biologic effect are compared in the absence (A) and presence (B, effect 2) of spare receptors. Some biologic effects in a tissue may be tightly coupled to binding, but others demonstrate the spare receptor phenomenon (e.g., compare effects 1 and 2 in B). (From Granner DK. Characteristics of hormone systems. In: Martin DW Jr, Mayer PA, Rodwell VW, Granner DK, eds. Harper's review of biochemistry, 20th ed. Los Altos, CA: Lange Medical Publications, 1985:501.)

FIGURE 4-3. Changes of receptor occupancy have large effects on the biologic response when effector and receptor occupancy are tightly coupled. This can occur when the receptor number changes (A) or when the affinity of the receptor for the hormone changes (B). In the hypothetical case shown in (A), a decrease from 20,000 receptors per cell to 10,000 results in a 50% decrease of the maximal response, a Vmax effect. A decrease in affinity (i.e., solid to interrupted line in [B], or rightward shift) means that more hormone is required for a given effect, but the same maximal response can be obtained. This is a Km effect. (From Granner DK. Characteristics of hormone systems. In: Murray RK, Granner DK, Mayer PA, Rodwell VW, eds. Harper's biochemistry, 21st ed. Norwalk, CT: Appleton & Lange, 1988.)

Receptors not involved in the elicitation of the response are called spare receptors. They are observed in the response of several polypeptide hormones and are thought to provide a means of increasing the sensitivity of a target cell to activation by low concentrations of hormone and to provide a reservoir of receptors. The concept of spare receptors is operational and may depend on which aspect of the response is examined and which tissue is involved. For example, there is excellent agreement between LH binding and cyclic adenosine monophosphate (cAMP) production in rat testis and ovarian granulosa cells (there generally are no spare receptors when any hormone activates adenylate cyclase), but steroidogenesis in these tissues, which is cAMP dependent, occurs when fewer than 1% of the receptors are occupied (see effects 1 and 2 in Fig. 4-2).18 Transcription of the phosphoenolpyruvate carboxykinase gene is repressed when far fewer than 1% of hepatoma cell insulin receptors are occupied, but there is a high correlation between insulin binding and amino acid transport in thymocytes.19 Other examples of the dissociation of receptor binding and biologic effects include the effects of catecholamines on muscle contraction, lipolysis, and ion transport.20 These end-responses presumably reflect a cascade or multiplier effect of the hormone. Different responses within the same cell can require various degrees of receptor occupancy. For example, successively greater degrees of occupancy of the adipose cell insulin receptor increase, in sequence, lipolysis, glucose oxidation, amino acid transport, and protein synthesis.21 AGONIST-ANTAGONIST CONCEPT Molecules can be divided into four groups according to their ability to elicit a hormone receptor–mediated response. These classes are agonists, partial agonists, antagonists, and inactive agents (Table 4-2).

TABLE 4-2. Classification of Steroids According to Their Action as Glucocorticoids

Agonists elicit the maximal response, although different concentrations may be required. In the example of Figure 4-4,1,2 and 3 could be porcine insulin, porcine proinsulin, and guinea pig insulin, respectively. In all systems tested, these insulins have the same rank order of potency, but each elicits a maximal response if present in sufficient concentration. Likewise, 1, 2, and 3 could be dexamethasone, cortisol, and corticosterone (see Table 4-2).Partial agonists evoke an incomplete response even when very large concentrations of the hormone are used, as shown by line B of Figure 4-5. Antagonists generally have no effects themselves, but they competitively inhibit the action of agonists or partial agonists (see lines A through C in Fig. 4-5). Many structurally similar compounds elicit no effect and have no effect on the action of the agonists or antagonists. These are classified as inactive agents and are represented as line D in Figure 4-5.

FIGURE 4-4. Within a class of hormones—glucocorticoids, for example—different molecules may have different potencies. In this case, hormones 1, 2, and 3 are all agonists, but very different concentrations are required to achieve a given biologic response. The binding of steroid to receptor would parallel each of these curves. (From Granner DK. Characteristics of hormone systems. In: Murray RK, Granner DK, Mayer PA, Rodwell VW, eds. Harper's biochemistry, 21st ed. Norwalk, CT: Appleton & Lange, 1988.)

FIGURE 4-5. Classification of hormones according to their biologic activity. Steroids, for example, can be classified as agonists (line A), partial agonists (line B), antagonists (C in A+C or B+C), or inactive agents (dotted line D). This drawing represents induction of the enzyme tyrosine aminotransferase. (From Granner DK. Characteristics of hormone systems. In: Murray RK, Granner DK, Mayer PA, Rodwell VW, eds. Harper's biochemistry, 21st ed. Norwalk, CT: Appleton & Lange, 1988.)

Partial agonists also compete with agonists for binding to and activation of the receptor, when they become partial antagonists. The extent of the inhibition of agonist activity caused by partial or complete antagonists depends on the relative concentration of the various steroids. Generally, much higher concentrations of the antagonist are required to inhibit an agonist than are necessary for the latter to exert its maximal effect. Because these concentrations are rarely achieved in vivo, this phenomenon is used for studies of the mechanism of action of hormones in vitro. The binding of a ligand to the receptor must facilitate a change in this molecule so that it can bind to DNA. This phenomenon was first suggested in studies that used the steroids in Table 4-2.22 The hypothesis assumes that agonists bind to and fully activate the receptor and elicit the maximal biologic response; that partial agonists fully occupy the receptor but afford incomplete activation and therefore a partial response; and that antagonists fully occupy the receptor, but because this complex is unable to bind to DNA, it elicits no intrinsic response but does inhibit the action of agonists. REGULATION OF RECEPTORS The number of hormone receptors on or in a cell is in a dynamic state and can be regulated physiologically or be influenced by diseases or therapeutic measures. The receptor concentration and affinity of hormone binding can be regulated. Some changes can be acute and can significantly affect hormone responsiveness of the cell. For instance, cells exposed to b-adrenergic agonists for minutes to hours no longer activate adenylate cyclase in response to more agonist, and the biologic response is lost. This desensitization occurs by two mechanisms.23 The loss of receptors, called down-regulation, involves the internal sequestration of receptors, segregating them from the other components of the response system, including the regulatory and catalytic subunits of adenylate cyclase. Removal of the agonist results in the return of receptors to the cell surface and restoration of hormonal sensitivity.23 An example of a second form of desensitization of the a-adrenergic system involves the covalent modification of receptor by phosphorylation.24 This cAMP-dependent process entails no change in receptor number and no translocation. Reconstitution experiments show that because the phosphorylated receptor is unable to activate adenylate cyclase, the activation and hormone binding domains are uncoupled.23 Other examples of physiologic adaptation that is accomplished through down-regulation of receptor number by the homologous hormone include insulin, glucagon, thyrotropin-releasing hormone, growth hormone, LH, follicle-stimulating hormone, and catecholamines. A few hormones, such as angiotensin II and prolactin, up-regulate their receptors. The changes in receptor number can occur over a period of minutes to hours and are probably an important means of regulating biologic responses. How the loss of receptor affects the biologic response elicited at a given hormone concentration depends on whether there are spare receptors (Fig. 4-6). Suppose there is a fivefold reduction in receptor number in a cell. With no spare receptors (see Fig. 4-6A), the maximal response obtained is 20% that of control, hence, the effect is on the Vmax. With spare receptors (see Fig. 4-6B), the maximal response is obtained, but at five times the originally effective hormone concentration, analogous to a Km effect.

FIGURE 4-6. The effect a five-fold loss of receptors has on a biologic system that lacks (A) or has (B) spare receptors. (From Granner DK. Characteristics of hormone systems. In: Martin DW Jr, Mayer PA, Rodwell VW, Granner DK, eds. Harper's review of biochemistry, 20th ed. Los Altos, CA: Lange Medical Publications, 1985:502.)

STRUCTURE OF RECEPTORS The acetylcholine receptor (AChR), which exists in relatively large amounts in the electric organ of Torpedo californica, was the first plasma membrane–associated receptor to be studied in detail. The AChR consists of four subunits: a2, b, d and g.25 The two a subunits bind acetylcholine.26 The technique of site-directed mutagenesis has been used to show which regions of this subunit participate in the formation of the transmembrane ion channel, which is the major function of the AChR.25 Other receptors occur in very small amounts, and recombinant DNA techniques have been used to deduce many of the structures and to find and characterize new receptors. The insulin receptor is a heterotetramer (a2b2) linked by multiple disulfide bonds, in which the extramembrane a subunit binds insulin and the membrane-spanning b subunit transduces the signal through the tyrosine kinase component of the cytoplasmic portion of this polypeptide27 (Fig.4-7). The insulin-like growth factor-I (IGF-I) receptor has a similar structure, and the epidermal growth factor (EGF) and low-density lipoprotein receptors are similar in many respects28,29 and 30 (see Fig. 4-7). Receptors that couple ligand binding to signal transduction through G-protein intermediaries characteristically have seven membrane-spanning domains.31

FIGURE 4-7. Schematic representation of the structures of the low-density lipoprotein (LDL), epidermal growth factor (EGF), and insulin receptors. The amino terminus (NH2) of each is in the extracellular portion of the molecule. The carboxyterminus (COOH) is in the cytoplasm. The open boxes represent cysteine-rich regions that are thought to be involved in ligand binding. Each receptor has a short domain (~25 amino acids) that traverses the plasma membrane (hatched line) and an intracellular domain of variable length. The EGF and insulin receptors have tyrosine kinase activity associated with the cytoplasmic domain ( ) and have autophosphorylation sites in this region. The insulin receptor is a heterotetramer connected by disulfide bridges (vertical bars).

Members of the nuclear receptor superfamily have several functional domains: a ligand-binding domain in the carboxyl-terminal region, an adjacent DNA-binding domain, and one or more trans-activation domains. There may also be dimerization, nuclear translocation, and heat shock protein domains, and regions that allow for interactions with a number of other accessory factor and coregulatory proteins17,32 (Fig. 4-8). The amino acid sequence homology is particularly strong in the various DNA-binding domains, and it was this feature that led to the elucidation of the nuclear receptor superfamily.17

FIGURE 4-8. Nuclear receptor family members have several general domains. The amino-terminal region is most variable and often contains a trans -activating domain (TAD1). The DNA-binding domain (DBD) is most conserved, and this feature led to the discovery that these receptors are part of a large family of DNA-binding proteins. The hormone-or ligand-binding domain (LBD), which affords specificity, is located in the carboxyl-terminal (COOH) region of the molecule and contains a second trans -activating domain (TAD2). Also shown are regions that allow for nuclear translocation, dimerization, and interaction with heat shock protein (Hsp90). Members of this family that have no known ligand are called orphan receptors.

CLASSIFICATION OF HORMONES A classification based on the location of receptors and the nature of the signal used to mediate hormonal action within the cell appears in Table 4-3, and general features of each group are listed in Table 4-4.

TABLE 4-3. Hormones and Their Actions: Classification According to Mechanism of Action

TABLE 4-4. General Features of Hormone Groups

The hormones in group I are lipophilic. After secretion, these hormones associate with transport proteins, a process that circumvents the solubility problem while prolonging the plasma half-life by preventing the hormone from being metabolized and excreted. These hormones readily traverse the plasma membrane of all cells and encounter receptors in the cytosol or the nucleus of target cells. The ligand-receptor complex is assumed to be the intracellular messenger in this group. The second major group consists of water-soluble hormones that bind to the plasma membrane of the target cell. These hormones regulate intracellular metabolic processes through intermediary molecules, called second messengers (the hormone itself is the first messenger), which are generated because of the ligand-receptor interaction. The second-messenger concept arose from the observation of Sutherland33 that epinephrine binds to the plasma membrane of pigeon erythrocytes and increases intracellular cAMP. This was followed by a series of experiments in which cAMP was found to mediate the metabolic effects of many hormones. Hormones that use this mechanism are shown in group IIA. Several hormones, some of which were previously thought to affect cAMP, appear to use cyclic guanosine

mono-phosphate (cGMP) (group IIB) or calcium or phosphatidylinositide metabolites (or both) as the intracellular signal (group IIC). The intracellular messenger has been identified as a protein kinase/phosphatase cascade for the hormones listed in group D. A few hormones fit in more than one category (i.e., some hormones act through cAMP and Ca2+), and assignments change with new information. MECHANISM OF ACTION OF GROUP I HORMONES A schematic representation of the mechanism of action of group I hormones (see Table 4-3) is shown in Figure 4-9. These lipophilic molecules probably diffuse through the plasma membrane of all cells but encounter their specific, high-affinity receptor only within target cells. The hormone-receptor complex then undergoes an “activation” reaction that causes size, conformation, and surface charge changes that render it able to bind to chromatin. In some cases—with the glucocorticoid receptor, for example—this process involves the disruption of a receptor–heat shock protein complex. Whether the association and activation processes occur in the cytoplasm or nucleus appears to depend on the specific hormone. The hormone-receptor complex binds to specific regions of DNA and activates or inactivates specific genes.34,35 By selectively affecting gene transcription and the production of the respective messenger RNAs (mRNAs), the amounts of specific proteins are changed, and metabolic processes are influenced. The effect of each of these hormones is specific; generally, the hormone affects 1 hour. The analysis of pulsatile variations may be considered at two levels. The researcher may wish to define and characterize significant variations in peripheral levels based on estimations on the size of measurement error (i.e., primarily assay error). However, under certain circumstances, it is possible to mathematically derive secretory rates from the peripheral concentrations. This procedure, often referred to as deconvolution, often reveals more pulses of secretion than the analysis of peripheral concentrations. It also more accurately defines the temporal limits of each pulse. The association between pulsatile GH secretion and sleep stages in a single individual is shown in Figure 6-7.28 The profile shown on the top represents the plasma levels of GH measured at 15-minute intervals. Three pulses were found significant using a pulse detection algorithm (i.e., ULTRA28). The corresponding profile of GH secretory rates calculated by deconvolution is shown in the second profile from the top. A single compartment model for GH disappearance with a half-life of 19 minutes and a volume of distribution of 7% of the body weight was used in this calculation. Pulse analysis of the secretory rates now reveals the occurrence of three additional pulses of GH secretion. The three lower profiles show the percentages of each 15-minute sampling interval spent in stages of wake, SW, and REM, respectively. When the profile of plasma concentrations is compared with the SW profile, it appears that subsequent to its initiation in concomitance with the beginning of the first SW period, the sleep-onset GH pulse spanned the first 3 hours of sleep, without apparent modulation by non-REM and REM stages. However, the profile of secretory rates clearly reveals that GH was preferentially secreted during the SW stage, with interruptions of secretory activity coinciding with the intervening REM or wake stages. Deconvolution demonstrated a closer association between SW stages and active GH secretion than the analysis of plasma levels, because the temporal limits of each pulse were more accurately defined and additional pulses were revealed. The validity of the deconvolution procedure is critically dependent on the knowledge of the clearance kinetics of the hormone under study; extra caution in interpreting the data must be exerted, because this procedure involves an amplification of measurement error, with increased risk of false-positive error.

FIGURE 6-7. 24-hour profile of plasma growth hormone (GH; top) sampled at 15-minute intervals in a normal man. The black bar indicates the sleep period. The second panel from the top shows the profile of GH secretory rates derived by deconvolution from the profile of plasma concentrations. Significant pulses of plasma levels and secretory rates are indicated by arrows. The three lower panels represent the temporal distribution of slow-wave (SW) stages (III + IV), wake, and rapid eye movement (REM) during sleep. Vertical lines indicate the temporal association between pulses of GH secretion and SW stages. (From Van Cauter E. Computer-assisted analysis of endocrine rhythms. In: Rodbard D, Forti G, eds. Computers in endocrinology. New York: Raven Press, 1990:59.)

Whether examining peripheral concentrations or secretory rates, there are two major approaches to analyzing the episodic fluctuations. The first, and most commonly used, is the time domain analysis in which the data are plotted against time and pulses are detected and identified. The second is the analysis in the frequency domain in which amplitude is plotted against frequency or period. These two approaches differ fundamentally both in the mathematical treatment of the data and in the questions they may help to resolve; therefore, they should be viewed as being complementary. The regularity of pulsatile behavior may be quantified by both approaches (i.e., by examining the distribution of interpulse intervals derived from a time domain analysis or by examining the distribution of spectral power in a

frequency domain analysis). Additionally, another analytical tool, the approximate entropy, has been introduced to quantify regularity of oscillatory behavior in endocrine and other physiologic time series.29,30

RHYTHMS IN THE SOMATOTROPIC AXIS In normal subjects, the 24-hour profile of plasma GH consists of stable low levels abruptly interrupted by bursts of secretion (see Fig. 6-2, Fig. 6-5, Fig. 6-6 and Fig. 6-7). The most reproducible secretory pulse occurs shortly after sleep onset, in association with the first phase of SW sleep.31 Other secretory pulses may occur in later sleep and during wakefulness in the absence of any identifiable stimulus. In women, daytime GH pulses are more frequent than in men, and the sleep-associated pulse, although still present, does not generally account for most of the 24-hour GH release. The secretory profile is less regular in women than in men.32 Circulating estradiol concentrations play an important role in determining overall levels of spontaneous GH secretion.33 Sleep onset elicits a pulse in GH secretion whether sleep is advanced, delayed, interrupted, or fragmented. Delta wave electroencephalographic activity consistently precedes the elevation in plasma GH levels. While SW sleep is clearly a major determinant of the 24-hour profile of GH secretion in humans, there is also evidence for the existence of a circadian modulation of the occurrence and height of GH pulses, reflecting decreased somatostatin inhibition in the evening and nocturnal hours.34 Administration of a specific GHRH antagonist results in a near total suppression of sleep-related GH release, indicating an important role for GHRH in the control of nocturnal GH secretion.35 Two studies involving pharmacologic stimulation of SW sleep have provided evidence for a common mechanism in the control of SW sleep and GH secretion and have indicated that compounds, which increase SW sleep, could represent a novel class of GH secretagogues. Indeed, enhancement of SW sleep by oral administration of low g-hydroxybutyrate (a naturally occurring metabolite of g-aminobutyric acid used in the treatment of narcolepsy) or ritanserin (a selective 5HT2 receptor antagonist) results in simultaneous and highly correlated increases of nocturnal GH release.36,37 The total amount and the temporal distribution of GH release are strongly dependent on age.31 A pulsatile pattern of GH release, with increased pulse amplitude during sleep, is present in prepubertal boys and girls. During puberty, the amplitude of the pulses, but not the frequency, is increased, particularly at night. Maximal overall GH concentrations are reached in early puberty in girls and in late puberty in boys. Age-related decreases in GH secretion have been well documented in both men and women and are illustrated in Figure 6-6. There is a marked suppression of GH levels throughout the 24-hour span in obese subjects. In normal-weight subjects, fasting, even for only 1 day, enhances GH secretion via an increase in both pulse amplitude and pulse frequency.38 Nonobese juvenile or maturity-onset diabetic patients hypersecrete GH during wakefulness as well as during sleep, primarily because of an increase in the amplitude of pulses.39 This abnormality may disappear when blood sugar levels are strictly controlled. In acromegaly, GH is hypersecreted throughout the 24-hour span, with a pulsatile pattern superimposed on elevated basal levels, indicative of the presence of tonic secretion.40,41 After trans-sphenoidal surgery, a normal circadian pattern of GH release can be restored.41 In contrast, bromocriptine therapy lowers the overall GH secretion but does not lead to the resumption of normal 24-hour profiles.

RHYTHMS IN THE PITUITARY-ADRENAL AXIS Twenty-four-hour profiles of cortisol typical of normal subjects are shown in Figure 6-2, Figure 6-3, Figure 6-4 and Figure 6-6. The patterns of plasma adrenocorticotropic hormone (ACTH) and cortisol variations show an early morning maximum, declining levels throughout daytime, a quiescent period of minimal secretory activity, and an abrupt elevation during late sleep. With a 15-minute sampling rate, ~15 pulses of ACTH and cortisol can be detected in a 24-hour span. The cortisol profiles shown in the upper panels of Figure 6-2 illustrate the remarkable persistence of the wave shape of the rhythm in the absence of sleep and support the notion that the 24-hour periodicity of corticotropic activity is primarily controlled by circadian rhythmicity. Nevertheless, modulatory effects of the sleep or wake condition have been clearly demonstrated. Sleep onset is reliably associated with a short-term inhibition of cortisol secretion,42,43 and 44 although this effect (which appears to be related to slow-wave stages45) may not be detectable when sleep is initiated at the peak of corticotropic activity (i.e., in the morning13). Conversely, awakening at the end of the sleep period is consistently followed by a pulse of cortisol secretion.46,47 During sleep deprivation, these rapid effects of sleep onset and sleep offset on corticotropic activity are obviously absent, and the amplitude of the rhythm is reduced as compared with normal conditions (see Fig. 6-2). In addition to the immediate modulatory effects of sleep-wake transitions on cortisol levels, nocturnal sleep deprivation—even partial—results in higher-than-normal cortisol concentrations on the following evening.48 Sleep loss, thus, appears to delay the normal return to evening quiescence of the corticotropic axis. This suggests that sleep loss may slow down the rate of recovery of the hypothalamic–pituitary–adrenal axis response after a challenge. A circadian variation parallel to that of cortisol occurs for the plasma levels of adrenal steroids.49 Figure 6-8 shows the mean 24-hour profiles of cortisol, 11-hydroxyandrostenedione, dehydroepiandrosterone, and androstenedione (AD) for 10 normal young men. The amplitude of the circadian variation in 11-hydroxyandrostenedione levels, a steroid derived from adrenal AD, is essentially similar to that of cortisol, whereas the amplitude of the variations in dehydroepiandrosterone and AD levels, two steroids of partially gonadal origin, is much lower.49 Pulses of the plasma concentrations of adrenal steroids occur in remarkable synchrony with bursts of cortisol secretion, indicating that pulsatile ACTH release is reflected in the temporal organization of all adrenal secretions.

FIGURE 6-8. Mean 24-hour profiles of plasma cortisol, 11-hydroxyan-drostenedione (11-OAD), dehydroepiandrosterone (DHEA), and andro-stenedione (AD) obtained at 15-minute intervals in 10 normal young men. For each subject and each hormone, the data were expressed as a percentage of the individual 24-hour mean before calculating group values. The vertical bars at each time point represent the standard error of the mean. (Data from Lejeune-Lenain C, Van Cauter E, Desir D, et al. Control of circadian and episodic variations of adrenal androgens secretion in man. J Endocrinol Invest 1987; 10:267.)

A distinct circadian rhythm of serum cortisol levels emerges at ~6 months of age. Once this periodicity has been established, it persists throughout adulthood and has been observed through the ninth decade.21 The overall pattern of the rhythm remains unchanged (see Fig. 6-6), but in older subjects, the nadir is advanced by 1 to 2 hours and the amplitude is decreased.21 In women, oral contraceptive therapy results in a large increase of the mean cortisol level and of the amplitude of the rhythm resulting from an estrogen-induced elevation of transcortin-binding capacity. Thus, when hypercortisolism is suspected in a female patient, it is essential to know whether the patient is receiving estrogen treatment. The 24-hour profile of pituitary-adrenal secretion remains unaltered in a wide variety of pathologic states. Disease states in which alterations of the cortisol rhythm have been observed50 include primarily (a) disorders involving abnormalities in binding and/or metabolism of cortisol, (b) the various forms of Cushing syndrome, and (c) severe depression. The relative amplitude of the circadian rhythm and of the episodic fluctuations of cortisol is blunted in patients with liver disease and in patients with anorexia nervosa, primarily because of the decreased metabolic clearance of cortisol. In contrast, in hyperthyroidism, for which cortisol production and peripheral metabolism are increased, episodic pulses are enhanced. In hypothyroid patients, there is diminished cortisol clearance, the mean level is markedly elevated, and the relative amplitude of the rhythm is, therefore, dampened. Figure 6-9 illustrates typical 24-hour cortisol patterns of plasma cortisol and cortisol secretory rates in a patient with pituitary Cushing disease and a patient with major endogenous depression as compared with a normal subject. In patients with Cushing syndrome secondary to adrenal adenoma or ectopic ACTH secretion, the circadian variation of plasma cortisol is invariably absent. However, in pituitary-dependent Cushing disease, a low-amplitude circadian variation may persist, suggesting that there is no defect in the neural clock generating the periodicity. Cortisol pulsatility is blunted in ~70% of patients with Cushing disease, suggesting autonomous tonic secretion of ACTH by a pituitary tumor. However, in ~30% of these patients, the magnitude of the pulses is instead enhanced. These “hyperpul-satile” patterns could be caused by enhanced hypothalamic release of CRH or persistent pituitary responsiveness to CRH. Hypercortisolism with persistent circadian rhythmicity and increased pulsatility is found in a majority of acutely depressed patients. In these patients, who do not

develop the clinical signs of Cushing syndrome despite the high cortisol levels, the quiescent period often occurs earlier than in normal subjects of comparable age. This phase-advance could reflect an alteration in the regulation of the circadian pacemaker system. When a clinical remission is obtained, the hypercortisolism and the abnormal timing of the quiescent period disappear, indicating that these disturbances are “state” rather than “trait” dependent. Contrasting with the increased cortisol pulsatility that characterizes many patients with major depression, a few studies have suggested that pulsatile variations are of lower amplitude in post-traumatic stress disorder and chronic fatigue syndrome.51,52

FIGURE 6-9. Twenty-four-hour profiles of cortisol secretory rate (top) and plasma cortisol (bottom) in a normal subject (left), a patient with pituitary Cushing disease (middle), and a patient with major endogenous depression of the unipolar subtype (right). Cortisol secretory rates were derived from plasma levels using a two-compartment model for cortisol distribution and metabolism. Note that circadian rhythmicity is markedly attenuated in the subject with Cushing disease, but it is preserved in the depressed patient. Cortisol secretion is entirely intermittent in the normal subject and the depressed patient, but the secretory pattern of the patient with Cushing disease shows evidence of tonic cortisol release.

RHYTHMS IN PROLACTIN SECRETION Under normal conditions, the 24-hour profile of plasma PRL levels (see Chap. 13) follows a bimodal pattern, with minimal concentrations around noon, an afternoon phase of augmented secretion, and a major nocturnal elevation starting shortly after sleep onset and culminating around midsleep. Episodic pulses occur throughout the 24-hour span. Morning awakening is consistently associated with a brief PRL pulse.53,54 Studies on PRL during daytime naps or after shifts of the sleep period have demonstrated that sleep onset is invariably associated with an increase in PRL secretion. This is well illustrated by the profiles shown in Figure 6-2 in the presence and in the absence of sleep and in Figure 6-4, which compares the profiles of day and night workers. However, a sleep-independent circadian component of PRL secretion may be observed in some individuals, particularly in women.54 An example may be seen in the PRL profiles of night workers (see Fig. 6-4) in whom a nocturnal elevation occurred despite nocturnal activity.15 When sleep structure is characterized by power spectral analysis of the electroencephalogram, a close temporal association between increased PRL secretion and SW activity is clearly apparent.55 Conversely, prolonged awakenings, which interrupt sleep, are consistently associated with decreasing PRL concentrations. Thus, shallow and fragmented sleep is generally associated with lower nocturnal PRL levels. This is indeed what is observed in elderly subjects (see Fig. 6-6),21 who have an increased number of awakenings and markedly decreased amounts of SW sleep, and in whom a dampening of the nocturnal PRL rise is evident. Benzodiazepine hypnotics taken at bedtime often cause an increase in the nocturnal PRL rise, resulting in concentrations in the pathologic range for part of the night.56 Absence or blunting of the nocturnal increase of plasma PRL has been reported in a variety of pathologic states, including uremia, breast cancer in postmenopausal women, and Cushing disease. In subjects with insulin-dependent diabetes, the circadian and sleep modulation of PRL secretion is preserved, but overall levels are markedly diminished.57 In hyperprolactinemia associated with prolactinomas, the nocturnal elevation of PRL is preserved in patients with microadenomas but altered in patients with macroadenomas.58 Selective removal of PRL-secreting microadenomas can result in the normalization of the PRL pattern.

RHYTHMS IN THE GONADOTROPIC AXIS Rhythms in the gonadotropic axis cover a wide range of frequencies, from episodic release in the ultradian range to diurnal rhythmicity and monthly and seasonal cycles. These various rhythms interact to provide a coordinated temporal program governing the development of the reproductive axis and its operation at every stage of maturation. The following summary is centered on 24-hour rhythms and their interaction with pulsatile release at the various stages of maturation of the human reproductive system. More detailed reviews may be found elsewhere.50,59,60 Before puberty, LH levels are very low. Both LH and FSH appear to be secreted in a pulsatile pattern but with a very low amplitude, which is insufficient to activate the gonad. The onset of puberty is associated with an augmentation of pulsatile activity in a majority of both girls and boys. In pubertal children, the magnitude of the nocturnal pulses of LH and FSH is consistently increased during sleep. As the pubescent child enters adulthood, the daytime pulse amplitude increases as well, eliminating or diminishing the diurnal rhythm. In pubertal girls, there is a diurnal variation of circulating estradiol levels, with higher concentrations during the daytime than during the nighttime. The lack of parallelism between gonadotropin and estradiol levels reflects a 6- to 8-hour delay between gonadotropin stimulation and the subsequent ovarian response. In pubertal boys, the nocturnal rise of testosterone coincides with the elevation of gonadotropins. Patterns of LH release in adult men exhibit large interindi-vidual variability. The diurnal variation is dampened and may become undetectable. A marked diurnal rhythm in circulating testosterone levels in young adults, with minimal levels in the late evening and maximal levels in the early morning, has been well demonstrated. In young male adults, the amplitude of the circadian variation averages 25% of the 24-hour mean.49 In older men, the amplitude of LH pulses is decreased, and no significant diurnal pattern can be detected. However, the circadian rhythm in testosterone remains apparent, although markedly dampened. In adult women, the 24-hour variation in plasma LH is modulated by the menstrual cycle. In the early follicular phase, LH pulses are large and infrequent, and a slowing of the frequency of secretory pulses occurs during sleep. In the midfollicular phase, pulse amplitude is decreased, pulse frequency is increased, and frequency modulation of LH pulsatility by sleep is less apparent. Pulse amplitude increases again by the late follicular phase. No modulation by sleep is apparent until the early luteal phase, when nocturnal slowing of pulsatility is again evident. During the luteal-follicular transition, there is a four- to five-fold increase in LH pulse frequency, which accompanies the selective FSH rise necessary for normal folliculogenesis. Toward menopause, gonadotropin levels are elevated but show no consistent circadian pattern. Abnormal ultradian and/or circadian hormonal profiles have been found in a wide variety of reproductive disorders. The findings pertaining to disorders of female and male reproduction for which an abnormal function of the hypothalamic pulse generator and/or its modulation by diurnal rhythmicity seem to be primarily involved have been reviewed elsewhere.50,59,60

RHYTHMS IN THE THYROTROPIC AXIS In normal adult men and women, TSH levels are low throughout the daytime and begin to increase in the late afternoon or early evening.61 Maximal levels occur shortly before sleep. During sleep, TSH levels generally decline slowly. A further decrease occurs in the morning hours (see Fig. 6-2 and Fig. 6-6). Studies involving sleep deprivation and shifts of the sleep-wake cycle have consistently indicated that an inhibitory influence is exerted on TSH secretion during sleep. Interestingly, when sleep occurs during daytime hours, TSH secretion is not suppressed significantly below normal daytime levels. When the depth of sleep is increased by prior sleep deprivation, the nocturnal TSH rise is even further blunted. There is a consistent association between descending slopes of TSH concentrations and SW stages.62 The pronounced enhancing effect of sleep deprivation on the nighttime TSH rise is illustrated in Figure 6-2. The timing of the evening rise seems to be controlled by circadian rhythmicity. The temporal pattern of TSH secretion seems to reflect both tonic and pulsatile release, with both the frequency and the amplitude of the pulses increasing during the nighttime. Pulses of TSH secretion persist during somatostatin or dopamine treatment, suggesting that the control of pulsatility is largely thyrotropin-releasing hormone dependent.61 Because triiodothyronine and thyroxine are largely bound to serum protein, the existence of a circadian rhythm independent of postural changes has been difficult to establish. However, under conditions of sleep deprivation, the increased amplitude of the TSH rhythm results in an increased amplitude of the tri-iodothyronine rhythm, which becomes detectable in a majority of subjects.

The fact that the inhibitory effects of sleep on TSH secretion are time dependent may cause, under certain circumstances, elevations of plasma TSH levels, which reflect the misalignment of sleep and circadian timing. Figure 6-10 shows the mean profiles of plasma TSH levels observed in a group of normal young men in the course of adaptation to simulated jet lag. 63 After a 24-hour baseline period, the sleep-wake cycle and the dark period were abruptly advanced by 8 hours. In the course of adaptation, TSH levels increased progressively, because daytime sleep failed to inhibit TSH, and nighttime wakefulness was associated with large circadian-dependent TSH elevations. As a result, mean TSH levels after awakening from the second shifted sleep period were more than two-fold higher than during the same time interval after normal nocturnal sleep. This study demonstrates that the subjective discomfort and fatigue associated with jet lag may involve a prolonged elevation of a hormonal concentration in the peripheral circulation.

FIGURE 6-10. Mean (and standard error of the mean) profiles of plasma thyrotropin (thyroid-stimulating hormone [TSH]) from eight normal young men who were subjected to an 8-hour advance of the sleep-wake and dark-light cycles. Black bars indicate bedtime periods. (Data from Hirschfeld U, Moreno-Reyes R, Akseki E, et al. Progressive elevation of plasma thyrotropin during adaptation to simulated jet lag: effects of treatment with bright light or zolpidem. J Clin Endocrinol Metab 1996; 81:3270.)

In older adults, there is an overall decrease of TSH levels (see Fig. 6-6), but the rhythm persists, albeit slightly dampened and with an earlier evening rise than in younger adults.64 Fasting decreases overall TSH levels by decreasing pulse amplitude, resulting in a dampening of the nocturnal surge.65 A decreased or absent nocturnal rise of TSH has been observed in a wide variety of nonthyroidal illnesses, suggesting that hypothalamic dysregulation generally affects the circadian TSH surge. The nocturnal TSH surge is diminished or absent in hyperthyroidism, central hypothyroidism, and in various conditions of hypercortisolism. The lack of normal nocturnal elevation of TSH levels also appears to be a sensitive index of preclinical hyperthyroidism. In poorly controlled diabetic states, whether type 1 or type 2, the surge also disappears.66 Correction of hyperglycemia is associated with a reappearance of the nocturnal elevation.66

RHYTHMS IN GLUCOSE REGULATION In normal humans, glucose tolerance varies with the time of day.67 Figure 6-11 illustrates circadian variations in glucose tolerance to oral glucose, identical meals, constant glucose infusion, and enteral nutrition in normal subjects. Plasma glucose responses to oral glucose, intravenous glucose, or meals are markedly higher in the evening than in the morning. Overnight studies of subjects sleeping in the laboratory have consistently observed that despite the prolonged fasting condition, glucose levels remain stable or fall only minimally across the night, contrasting with the clear decrease that is associated with daytime fasting. Thus, a number of mechanisms operative during nocturnal sleep must intervene to maintain stable glucose levels during the overnight fast. Experimental protocols using intravenous glucose infusion or enteral nutrition (the two experimental conditions allowing for the study of nighttime glucose tolerance during sleep without awakening the subjects) have shown that glucose tolerance deteriorates further as the evening progresses, reaches a minimum around midsleep, and then improves to return to morning levels (see Fig. 6-11).67 There is evidence to indicate that this diurnal variation in glucose tolerance is partly driven by the wide and highly reproducible 24-hour rhythm of circulating levels of cortisol, an important counterregulatory hormone.68 Diminished insulin sensitivity and decreased insulin secretion in relation to elevated glucose levels are both involved in causing reduced glucose tolerance later in the day. During the first part of the night, decreased glucose tolerance is due to decreased glucose utilization both by peripheral tissues (relaxed muscles and rapid insulin-like effects of sleep-onset GH secretion) and by the brain (imaging studies have demonstrated a reduction in glucose uptake during SW sleep). During the second part of the night, these effects subside, as sleep becomes more shallow and fragmented. Thus, complex interactions of circadian and sleep effects result in a consistent and predictable pattern of changes of the setpoint of glucose regulation over the 24-hour cycle.

FIGURE 6-11. Twenty-four-hour pattern of blood glucose changes in response to oral glucose (top panel; 50 g glucose every 3 hours), identical meals, constant glucose, and continuous enteral nutrition in normal young adults. At each time point, the mean glucose level is shown with the standard error of the mean. (Reproduced from Van Cauter E, Polon-sky KS, Scheen AJ. Roles of circadian rhythmicity and sleep in human glucose regulation. Endocr Rev 1997; 18:716.)

Human insulin secretion is a complex oscillatory process including rapid pulses recurring every 10 to 15 minutes superimposed on slower, ultradian oscillations with periods in the 90- to 120-minute range.26 The ultradian oscillations are tightly coupled to glucose, and the periodicity of the insulin secretory oscillations can be entrained to the period of an oscillatory glucose infusion, supporting the concept that these ultradian oscillations are generated by the glucose-insulin feedback mechanism.69 Stimulatory effects of sleep on insulin secretion are mediated by an increase in the amplitude of the oscillation.70 The rapid 10- to 15-minute pulsations seem to have a different origin from that of the ultradian oscillations.26 Indeed, they may appear independently of glucose, because they have been observed in the isolated perfused pancreas and in isolated islets. Thus, the existence of an intrapancreatic pacemaker generating rapid oscillations has been postulated. In obese and diabetic subjects, the diurnal and ultradian variations in glucose regulation are abnormal. In obesity, the morning versus evening difference in glucose tolerance observed in normal subjects is abolished. In type 1 diabetic patients, the increase in glucose levels and/or insulin requirements, which occurs in a prebreakfast period ranging from 5:00 a.m. to 9:00 a.m., has been called the dawn phenomenon.71 A role for nocturnal GH secretion in the pathogenesis of the dawn phenomenon has been demonstrated in some, but not all, studies. The observation of a dawn phenomenon in type 2 diabetes patients under normal dietary conditions has been less consistent. Prominent diurnal variations in glucose levels and insulin secretion in both normal subjects and diabetic patients become apparent during prolonged fasting.72 Figure 6-12 illustrates these variations in diabetic patients and age-, sex-, and weight-matched controls studied during a 24-hour fast following an overnight fast.72 As expected, glucose levels initially declined as a result of the fasting condition, but started rising again in the late evening to reach a morning maximum. The nocturnal rise of glucose during prolonged fasting could represent a normal diurnal variation in the set-point of glucose regulation amplified by counterregulatory mechanisms activated by the fasting condition.

FIGURE 6-12. Individual 24-hour glucose profiles from four subjects with type 2 diabetes who remained fasted throughout the study period. The two upper profiles were obtained from 8 a.m. the first day until 8 a.m. the next day. The two lower profiles were obtained from 2 p.m. the first day until 2 p.m. the next day. Irrespective of the timing of the beginning of the fasting period, glucose concentrations started increasing in the evening and peaked in the morning. The dashed lines represent the best-fit curve. (From Shapiro ET, Polonsky KS, Copinschi G, et al. Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. J Clin Endocrinol Metab 1991; 72:444.)

The rapid and ultradian oscillations of insulin secretion are perturbed in type 2 diabetes and in impaired glucose tolerance without hyperglycemia. The rapid pulses appear to be less regular and of shorter duration than in normal subjects.73,74 A less regular oscillatory pattern may already be detected in relatives of patients with type 2 diabetes. The ultradian oscillations, which have an exaggerated amplitude in obese subjects without apparent changes in frequency or pattern of recurrence, are irregular and of lower amplitude in subjects with established type 2 diabetes.75,76 and 77 Disturbances in the pattern of entrainment of insulin secretion to oscillatory glucose infusions are evident in type 2 diabetes patients, in nondiabetic subjects with impaired glucose tolerance,77 and in nondiabetic first-degree relatives of subjects with type 2 diabetes.26 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6.

Turek FW. Circadian rhythm. Horm Res 1998; 49:109. Borbely AA. Processes underlying sleep regulation. Horm Res 1998; 49:114. Vitaterna MH, King DP, Chang AM, et al. Mutagenesis and mapping of a mouse gene, Clock, essential for circadian behavior. Science 1994; 264:719. King DP, Zhao Y, Sangoram AM, et al. Positional cloning of the mouse circadian Clock gene. Cell 1997; 89:641. Dunlap JC. Molecular bases for circadian clocks. Cell 1999; 96:271. Porkka-Heiskanen T, Strecker RE, Thakkar M, et al. Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science 1997; 276:1255.

6a. Spiegel K, Leproult R, Van Cauter E. Impact of sleep debt on metabolic and endocrine function. Lancet 1999; 9:1435. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17.

Reinberg AE. Concepts in chronopharmacology. Annu Rev Pharmacol Toxicol 1992; 32:51. Czeisler CA, Johnson MP, Duffy JF, et al. Exposure to bright light and darkness to treat physiologic maladaptation to night work. N Engl J Med 1990; 322:1253. Tepas DI, Carvalhais AB. Sleep patterns of shiftworkers. Occup Med 1990; 5:199. Van Reeth O. Sleep and circadian disturbances in shift work: strategies for their management. Horm Res 1998; 49:158. Roden M, Koller M, Pirich K, et al. The circadian melatonin and cortisol secretion pattern in permanent night shift workers. Am J Physiol 1993; 265:R261. Folkard S, Minors DS, Waterhouse JM. Chronobiology and shift work: current issues and trends. Chronobiologia 1985; 12:31. Weibel L, Follenius M, Spiegel K, et al. Comparative effect of night and daytime sleep on the 24-hour cortisol secretory profile. Sleep 1995; 18:549. Weibel L, Spiegel K, Follenius M, et al. Internal dissociation of the circadian markers of the cortisol rhythm in night workers. Am J Physiol 1996; 270:E608. Spiegel K, Weibel L, Gronfier C, et al. Twenty-four hour prolactin profiles in night workers. Chronobiol Int 1996; 13:283. Wehr TA, Goodwin FK. Biological rhythms in manic-depressive illness. In: Wehr TA, Goodwin FK, eds. Circadian rhythms in psychiatry. Pacific Grove, CA: Boxwood Press, 1983:129. Van Cauter E, Spiegel K. Circadian and sleep control of hormonal secretions. In: Turek FW, Zee PC, eds. Regulation of sleep and circadian rhythms. New York: Marcel Dekker, Inc., 1999:397.

17a. Saini J, Krieger J, Brandenberger G, et al. Continuous positive airway pressure treatment: effects on growth hormone, insulin and glucose profiles in obstructive sleep apnea patients. Horm Metab Res 1993; 25:375. 18. Grunstein RR, Handelsman DJ, Lawrence SJ, et al. Neuroendocrine dysfunction in sleep apnea: reversal by continuous positive airways pressure therapy. J Clin Endocrinol Metab 1989; 68:352. 19. Goldstein SJ, Wu RHK, Thorpy MJ, et al. Reversibility of deficient sleep entrained growth hormone secretion in a boy with achondroplasia and obstructive sleep apnea. Acta Endocrinol 1987; 116:95. 20. Czeisler CA, Chiasera AJ, Duffy JF. Research on sleep, circadian rhythms and aging: applications to manned spaceflight. Exp Gerontol 1991; 26:217. 21. van Coevorden A, Mockel J, Laurent E, et al. Neuroendocrine rhythms and sleep in aging. Am J Physiol 1991; 260:E651. 22. Chapotot F, Gronfier C, Jouny C, et al. Cortisol secretion is related to electroencephalographic alertness in human subjects during daytime wakefulness. J Clin Endocrinol Metab 1998; 83:4263. 23. Hotchkiss J, Knobil E. The menstrual cycle and its neuroendocrine control. In: Knobil E, Neil JD, eds. The physiology of reproduction, 2nd ed. New York: Raven Press, 1994:711. 24. Wetsel WC, Valenca MM, Merchenthaler I, et al. Intrinsic pulsatile secretory activity of immortalized luteinizing hormone-releasing hormone-secreting neurons. Proc Natl Acad Sci U S A 1992; 89:4149. 25. Simon C, Brandenberger G, Follenius M. Ultradian oscillations of plasma glucose, insulin, and C-peptide in man during continuous enteral nutrition. J Clin Endocrinol Metab 1987; 64:669. 26. Polonsky KS, Sturis J, Van Cauter E. Temporal profiles and clinical significance of pulsatile insulin secretion. Horm Res 1998; 49:178. 27. Conn PM, Crowley WF. Gonadotropin-releasing hormone and its analogues. N Engl J Med 1991; 324:93. 28. Van Cauter E. Computer-assisted analysis of endocrine rhythms. In: Rod-bard D, Forti G, eds. Computers in endocrinology. New York: Raven Press, 1990:59. 29. Pincus SM, Keefe DL. Quantification of hormone pulsatility via an approximate entropy algorithm. Am J Physiol 1992; 262:E741. 30. Pincus SM, Goldberger AL. Physiological time-series analysis: what does regularity quantify? Am J Physiol 1994; 266:H1643. 31. Van Cauter E, Plat L, Copinschi G. Interrelations between sleep and the somatotropic axis. Sleep 1998; 21:553. 32. Pincus SM, Gevers EF, Robinson IC, et al. Females secrete growth hormone with more process irregularity than males in both humans and rats. Am J Physiol Physiol 1996; 270:E107. 33. Ho KY, Evans WS, Blizzard RM, et al. Effects of sex and age on the 24-hour profile of growth hormone secretion in man: importance of endogenous estradiol concentrations. J Clin Endocrinol Metab 1987; 64:51. 34. Jaffe C, Turgeon D, DeMott Friberg R, et al. Nocturnal augmentation of growth hormone (GH) secretion is preserved during repetitive bolus administration of GH-releasing hormone: potential involvement of endogenous somatostatin: a clinical research center study. J Clin Endocrinol Metab 1995; 80:3321. 35. Ocampo-Lim B, Guo W, DeMott Friberg R, et al. Nocturnal growth hormone (GH) secretion is eliminated by infusion of GH-releasing hormone antagonist. J Clin Endocrinol Metab 1996; 81:4396. 36. Gronfier C, Luthringer R, Follenius M, et al. A quantitative evaluation of the relationships between growth hormone secretion and delta wave electroencephalographic activity during normal sleep and after enrichment in delta waves. Sleep 1996; 19:817. 37. Van Cauter E, Plat L, Scharf M, et al. Simultaneous stimulation of slow-wave sleep and growth hormone secretion by g-hydroxybutyrate in normal young men. J Clin Invest 1997; 100:745. 38. Hartman ML, Veldhuis JD, Johnson ML, et al. Augmented growth hormone (GH) secretory burst frequency and amplitude mediate enhanced GH secretion during a two-day fast in normal men. J Clin Endocrinol Metab 1992; 74:757. 39. Edge JA, Dunger DB, Matthews DR, et al. Increased overnight growth hormone concentrations in diabetic compared with normal adolescents. J Clin Endocrinol Metab 1990; 71:1356. 40. Gelato MC, Oldfield E, Loriaux DL, Merriam GR. Pulsatile growth hormone secretion in patients with acromegaly and normal men: the effects of growth hormone-releasing hormone infusion. J Clin Endocrinol Metab 1990; 71:585. 41. Hartman ML, Veldhuis JD, Vance ML, et al. Somatotropin pulse frequency and basal concentrations are increased in acromegaly and are reduced by successful therapy. J Clin Endocrinol Metab 1990; 70:1375. 42. Weitzman ED, Zimmerman JC, Czeisler CA, Ronda JM. Cortisol secretion is inhibited during sleep in normal man. J Clin Endocrinol Metab 1983; 56:352. 43. Born J, Muth S, Fehm HL. The significance of sleep onset and slow wave sleep for nocturnal release of growth hormone (GH) and cortisol. Psycho-neuroendocrinology 1988; 13:233. 44. Van Cauter E, Blackman JD, Roland D, et al. Modulation of glucose regulation and insulin secretion by circadian rhythmicity and sleep. J Clin Invest 1991; 88:934. 45. Follenius M, Brandenberger G, Bardasept J, et al. Nocturnal cortisol release in relation to sleep structure. Sleep 1992; 15:21. 46. Spath-Schwalbe E, Gofferje M, Kern W, et al. Sleep disruption alters nocturnal ACTH and cortisol secretory patterns. Biol Psychiatry 1991; 29:575. 47. Pruessner JC, Wolf OT, Hellhammer DH, et al. Free cortisol levels after awakening: a reliable biological marker for the assessment of adrenocortical activity. Life Sci 1997; 61:2539. 48. Leproult R, Copinschi G, Buxton O, Van Cauter E. Sleep loss results in an elevation of cortisol levels the next evening. Sleep 1997; 20:865. 49. Lejeune-Lenain C, Van Cauter E, Desir D, et al. Control of circadian and episodic variations of adrenal androgens secretion in man. J Endocrinol Invest 1987; 10:267. 50. Van Cauter E, Turek FW. Endocrine and other biological rhythms. In: DeGroot LJ, ed. Endocrinology. Philadelphia: WB Saunders, 1995:2487. 51. Yehuda R, Teicher MH, Trestman RL, et al. Cortisol regulation in posttraumatic stress disorder and major depression: a chronobiological analysis. Biol Psychiatry 1996; 40:79. 52. MacHale SM, Cavanaugh JTO, Bennie J, et al. Diurnal variation of adreno-cortical activity in chronic fatigue syndrome. Neuropsychobiology 1998; 38:213. 53. Spiegel K, Follenius M, Simon C, et al. Prolactin secretion and sleep. Sleep 1994; 17:20. 54. Waldstreicher J, Duffy JF, Brown EN, et al. Gender differences in the temporal organization of prolactin (PRL) secretion: evidence for a sleep-independent circadian rhythm of circulating PRL levels: a clinical research center study. J Clin Endocrinol Metab 1996; 81:1483. 55. Spiegel K, Luthringer R, Follenius M, et al. Temporal relationship between prolactin secretion and slow-wave electroencephalographic activity during sleep. Sleep 1995; 18:543. 56. Copinschi G, Van Onderbergen A, L'Hermite-Balériaux M, et al. Effects of the short-acting benzodiazepine triazolam, taken at bedtime, on circadian and sleep-related hormonal profiles in normal men. Sleep 1990; 13:232. 57. Iranmanesh A, Veldhuis JD, Carlsen EC, et al. Attenuated pulsatile release of prolactin in men with insulin-dependent diabetes mellitus. J Clin Endo-crinol Metab 1990; 71:73. 58. Seki K, Uesato T, Kato K, Shima K. Twenty-four hour secretory pattern of prolactin in hyperprolactinaemic patients with pituitary micro- and macroadenomas. Acta Endocrinol 1984; 106:433. 59. Hayes FJ, Crowley WFJ. Gonadotropin pulsations across development. Horm Res 1998; 49:163. 60. Filicori M, Tabarelli C, Casadio P, et al. Interaction between menstrual cyclicity and gonadotropin pulsatility. Horm Res 1998; 49:169. 61. Behrends J, Prank K, Dogu E, Brabant G. Central nervous system control of thyrotropin secretion during sleep and wakefulness. Horm Res 1998; 49:173. 62. Goichot B, Brandenberger G, Saini J, et al. Nocturnal plasma thyrotropin variations are related to slow-wave sleep. J Sleep Res 1992; 1:186.

63. Hirschfeld U, Moreno-Reyes R, Akseki E, et al. Progressive elevation of plasma thyrotropin during adaptation to simulated jet lag: effects of treatment with bright light or zolpidem. J Clin Endocrinol Metab 1996; 81:3270. 64. van Coevorden A, Laurent E, Decoster C, et al. Decreased basal and stimulated thyrotropin secretion in healthy elderly men. J Clin Endocrinol Metab 1989; 69:177. 65. Romijn JA, Adriaanse R, Brabant G, et al. Pulsatile secretion of thyrotropin during fasting: a decrease of thyrotropin pulse amplitude. J Clin Endo-crinol Metab 1990; 70:1631. 66. Bartalena L, Cossu E, Grasso L, et al. Relationship between nocturnal serum thyrotropin peak and metabolic control in diabetic patients. J Clin Endocrinol Metab 1993; 76:983. 67. Van Cauter E, Polonsky KS, Scheen AJ. Roles of circadian rhythmicity and sleep in human glucose regulation. Endocr Rev 1997; 18:716. 68. Plat L, Byrne MM, Sturis J, et al. Effects of morning cortisol elevation on insulin secretion and glucose regulation in humans. Am J Physiol 1996; 270:E36. 69. Sturis J, Polonsky KS, Mosekilde E, Van Cauter E. Computer model for mechanisms underlying ultradian oscillations of insulin and glucose. Am J Physiol 1991; 260:E801. 70. Simon C. Ultradian pulsatility of plasma glucose and insulin secretion rates: circadian and sleep modulation. Horm Res 1998; 49:185. 71. Bolli GB, Gerich JE. The “dawn phenomenon”: a common occurrence in both non-insulin-dependent and insulin-dependent diabetes mellitus. N Engl J Med 1984; 310:746. 72. Shapiro ET, Polonsky KS, Copinschi G, et al. Nocturnal elevation of glucose levels during fasting in noninsulin-dependent diabetes. J Clin Endo-crinol Metab 1991; 72:444. 73. Lang DA, Matthews DR, Burnett M, Turner RC. Brief, irregular oscillations of basal plasma insulin and glucose concentrations in diabetic man. Diabetes 1981; 30:435. 74. O'Rahilly S, Turner R, Matthews D. Impaired pulsatile secretion of insulin in relatives of patients with non-insulin-dependent diabetes. N Engl J Med 1988; 318:1225. 75. Polonsky KS, Given BD, Hirsch LJ, et al. Abnormal patterns of insulin secretion in non-insulin-dependent diabetes mellitus. N Engl J Med 1988; 318:1231. 76. Simon C, Brandenberger G, Follenius M, Schlienger JL. Alteration in the temporal organisation of insulin secretion in Type 2 (non-insulin-dependent) diabetic patients under continuous enteral nutrition. Diabetologia 1991; 34:435. 77. O'Meara NM, Sturis J, Van Cauter E, Polonsky KS. Lack of control by glucose of ultradian insulin secretory oscillations in impaired glucose tolerance and in non-insulin-dependent diabetes mellitus. J Clin Invest 1993; 92:262.

CHAPTER 7 GROWTH AND DEVELOPMENT IN THE NORMAL INFANT AND CHILD Principles and Practice of Endocrinology and Metabolism

CHAPTER 7 GROWTH AND DEVELOPMENT IN THE NORMAL INFANT AND CHILD GILBERT P. AUGUST Height and Weight Growth Velocity Skinfold Thickness Body Mass Index Bone Age Height Prediction Body Proportions Genital Development Developmental Endocrinology Growth Hormone and Insulin-Like Growth Factors Adrenocorticotropic Hormone and Adrenocortical Steroids Thyroid-Stimulating Hormone and Thyroid Hormones Luteinizing Hormone, Follicle-Stimulating Hormone, and the Sex Steroids Renin and Aldosterone Chapter References

The pediatric population is composed of continually changing individuals. Thus, a knowledge of normal developmental changes is required for the clinician to recognize deviant growth and development and abnormalities in the hormonal milieu.

HEIGHT AND WEIGHT Height and weight standards derived from one ethnic group cannot always be applied to other ethnic groups. Furthermore, use of current standards is important, because growth data obtained from a previous generation may not apply to the present generation. The National Health Survey collected growth and anthropometric data on American children from 1963 to 1974 (Table 7-1).1,2,3,4,5,6,7,8 and 9 These data provide standards that can be correlated with sex, race, and socioeconomic status. The growth charts distributed by pharmaceutical companies are derived from these National Health Survey data. Because the growth standards of American white and black children do not differ significantly, a single growth standard can be used. Significant differences do exist, however, in the growth standards of American children of Asian descent.7

TABLE 7-1. Height and Weight of Children, Body Proportions, and Error of Prediction of Adult Height (United States)

Figure 7-1, Figure 7-2, Figure 7-3 and Figure 7-4 show the current growth data from the National Health Survey. In Table 7-1, the mean heights, mean weights, and standard deviations (SDs) for children 2 to 18 years of age are detailed. The standard deviation is useful in evaluating extreme deviations in growth (e.g., heights and weights below the standard curves).10 Children whose heights or weights are below the standard curves constitute a significant proportion of the population, because the commonly used growth charts provide data from only the 5th through the 95th percentiles.

FIGURE 7-1. Length and weight of boys. Birth to 36 months of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May 30, 2000.

FIGURE 7-2. Height and weight of boys. 2 to 20 years of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May 30, 2000.

FIGURE 7-3. Length and weight of girls. Birth to 36 months of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May30, 2000.

FIGURE 7-4. Height and weight of girls. 2 to 20 years of age. Centers for Disease Control and Prevention, National Center for Health Statistics. CDC growth charts: United States. Full size charts are available on the internet at http://www.cdc.gov/nchs/about/major/nhanes/growthcharts/charts.htm May 30, 2000.

For children aged 2 to 18 years, growth curves for as low as 2 SD below the mean are available. These curves are more in keeping with the usual definition of “normal” as comprising 95% of the population. Only 2.5% of children would be considered as “short” by these standards (Fig. 7-1, Fig. 7-2, Fig. 7-3 and Fig. 7-4). The National Center for Health Statistics has revised the current growth charts to reflect a more contemporaneous standard. The new charts contain data from the 3rd to 97th percentiles. The new growth charts are available on the internet at http://www.cdc.gov/growthcharts. Socioeconomic status was found to play a small but significant adverse role in the growth of children with very low family incomes.6 Beyond age 2 to 3 years, children grow throughout childhood until puberty along a particular height percentile channel that is determined by genetic factors. The normal pubertal growth spurt is reflected by an upward shift in height percentile. As growth decelerates later in puberty with impending epiphyseal closure, a shift downward in height percentile occurs, so that adult height more closely approximates the prepubertal percentile ranking. The reason for these shifts lies in the cross-sectional character of the standard growth charts. Cross-sectional data mask normal longitudinal growth patterns. The growth charts derived from studies in Britain have superim-posed longitudinal growth lines that represent temporal variations in the onset of puberty.11 Similar data are available for North American children.12 A shift in height percentile can also occur during infancy when growth velocities are adjusted in children whose birth lengths are either “too long” or “too short” in comparison with their parents' heights. The shift upward in percentile generally takes place within the first 6 months of life and is usually completed by 12 months of age. The shift downward in percentile begins later, generally after 3 months of age, and is completed by 18 months of age.13 The influence of midparental stature on the interpretation of a child's growth status is considered to be of such importance by some workers that conversion graphs and adjustment formulas have been developed. Tanner and coworkers14 developed such curves for British children, and adjustment tables are now available for American children.15 Parental heights can also be used to predict target adult height.16 This can be valuable when the effects of a treatment are analyzed. The standard growth charts derived from the National Health Survey data are not as applicable for tracking the growth of premature infants or children small for gestational age. However, growth standards are available for such children through 4 years of age.17,18 These growth charts must be adjusted for the degree of low birth weight.19 Further, infants small for gestational age may not truly “catch up” when their heights are compared with those of their siblings.20,21 Whenever sufficient data exist to construct growth charts for children with constitutional diseases, those standards should be used for comparison of current growth status and for expected adult height. Such growth standards are available for children with achondroplasia, Down syndrome, diastrophic dysplasia, Duchenne muscular dystrophy, Klinefelter syndrome, Noonan syndrome, pseudoachondroplasia, Russell-Silver syndrome, spondyloepiphyseal dysplasia, and William syndrome.22 Similar information is available for children with gonadal dysgenesis.23,24 Careful attention to detail is important to produce accurate height measurements that can be used to track a child's growth (Fig. 7-5). Length should be measured on a horizontal board with the head held firmly against the upright at one end and the length determined by sliding the movable upright against the child's heels. Length should be determined until 3 years of age. The growth curve from birth until 3 years of age is standardized on length. Height must be measured using a wall-mounted instrument or tape. This permits the child to stand fully upright. The feet should be together, the back and heels should be pressed against the wall or vertical part of the measuring instrument, and the head should be held in the Frankfort horizontal plane with upward pressure exerted under the child's mastoid process to hold the correct position and to measure maximal height. The measurement of height is taken by sliding a right-angle block downward until it touches the child's head; the height can then be read from the vertically mounted rule. Accurate, reproducible height measurements cannot be obtained from the commonly used height-measuring devices attached to weighing scales. The importance of adhering to the correct technique is documented in the National Health Survey data.1,2,3,4,5 and 6

FIGURE 7-5. Technique of accurate measurement of standing height.

GROWTH VELOCITY The standard growth charts with their height percentiles are growth attainment charts, which represent how much height the child has attained by a particular age. A concept that is often more useful, especially when evaluating the longitudinal growth of an individual child, is growth velocity, which depicts growth during a given

period. The difference in height at the beginning and end of a given period of time is annualized and then expressed as centimeters per year.11,12 The X-axis on a growth velocity chart is the chronologic age; the Y-axis is the growth velocity in centimeters per year. These charts take into consideration the variability in height velocity caused by the normal variability in age of onset of puberty. An advantage of height velocity charts is their ability to demonstrate more quickly an aberration of growth or the effects of treatment. Short-term growth velocity data, however, may be subject to error because of marked seasonal variation; children generally grow faster in the spring through early summer.24 The growth velocity charts clearly demonstrate that a child's growth rate is most rapid during infancy and puberty and is relatively stable during the elementary school years. Although these charts present height velocity percentiles ranging from the 3rd to 97th percentiles, these may not represent the biologic normal range. Longitudinal growth that is consistently below the 10th percentile may not be sufficient to maintain a child in a single height attainment percentile; instead, that child may demonstrate a downward shift in height percentile.25 SKINFOLD THICKNESS Skinfold thickness is used as a measure of total body fat.26 In research studies, more than one site needs to be measured, but for clinical use, the more readily accessible triceps skin-fold is commonly used. The National Health Survey used a carefully standardized technique to measure the triceps skin-fold. The Lange caliper (Fig. 7-6) was used on a skinfold parallel to the long axis of the arm over the triceps muscle halfway between the elbow and acromial process of the scapula; care was taken to apply the caliper so that the pressure plates remained parallel to each other. Table 7-2 presents the normal values.27,28 and 29

FIGURE 7-6. Measurement of skinfold thickness with the Lange caliper.

TABLE 7-2. Percentiles for Triceps Skinfold Thickness (in Millimeters) by Year of Age27,28

BODY MASS INDEX Body mass index (BMI) is another convenient measure of body fat. It is derived by calculating the weight in kilograms divided by the the height in meters squared. Population ranges for American children are now available according to ethnic group for ages 5 to 17 years.30 The 15th, 50th, and 85th percentiles for each ethnic group appear in Table 7-3. The BMI will assume greater applicability once the new BMI curves are available from the National Center for Health Statistics. For adults, grade 1 obesity is defined as a BMI of over 25 and grade 2 obesity as a BMI over 30. Similar cutoff levels have been suggested for late adolescence. These BMI values are close to the 80th and 95th percentiles, respectively, on the National Center for Health Statistics charts.31

TABLE 7-3. Body Mass Index Centiles for Children 5–17 Years of Age

BONE AGE The concept of bone age is based on the skeletal changes that occur with the physical growth and maturation of the child. These skeletal changes include the calcification, growth, and shaping of the epiphyseal centers of the bones and their eventual fusion with the diaphyses. The fact that these changes occur in a regular sequence in the different bones as determined by the radiographic examination of normally developing children permits one to estimate a bone age for a particular child. Any of a number of techniques can be used, as well as any part of the skeleton. The most popular method in the United States compares a single anteroposterior radiograph of the left hand and wrist with the series of standard films in the atlas compiled by Greulich and Pyle.32 In Europe, the Tanner method is used to calculate the bone age of the left hand and wrist by using maturity indicators for each bone, so that a composite score is derived.33 The Tanner-Whitehouse method has been standardized for American children aged 8 to 16 years.34 As with any other laboratory test, there is a mean population bone age that corresponds to the particular best fit in the Greulich and Pyle atlas and a standard deviation or range of expected values. The normal range of expected bone ages for a child younger than 1 year of age is ±3 to 6 months; for a child of 6 to 11 years of age, the range is ±2 years; and for a child of 12 to 14 years of age, the range is ±2 years.35 The National Health Survey compared the bone ages of contemporary children with those derived from the original study of Greulich and Pyle between 1917 and 1942. The original study population was composed of upper-middle-class children in Cleveland. In children aged 6 to 11 years, good congruence was seen except that

contemporary 10- and 11-year-old children showed a significant retardation in bone age of 0.2 to 0.65 years.36 The assumption that bone age advances 1 year for each calendar year may not be correct; bone age advances more rapidly than chronologic age during the onset of puberty and during peak height velocity of puberty. At these times, the average advancement in bone age over chronologic age is 1.73 years, with a range of 0.8 to 2.8 years.37 Bone age advancement should not be attributed falsely to therapeutic intervention. These results serve as another example of the difference between normal longitudinal growth of a single child and the cross-sectional standards that are usually available for tracking a child's growth and development.37 The determination of bone age can be useful if one remains aware of its limitations. The particular standard used must be verified for the population being studied. The person interpreting the bone age radiograph must be consistent. Separate male and female standards exist, and there is a range of normal (see Chap. 18 and Chap. 217). HEIGHT PREDICTION One of the most common applications of a bone age determination is the prediction of eventual adult height. Bayley and Pinneau38 published prediction tables based on the percentage of adult height attained at various bone ages. The tables were further refined by separating the children whose bone ages were>1 year advanced or retarded compared with chronologic age. Separate prediction tables are provided for these children as well as for males and females. More mathematically refined systems have been proposed that use regression formulas with the addition of variables such as weight, midparental stature, growth velocity, and menarcheal status. The Roche-Wainer-Thissen (RWT) method uses the child's recumbent length rather than standing height and is standardized for American children.39 The Tanner-Whitehouse system (TW2 Mark 2) is standardized for British children.33 It includes a larger number of children at the extremes of the standardizing group and thus tries to counter one criticism of height prediction methods—that they are most accurate for the prediction of the final adult height of normal children whose bone ages are not significantly retarded. In addition, the TW2 Mark 2 system no longer takes into account midparental stature in the prediction regression formulas, as did the previous formulas. Regardless of which system is used for the prediction of adult height, the prediction always has an error. For the RWT method, the ±90% confidence range for a prediction is given in Table 7-1. For the TW2 Mark 2 system, the error is stated as a residual standard deviation. For comparison purposes, in Table 7-1, the residual standard deviation has been multiplied by 1.645 to provide the 90% confidence range for a predicted adult height. To obtain a 95% confidence range, the residual standard deviation is multiplied by ±1.96. For normal children, these two methods are comparable and appear more accurate than the older Bailey and Pinneau tables39; the stated error in the latter method is 5.1 cm for ±2 SD. In a comparative study of the three methods that examined Finnish children, the RWT method was slightly more accurate.40,41 The important issue is not how well the various height prediction systems work in normal children but rather how well they work in children with abnormal growth patterns. In children with familial tall stature, the Tanner method was slightly more accurate than the other two methods.41 Both the RWT and Tanner methods, however, overestimated adult height in children with precocious puberty or gonadal dysgenesis; the Bailey and Pinneau tables were more accurate in these conditions. The overestimation for children with precocious puberty ranged from 30 cm at 5 years of age to just 5 cm at 13 years of age. The overestimation of adult height was 10 cm at the earlier ages in children with gonadal dysgenesis. The greater accuracy of height prediction with increasing age of the patient also holds true for children with constitutional tall stature.42 In addition, the RWT and Tanner methods both overestimated the adult heights of children with primordial short stature.41 These studies used the older Tanner methods. Tanner and coworkers33 point out several limitations to the accuracy of their new formulas that should apply to other height prediction systems. Although the formulas take into account several variables, some unpredictability still exists, as reflected in the residual standard deviation. Tanner and co-workers33 point out that no particular reason exists that the equations should be valid in cases of precocious puberty or achondroplasia. In part, the error in height prediction, measured by whatever means, may be due to the variability in total height gained during puberty. In a review43 of four studies of American children, from the onset of puberty, girls had an average gain in height of 25.4 to 31 cm, and boys had an average gain in height of 28.2 to 31 cm. The standard deviations ranged from 4 to 6.8 cm for girls and 4.3 to 5.5 cm for boys. Thus, although considerable height was gained during puberty, considerable variability was seen as well. Similar results were reported in European studies.43 Confounding this further is the inverse relationship between height obtained during puberty and the age of onset of puberty.44 This phenomenon may be related to the shortened duration of puberty, at least in girls, with the later onset of puberty.45 Finally, an additional caution arises from a study that compared the final adult height achieved by short children with normal bone ages and by short children with significantly retarded bone ages with predictions of the children's height using the Bailey and Pinneau tables.46 Although only 1 of 17 children with normal bone age had an overpredicted adult height, 5 of 10 children with a retarded bone age had an over-prediction of their adult height—they failed to reach an adult height within 5.1 cm of the predicted height.46 This is the group of children for whom the physician is most often called on to use height prediction tables and provide reassurance to the patient and family. A comparison of predicted adult height, based on bone age and current height, with the predicted genetic target adult height based on parental heights is often useful.16 This is of importance when determining if the child's current height is inconsistent with genetic potential or when one is evaluating the effects of a therapeutic intervention. BODY PROPORTIONS Significant racial differences exist in regard to body proportions. The sitting/standing height ratio was determined during the National Health Survey.4,5 These data are presented in Table 7-1. Careful attention to detail is required. Height measurement using a wall-mounted instrument is required. Sitting height is measured on a sitting height table, with the child sitting erect in a standardized manner and with the head held in the Frankfort plane4,5 (Fig. 7-7).

FIGURE 7-7. Technique of accurate measurement of sitting height.

GENITAL DEVELOPMENT Table 7-4 presents the normal standards for penile and testicular size.47,48 and 49 Penile length is a stretched length; traction is applied to the penis and the rule is applied firmly to the root of the penis. Too often, the suprapubic fat pad is not compressed enough so that one obtains a penile length that is artifactually too short (see Chap. 93).

TABLE 7-4. Penile and Testicular Size During Childhood

DEVELOPMENTAL ENDOCRINOLOGY The pediatric patient undergoes continuous change that is reflected not only in growth rates but also in the secretory activity of the endocrine system. The perinatal period is one of rapid change in the endocrine system when the child is adjusting to extrauterine life (Table 7-6). This is further complicated by the differences that may exist in hormonal levels in the full-term versus preterm infant. This section outlines some of these hormonal variations. As a general rule, one can divide the analysis of hormonal secretory activity into the following developmental periods: fetal, perinatal, childhood, and pubertal.50,51

TABLE 7-6. Blood Concentrations of Gonadotropins, Gonadal Steroids, Renin, and Aldosterone in Infancy and Childhood

GROWTH HORMONE AND INSULIN-LIKE GROWTH FACTORS Immunoreactive growth hormone is found in human anterior pituitary glands as early as the seventh week of gestation.49 Pituitary growth hormone increases from 0.44µg per gland at 10 to 14 weeks of gestation to a mean of 675µg per gland at term. However, pituitary growth hormone level is fairly constant after 25 weeks of gestation. Growth hormone is also detectable in fetal serum in high concentration. These high concentrations of growth hormone may be a reflection of hypothalamic-pituitary neuroendocrine immaturity. At 10 to 14 weeks of gestation, the mean growth hormone level in fetal serum is 65 ng/mL, and it rises rapidly to a peak fetal concentration of 132 ng/mL at 20 to 25 weeks of gestation. Thereafter, serum growth hormone declines to 35 ng/mL at 35 to 40 weeks of gestation. The postulated immature neuroendocrine control of growth hormone secretion persists into the neonatal period. Growth hormone is not suppressed by glucose until 1 month of age, and the normal sleep-induced rise in serum growth hormone levels does not appear until 3 months of age.51 Insulin-like growth factor-I (IGF-I) plasma concentrations show a marked age-related pattern. The normal ranges, as reported by the Quest Diagnostic Nichols Institute, are shown in Figure 7-8 and Figure 7-9. The range is large, and a marked overlap of values is seen when one compares the lower range of normal with values typically associated with growth hormone deficiency. Plasma IGF-I levels increase with age; during puberty concentrations are reached that would be associated with acromegaly if encountered in an adult. After puberty, the levels decline to the normal adult range. Females generally have IGF-I levels ~20% higher than those of males. The normal level of IGF-I may correlate better with the child's physiologic development. Such a relation is seen during puberty (see Fig. 7-9). For prepubertal boys, a similar relation with bone age may be valid as well.52,53 Insulin-like growth factor–binding protein-3 (IGFBP-3) has been suggested as another serum protein reflective of growth hormone secretion and function.54 In a manner similar to IGF-I, normal values for IGFBP-3 vary according to age and stage of puberty (Fig. 7-10).

FIGURE 7-8. Variation of insulin-like growth factor-I with age and sex. (Data derived from the normal values provided by Quest Diagnostic Nichols Institute, San Juan Capistrano, California.)

FIGURE 7-9. Variation of insulin-like growth factor-I with stage of puberty and sex. (Data derived from the normal values provided by Quest Diagnostic Nichols Institute, San Juan Capistrano, California.)

FIGURE 7-10. Variation of insulin-like growth factor–binding protein-3 with age. (Data derived from the normal values provided by Quest Diagnostic Nichols Institute, San Juan Capistrano, California.)

Measurement of insulin-like growth factor–binding protein-2 (IGFBP-2) has been proposed as another aid in the diagnosis of growth hormone deficiency55 (see Table 7-5). Interestingly, IGFBP-2 is increased in growth hormone deficiency, rather than being decreased as is the case with both IGF-I and IGFBP-3.

TABLE 7-5. Normal Values for Insulin-Like Binding Protein-2 (IGFBP-2)

ADRENOCORTICOTROPIC HORMONE AND ADRENOCORTICAL STEROIDS Adrenocorticotropic hormone (ACTH) is detectable in fetal serum as early as 10 to 14 weeks of gestation. The concentration of ACTH declines toward term, falling from 249 pg/mL ± 65 SE (standard error) to 143 pg/mL ± 9 SE at term. By 1 week of age, the serum ACTH levels are similar to those found in older children. The pattern of adrenal corticosteroids secreted varies from those found in older children. This is due to the presence of a fetal zone in the adrenal gland that constitutes 70% to 85% of the total weight of the adrenal gland at term. In both full-term and premature infants, the fetal zone undergoes rapid involution after birth, so that it constitutes just 10% of the adrenal gland by 4 to 5 months of age.56 Plasma concentrations of adrenal steroids are shown in Figure 7-11.

FIGURE 7-11. Variation of selected adrenal steroids with age and onset of puberty. (DHEA, dehydroepiandrosterone.) (Data derived from the normal values provided by Endocrine Sciences, Calabasas Hills, California.)

THYROID-STIMULATING HORMONE AND THYROID HORMONES As with the previously discussed hormone systems, thyroid function varies greatly depending on gestational age and perinatal circumstances (see Chap. 47). Thyroid-stimulating hormone (TSH) is detected in fetal serum as early as 10 weeks of gestation; however, the levels are low—only 2.4µU/mL ± 0.14 SE at 10 to 20 weeks of gestation. This increases to 9.6µU/mL ± 0.93 SE at 25 to 30 weeks of gestation. The increase in TSH is accompanied by increases in mean free thyroxine (T4) and total T4 serum concentrations. Free T4 increases from 1.8 ng/dL at 10 to 20 weeks of gestation, to 2.6 ng/dL at 22 weeks, to 3 ng/dL at 30 weeks, and to 4 ng/dL at term. The rise in mean T 4-binding globulin during this period helps to produce an increase in serum T4 levels, which rise from 3µg/dL at 20 weeks of gestation to 9.4µg/dL (range, 5.7–15.6) at 30 weeks of gestation. Without any further rise in T4-binding globulin, the fetal T4 rises to 10.1µg/mL at 35 weeks of gestation (range, 6.1–16.8) and to 10.9µg/dL by term (range, 6.6–18.1).57,58 Between 30 and 40 weeks of gestation, some further maturation seems to take place of the hypothalamic systems responsible for the feedback inhibition of TSH by circulating thyroid hormones. This is suggested by the fall in serum TSH concentration to 8.9µU/mL ± 0.93 SE at term as the serum free T4 rises.59 New second- and third-generation TSH assays have been introduced, and the absolute values listed should be viewed relative to the newer assays. The perinatal period is associated with marked changes in thyroid hormone levels. Within a half hour of birth, the serum TSH level rises to a mean of slightly over 18µU/mL and then rapidly returns to normal within a day. The rise in serum TSH is paralleled within 24 hours by rises in serum T4 and triiodothyronine (T3) levels; T3 levels then decline in a few days, and T4 levels decline in a few weeks. Similar but less marked changes are seen in premature infants.59 The rise in serum T3 is not completely dependent on the rise in TSH but rather is also due to a change in the peripheral conversion of T4 to T3, as opposed to reverse T3, which occurs in the prenatal period59 (see Chap. 15, Chap. 33 and Chap. 47). The changes in thyroid hormone levels during the neonatal period in premature infants is also dependent on the degree of prematurity and the infant's health status.60 LUTEINIZING HORMONE, FOLLICLE-STIMULATING HORMONE, AND THE SEX STEROIDS As early as the 10th week of gestation, luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are found in fetal pituitary glands. Serum LH and FSH levels in the fetus are high, well within the adult castrate range. Higher fetal serum FSH levels are found in females. Toward term, serum LH and FSH levels decline to the low levels typically found at term, but shortly after birth, serum levels of LH and FSH rise again, with LH levels higher in the male and FSH levels higher in the female. Serum LH and FSH levels may not decline to the concentrations deemed typical of childhood for 6 to 12 months (Table 7-6; see also Chap. 16 and Chap. 91). Marked differences appear to exist in the patterns of immunoreactive LH and FSH as compared with bioactive LH and FSH. The ratios also appear to vary with age and stage of development.61,62 and 63 The fetal serum estradiol level does not differ in males and females, but males have higher serum levels of testosterone before the 20th week of gestation,

corresponding to the period of development of the male external genitalia. The serum testosterone level ranges from 100 to 600 ng/dL.64 The fetal serum testosterone level declines toward term, but in cord blood, it remains higher in males than in females. Coincident with the postnatal surge of LH, serum testosterone levels in males rise again to mid-pubertal levels, remaining higher than concentrations thought appropriate for childhood until 4 to 9 months of age.65,66,67 and 68 The normal concentrations of these hormones and some other adrenal hormones are presented in Table 7-6 and Figure 7-11. Although the serum androstenedione level is in the typical prepubertal range by 1 year of age, it begins to rise again before any manifestations of puberty. Androstenedione begins to rise at 8 years of age in boys and at 7 years of age in girls. A similar pattern can be seen for dehydroepiandrosterone and for dehydroepiandrosterone sulfate. The rise in these hormones starts at 7 years of age in boys and at 6 years of age in girls. Similar to other adrenal androgens, dehydroepiandrosterone sulfate also manifests serum concentrations that vary with the child's age. In premature infants, the mean serum concentration is 263µg/dL ± 40.1 SD; in full-term infants, 58.9µg/dL ± 4.5 SD; in male children 1 to 6 years of age, 15.4µg/dL ± 6.8 SD; in female children 1 to 6 years of age, 24.7µg/dL ± 11.1 SD; in male children 6 to 8 years of age, 18.8µg/dL ± 4.1 SD; in female children 6 to 8 years of age, 30.4µg/dL ± 7.6 SD; in male children 8 to 10 years of age, 58.6µg/dL ± 10.1 SD; and in female children 8 to 10 years of age, 117.3µg/dL ± 41.7 SD69 (see Chap. 91 and Chap. 92). RENIN AND ALDOSTERONE The plasma concentrations of renin and aldosterone are markedly elevated in the newborn and decline slowly to accepted normal levels70 (see Table 7-6). CHAPTER REFERENCES 1. National Center for Health Statistics. Height and weight of children, United States. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 104. Public Health Service. Washington: Government Printing Office, September 1970. 2. National Center for Health Statistics. Height and weight of youths 12–17 years, United States. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 124. Public Health Service. Washington: Government Printing Office, January 1973. 3. National Center for Health Statistics. NCHS growth curves for children birth–18 years, United States. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 165. Public Health Service. Washington: Government Printing Office, November 1977. 4. National Center for Health Statistics. Body dimensions and proportions, white and Negro children 6–11 years, United States. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 143. Public Health Service. Washington: Government Printing Office, December 1974. 5. National Center for Health Statistics. Body weight, stature, and sitting height, white and Negro youths 12–17 years, United States. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 126. Public Health Service. Washington: Government Printing Office, August 1973. 6. National Center for Health Statistics: Height and weight of children: socioeconomic status. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 119. Public Health Service. Washington: Government Printing Office, October 1972. 7. Barr GD, Allen CM, Shinefield HR. Height and weight of 7500 children of three skin colors. Am J Dis Child 1972; 124:866. 8. Wingerd J, Schoen EJ, Solomon IL. Growth standards in the first two years of life based on measurements of white and black children in a prepaid healthcare program. Pediatrics 1971; 47:818. 9. Wingerd J, Solomon IL, Schoen EJ. Parent-specific height standards for preadolescent children of three racial groups, with method for rapid determination. Pediatrics 1973; 52:555. 10. Lippe BM. Short stature in children: evaluation and management. J Pediatr Health Care 1987; 1:313. 11. Tanner JM, Whitehouse RM. Clinical longitudinal standards for height, weight, height velocity, weight velocity, and stages of puberty. Arch Dis Child 1976; 51:170. 12. Tanner JM, Davies PSW. Clinical longitudinal standards for height and weight velocity for North American children. J Pediatr 1985; 107:317. 13. Smith DW, Truog W, Rogers JE, et al. Shifting linear growth during infancy: illustration of genetic factors in growth from fetal life through infancy. J Pediatr 1976; 89:225. 14. Tanner JM, Goldstein H, Whitehouse RH. Standards for children's heights at ages 2–9 years allowing for heights of parents. Arch Dis Child 1970; 45:755. 15. Himes JH, Roche AF, Thissen D, Moore WM. Parent-specific adjustments for evaluation of recumbent length and stature of children. Pediatrics 1985; 75:304. 16. Luo ZC, Albertson-Wikland K, Karlberg J. Target height as predicted by parental heights in a population-based study. Pediatr Res 1998; 44:563. 17. Cruise MO. A longitudinal study of the growth of low birth weight infants. I. Velocity and distance growth, birth to 3 years. Pediatrics 1973; 51:620. 18. Fritzhardinge PM, Steven EM. The small-for-date infant. I. Later growth patterns. Pediatrics 1972; 49:671. 19. Guo SS, Roche AE, Chumlea W, et al. Adjustments to the observed growth of preterm low birth weight infants for application to infants who are small for gestational age at birth. Acta Med Auxol 1998; 30:71. 20. Strauss RS, Dietz WH. Growth and development of term infants born with low birth weight: effects of genetic and environmental factors. J Pediatr 1998; 133:67. 21. Leger J, Limoni C, Collin D, Czernichow P. Prediction factors in the determination of final height in subjects born small for gestational age. Pediatr Res 1998; 43:808. 22. Ranke MB. Disease-specific growth charts: do we need them? Acta Paediatr Scand 1989; 356(Suppl):17. 23. Park E, Bailey JD, Cowell CA. Growth and maturation of patients with Turner's syndrome. Pediatr Res 1983; 17:1. 24. Lyon AJ, Preece MA, Grant DB. Growth curves for girls with Turner syndrome. Arch Dis Child 1985; 60:932. 25. Marshall WA. Evaluation of growth rate in height over periods of less than one year. Arch Dis Child 1971; 46:414. 26. Cureton KJ, Boileau RA, Lohman TG. A comparison of densitometric, potassium-40 and skinfold estimates of body composition in prepubescent boys. Hum Biol 1975; 47:321. 27. National Center for Health Statistics. Skinfold thickness of children 6–11 years. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 120. Public Health Service, Washington: Government Printing Office, October 1972. 28. National Center for Health Statistics. Skinfold thickness of youths 12–17 years, United States. Vital and Health Statistics. PHS Pub. No. 1000—series 11, No. 132. Public Health Service, Washington: Government Printing Office, January 1974. 29. Sann L, Durand M, Picard J, et al. Arm fat and muscle areas in infancy. Arch Dis Child 1988; 63:256. 30. Rosner B, Prineas R, Loggie J, Daniels SR. Percentiles for body mass index in U.S. children 5 to 17 years of age. J Pediatr 1998; 132:211. 31. Dietz WH, Robinson TN. Use of the body mass index (BMI) as a measure of overweight in children and adolescents. J Pediatr 1998; 132:191. 32. Greulich W, Pyle SL. A radiographic standard of reference for the growing hand and wrist. Chicago: Year Book Medical Publishers, 1971. 33. Tanner JM, Whitehouse RH, Cameron N, et al. Assessment of skeletal maturity and prediction of adult height (TW2 method), 2nd ed. London: Academic Press, 1983. 34. Tanner J, Oshman D, Bahhage F, Healy M. Tanner-Whitehouse bone age reference values for North American children. J Pediatr 1997; 131:34. 35. Graham CB. Assessment of bone maturation: methods and pitfalls. Radiol Clin North Am 1972; 10:185. 36. National Center for Health Statistics. Skeletal maturity of children 6–11 years. Vital and Health Statistics, PHS Pub. No. 1000—series 11, No. 140. Public Health Service. Washington: Government Printing Office, November 1974. 37. Buckler JMH. Skeletal age changes in puberty. Arch Dis Child 1984; 59:115. 38. Bailey N, Pinneau SR. Tables for predicting adult height from skeletal age: revised for use with the Greulich-Pyle hand standards. J Pediatr 1952; 14:423. 39. Roche AF, Wainer H, Thissen D. The RWT method for the prediction of adult stature. Pediatrics 1975; 56:1026. 40. Lenko HL. Prediction of adult height with various methods in Finnish children. Acta Paediatr Scand 1979; 68:85. 41. Zachmann M, Sobradillo B, Frank H, Prader A. Bayley-Pinneau, Roche-Wainer-Thissen and Tanner height predictions in normal children and in patients with various pathologic conditions. J Pediatr 1978; 93:749. 42. de Waal WJ, Greyn-Fokker MH, Stijnen TH, et al. Accuracy of final height prediction and effect of growth-reductive therapy in 362 constitutionally tall children. J Clin Endocrinol Metab 1996; 81:1206. 43. Abbassi V. Growth and normal puberty. Pediatrics 1998; 102:507. 44. August GP, Julius JR, Blethen SL. Adult height in children with growth hormone deficiency who are treated with biosynthetic growth hormone: the National Cooperative Growth Study experience. Pediatrics 1998; 102:512. 45. Marti-Henneberg C, Vizmanos B. The duration of puberty in girls is related to the timing of its onset. J Pediatr 1997; 131:618. 46. Blethen SL, Gaines S, Weldon V. Comparison of predicted and adult heights in short boys: effect of androgen therapy. Pediatr Res 1984; 18:467. 47. Winter JSD, Faiman C. Pituitary-gonadal relations in male children and adolescents. Pediatr Res 1972; 6:126. 48. Zachmann M, Prader A, Kind HP, et al. Testicular volume during adolescence: cross sectional and longitudinal studies. Helv Paediatr Acta 1974; 29:61. 49. Schonfeld WA. Primary and secondary sexual characteristics: study of their development in males from birth through maturity, with biometric study of penis and testes. Am J Dis Child 1943; 65:535. 50. Belisle S, Tulchinsky D. Amniotic fluid hormones. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:169. 51. Gluckman PD, Grumbach MM, Kaplan SL. The human fetal hypothalamus and pituitary gland: the maturation of neuroendocrine mechanisms controlling the secretion of fetal pituitary growth hormone, prolactin, gonadotropin, and adrenocorticotropin-related peptides. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:196. 52. Cacciari E, Cicognani A, Pirazzoli P, et al. Differences in somatomedin-C between short-normal subjects and those of normal height. J Pediatr 1985; 106:891. 53. Juul A, Bang P, Hertel NT, et al. Serum insulin-like growth factor-I in 1030 healthy children, adolescents, and adults: relation to age, sex, stage of puberty, testicular size, and body mass index. J Clin Endocrinol Metab 1994; 78:744. 54. Blum WF, Albertsson-Wikland K, Rosberg S, Ranke MB. Serum levels of insulin-like growth factor I (IGF-I) and IGF binding protein 3 reflect spontaneous growth hormone secretion. J Clin Endocrinol Metab 1993; 76:1610. 55. Smith WJ, Nam TJ, Underwood LE, et al. Use of Insulin-like growth factor-binding protein-2 (IGFBP-2), IGFBP-3, and IGF-I for assessing growth hormone status in short children. J Clin Endocrinol Metab 1993; 77:1294. 56. Sperling MA. Newborn adaptation: adrenocortical hormones and ACTH. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:387. 57. Oddie TH, Fisher DA, Bernard B, Lam RW. Thyroid function at birth in infants of 30 to 45 weeks gestation. J Pediatr 1977; 90:803. 58. Burrow GN, Fisher DA, Larsen PR. Maternal and fetal thyroid function. N Engl J Med 1994; 331:1072. 59. Fisher DA, Klein AH. The ontogenesis of thyroid function and its relationship to neonatal thermogenesis. In: Tulchinsky D, Ryan KJ, eds. Maternal-fetal endocrinology. Philadelphia: WB Saunders, 1980:281. 60. Van Wassenaer AG, Kok JH, Dekker FW, De Vulder JJM. Thyroid function in very preterm infants: influences of gestational age and disease. Pediatr Res 1997; 42:604. 61. Beitins IZ, Padmanabhan V. Bioactivity of gonadotropins. Endocrinol Metab Clin North Am 1991; 20:85. 62. Beck-Peccoz P, Padman V, Baggiani AM, et al. Maturation of hypotha-lamic-pituitary-gonadal function in normal human fetuses: circulating levels of gonadotropins, their common a-subunits and free testosterone, and discrepancy between immunological and biological activities of circulating follicle-stimulating hormone. J Clin Endocrinol Metab 1991; 73:525. 63. Kletter GB, Padmanabhan V, Brown MB, et al. Serum bioactive gonadotropin during male puberty: a longitudinal study. J Clin Endocrinol Metab 1993; 76:432. 64. Reyes FI, Boroditsky RS, Winter JSD, Faiman C. Studies on human sexual development. II. Fetal and maternal serum gonadotropin and sex steroid concentrations. J Clin Endocrinol Metab 1974; 38:612. 65. Forest MC, Cathiard AM, Bertrand JA. Evidence of testicular activity in early infancy. J Clin Endocrinol Metab 1973; 37:148. 66. Forest M, Cathiard AM. Pattern of plasma testosterone and D-4-andro-stenedione in normal newborns: evidence for testicular activity at birth. J Clin Endocrinol Metab 1975; 41:977. 67. Winter JSD, Faiman C, Hobson WC, et al. Patterns of serum gonadotropin concentrations from birth to four years of age in man and chimpanzee. J Clin Endocrinol Metab 1975; 40:545. 68. Winter JSD, Hughes IA, Reyes FI, et al. Pituitary-gonadal relations in infancy. II. Patterns of serum gonadal steroid concentration in man from birth to two years of age. J Clin Endocrinol Metab 1976; 42:679. 69. Reiter ED, Fuldaurer VG, Root AW. Secretion of the adrenal androgen, dehydroepiandrosterone sulfate, during normal infancy, childhood, in sick infants and in children with endocrinologic abnormalities. J Pediatr 1977; 90:766.

70. Fiselier T, Monnens L, van Munster P, et al. The renin-angiotensin-aldoster-one system in infancy and childhood in basal conditions and after stimulation. Eur J Pediatr 1984; 143:18. 82—blank

CHAPTER 8 MORPHOLOGY OF THE ENDOCRINE BRAIN, HYPOTHALAMUS, AND NEUROHYPOPHYSIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 8 MORPHOLOGY OF THE ENDOCRINE BRAIN, HYPOTHALAMUS, AND NEUROHYPOPHYSIS JOHN R. SLADEK, JR., AND CELIA D. SLADEK Overview of the Hypothalamus Nuclear Topography Hypothalamic Zones Preoptic Area Classic Pathways Afferent Pathways Efferent Pathways Hypothalamo-Neurohypophysial System Chemical Neuroanatomy Hormones as Chemical Neurotransmitters Vasopressin Release as a Model of Interactive Chemical Circuitry Chapter References

OVERVIEW OF THE HYPOTHALAMUS The human hypothalamus is a tiny wedged-shaped mass of tissue, composed primarily of gray matter, that subserves widespread functions, ranging from those considered somewhat automatic (i.e., autonomic) to more complex behaviors requiring a high level of integration. The hypothalamus is situated in the ventralmost diencephalon (Fig. 8-1). It is bounded laterally by portions of the subthalamus; medially by the vertically oriented, slitlike third ventricle; rostrally by the lamina terminalis; caudally by the mesencephalon; and dorsally by the thalamus. Ventrally, the hypothalamus is contiguous with the infundibulum and pituitary stalk; the latter serves to transmit axons en route to the neurohypophysis. On the ventral surface of the brain, the hypothalamus appears as a prominent set of protuberances, the paired mammillary bodies caudally, and the midline infundibular eminence. The boundaries of the hypothalamus are easily identified by the optic chiasm rostrally and by the caudal edge of the mammillary bodies as they are situated at the rostral limit of the interpeduncular fossa. Here, the caudal limits of the hypothalamus merge with the midbrain.

FIGURE 8-1. Human hypothalamus in sagittal section. The hypothalamus is bounded anteriorly by the lamina terminalis, the tissue that bridges the anterior commissure and optic chiasm; posteriorly by the brainstem; and dorsally by the overlying thalamus (adjacent to the label for the third ventricle). The hypothalamus is relatively small in comparison to other brain regions, yet is intimately involved in mediating or moderating many brain and endocrine functions. (From Krieger DT, Hughes JC, eds. Neuroendocrinology. Sunderland, MA: Sinauer Associates, 1980:3.)

Classic neuroanatomic techniques have been used to identify hypothalamic nuclei cytoarchitectonically. Some nuclei are relatively easy to identify microscopically, based on size or packing density of neurons; others benefit from placement close to readily identified landmarks such as the third ventricle or the optic chiasm. Prominent myelinated fiber bundles punctuate the relatively nondescript nuclear divisions of the hypothalamus, indicating the diverse reciprocal connections that exist between the hypothalamus and numerous brain regions. The fornix and mammillary fasciculus interconnect the hypothalamus with, for example, the hippocampus, the thalamus, and the brainstem. Other brain regions influence hypothalamic function through pathways that are seen to advantage with current histochemical methods. These systems, which include the medial forebrain bundle and the dorsal longitudinal fasciculus, interconnect the hypothalamus with olfactory areas rostrally and with autonomic centers of the brainstem and spinal cord caudally. Thus, the hypothalamus is an important link between the forebrain and brainstem, integrating visceral, endocrine, and behavioral functions through a highly ordered set of reciprocating circuits.

NUCLEAR TOPOGRAPHY HYPOTHALAMIC ZONES Because the hypothalamus is so rich in perikaryal groups and so poor in landmarks (e.g., myelinated fiber bundles), the prudent approach is to consider easily identified hypothalamic zones as geographic points with respect to the more than two dozen classically identified hypothalamic nuclei, as detailed by Nauta and Haymaker.1 Zones in both longitudinal and coronal planes are illustrated in Figure 8-2.

FIGURE 8-2. Diagram showing anatomic relationship of hypothalamic nuclei, by zones. Here, the hypothalamus is considered as a hexagon, bordered anteriorly by the preoptic area and posteriorly by the midbrain. The third ventricle (wide black line) forms its midline. Most hypothalamic nuclei can be found within three longitudinal zones, designated as midline, medial, and lateral in relation to the third ventricle and fornix. These zones can be further subdivided into three anteroposterior regions or levels (supraoptic, tuberal, and mammillary), allowing identification of hypothalamic nuclei by position, an approach that is much simpler than distinguishing the nuclei by the classic cytoarchitectonic method. The three levels, shown in the coronal plane in Figure 8-3, depict the essential hypothalamic nuclei.

Longitudinal Plane. The longitudinal zones are based on the phylogenetically primitive organization of lower vertebrates, in which the hypothalamus is characterized by a relatively cell-rich medial zone. A somewhat acellular lateral zone is separated from the medial zone by a prominent fiber bundle, the fornix. The medial zone can be subdivided further into a midline zone that is immediately adjacent to the third ventricle and contains a relatively homogeneous nuclear mass called the periventricular stratum. The more laterally placed portions of the medial hypothalamic zone contain reasonably differentiated cell clusters, including the medial preoptic,

anterior hypothalamic, ventromedial, dorsomedial, paraventricular, posterior, and premammillary nuclei. Coronal Plane. For convenience, the hypothalamus in the coronal plane can be further subdivided anteroposteriorly into three regions: the supraoptic, tuberal, and mammillary regions (Fig. 8-3 and Fig. 8-4). The supraoptic region includes that portion of the hypothalamus situated between the optic chiasm and tuber cinereum. The tuberal zone extends caudally to the most anterior portion of the mammillary bodies. The mammillary region includes the most caudal hypothalamus to the mesodiencephalic junction.

FIGURE 8-3. The hypothalamus in the coronal plane, demonstrating the relative positions of hypothalamic nuclei. At the supraoptic level, the midline zone (dashed line) consists of the periventricular nucleus (Peri) and its ventral expansion, the suprachiasmatic nucleus (SCN). The periventricular nucleus continues through the more caudal levels. At tuberal levels, the SCN is replaced by the arcuate nucleus (Arc). The medial zone, defined as all the tissue between the fornix laterally and the periventricular nucleus medially, contains five major hypothalamic nuclei that essentially replace each other positionally at each level. Thus, the paraventricular (PVN) and anterior (AN) nuclei, which occupy the supraoptic level, are replaced by the dorsomedial (DM) and ventromedial (VM) nuclei at tuberal levels; these nuclei in turn are replaced by the posterior nucleus (POST) at the most caudal, mammillary level. The lateral zone consists almost exclusively of the lateral hypothalamic area (LHA), with the exception of the supraoptic nucleus (SON), which is seen at supraoptic levels. The third ventricle appears as the dark structure in the midline. (F, fornix; ME, median eminence; MM, mammillary body; OC, optic chiasm; OT, optic tract.)

FIGURE 8-4. The hypothalamus in a parasagittal plane, showing the relative positions of major nuclei. These relationships can be better appreciated by comparing this figure with Figure 8-1 and Figure 8-3. For example, three nuclear groups, the paraventricular, dorsomedial, and posterior, appear to replace one another as one proceeds anatomically, rostrally to caudally, through the three major coronal levels—supraoptic, tuberal, and mammillary (see Fig. 8-3). (From Carpenter MB, Sutin J, eds. Human neuroanatomy. Baltimore: Williams & Wilkins, 1983:553.)

The midline zone accordingly contains the periventricular stratum throughout most of its rostrocaudal extent. At tuberal regions, the ventral periventricular stratum expands laterally to accommodate the arcuate (infundibular) nucleus. At supraoptic levels, a ventral region of the periventricular stratum is identified as the suprachiasmatic nucleus. Although numerous other fine points of periventricular zone anatomy exist, one could consider the two subcomponents of this region as the suprachiasmatic and arcuate nuclei. The medial hypothalamic zone, defined as all remaining tissue lateral to the periventricular stratum and medial to the descending limb of the fornix, contains the paraventricular and anterior hypothalamic nuclei at supraoptic levels, the dorsomedial and ventromedial nuclei at tuberal levels, and the posterior hypothalamic nucleus at mammillary levels (see Fig. 8-3). Thus, as one proceeds rostrally to caudally through the three hypothalamic regions, a gradual replacement of these well-differentiated nuclei is seen. The lateral hypothalamic zone is relatively undistinguished, with the lateral hypothalamic area occupying most of this zone in all three regions. The lone exception is the supraoptic nucleus, which is seen at the supraoptic level. Because of the complex nature of the paraventricular nucleus, it is sometimes considered as part of the periventricular zone, but the alarlike lateral extensions of this nucleus clearly place a large part of it within the medial zone. PREOPTIC AREA The preoptic area, located rostral to the strict limits of the hypothalamus, contains nuclei in all three (medial to lateral) zones, including the preoptic periventricular, medial preoptic, and lateral preoptic nuclei (see Fig. 8-4). Functionally, the preoptic region is integrated with the remainder of the hypothalamus because it includes neurons that regulate anterior pituitary function as well as structures that are essential for fluid and electrolyte balance. A full description of these hypothalamic regions and nuclei may be found in two classic accounts.1,2

CLASSIC PATHWAYS AFFERENT PATHWAYS Afferent pathways to hypothalamic nuclei arise primarily from the brainstem, thalamus, basal ganglia, cerebral cortex, and olfactory areas. The detailed anatomy of these connections and hypothalamic interconnections is described in several fine reviews.1,3,4 Briefly, the reticular formation and visceral centers of the brainstem connect with the hypothalamus through two prominent pathways, the mammillary peduncle and the dorsal longitudinal fasciculus. Visceral and somatic information also reaches the hypothalamus from the locus ceruleus, vagal nuclei, periaqueductal gray area, and nuclei of the solitary tract (Fig. 8-5). The fornix transmits fibers from the hippocampus by direct projections to the mammillary bodies. Additional afferents from the piriform cortex and the amygdala also reach the hypothalamus, probably by other routes. Olfactory information through the medial forebrain bundle and stria terminalis reaches the hypothalamus either directly or indirectly through the previously mentioned cortical regions, the stria terminalis being the primary pathway from the amygdala. Most of the afferent pathways are reciprocal; thus, the hypothalamic efferent connections are extensive.

FIGURE 8-5. Major hypothalamic pathways. Pathways to the hypothalamus arise from several areas, including olfactory, limbic, and brainstem regions; reciprocal

circuits exist, particularly to autonomic centers of the caudal medulla and spinal cord. The dorsal longitudinal fasciculus conveys most of the descending information to brainstem nuclei and is identified by the parallel, opposite arrows in this drawing. Another major pathway, the fornix, carries higher cortical information from the hippocampus (Hipp.) to the mammillary nucleus. (AV, nucleus ventralis anterior of thalamus; BL, Ce., and Co., basolateral, central, and cortical amygdaloid nuclei; Bulb. olf., olfactory bulb; Coll. sup., superior colliculus; H, hypothalamus; M, mammillary body; MD, dorsomedial thalamic nucleus; N. Dors. n. X, dorsal motor vagal nucleus; N. loc. coer., locus coeruleus; N. raphe, nuclei of the raphe; N. tr. solit., nucleus tractus solitarius; Olf. Cort., olfactory cortex; Optic ch., optic chiasma; Periaq. gr., periaqueductal gray; Prefr. cort., prefrontal cortex; RF, reticular formation; S, septum; VM, ventromedial hypothalamic nucleus.) (From Brodal A, ed. Neurological anatomy in relation to clinical medicine, 3rd ed. New York: Oxford University Press, 1981:698.)

EFFERENT PATHWAYS The dorsal longitudinal fasciculus transmits information to brainstem reticular centers as well as to visceral and somatic efferent nuclei. This system descends caudally to innervate preganglionic autonomic centers of the spinal cord, particularly the intermediolateral gray column.5,6 Although the hypothalamus was once thought to interconnect with autonomic centers by multisynaptic pathways, it is now clear that the dorsal motor nucleus of the vagus, the nuclei of the solitary tract, and the nucleus ambiguus receive direct projections from the paraventricular nucleus of the hypothalamus. Thus, reciprocal connections between these regions form an integrated autonomic circuit. Other fibers reach the brainstem through the medial forebrain bundle. Hypothalamic efferents projecting to the thalamus traverse the mammillothalamic tract to the anterior nucleus of the thalamus, where they are relayed to the cerebral cortex. Other efferent fibers ascend over the more diffuse periventricular system. The hypothalamus is connected with the neurohypophysis by the hypothalamo-neurohypophysial tract, a system of unmyelinated fibers that terminates in the neurohypophysis for the purpose of storing and then delivering vasopressin, oxytocin, and possibly other endogenous peptides such as dynorphin to the pituitary blood. HYPOTHALAMO-NEUROHYPOPHYSIAL SYSTEM In contrast to other hypothalamic nuclei, the paraventricular and supraoptic nuclei are easy to identify (Fig. 8-6). Each nucleus contains the largest nerve cells in the hypothalamus (Fig. 8-7). The cells stain densely with common dyes that reveal the abundant ribonucleoprotein (i.e., Nissl substance) in their perikaryal cytoplasm. Ultrastructurally, the protein-synthesizing machinery of these neurons also is evident and reflects their high level of activity with respect to the manufacture of vasopressin, oxytocin, and coexistent peptides. The neurons are revealed in excellent detail when stained immunohistochemically with antibodies directed against specific hypothalamic peptides (see Fig. 8-7).7,8 Such analysis has revealed that oxytocin neurons also contain cholecystokinin9 and that vasopressin neurons also contain dynorphin.10 Axons from the supraoptic and paraventricular nuclei as well as associated accessory nuclei of the hypothalamus project ventrally and caudally (see Fig. 8-6) to the underlying neurohypophysis, where they release oxytocin, vasopressin, and associated peptides into the peripheral circulation to regulate fluid and electrolyte balance and to initiate the smooth muscle contraction that is associated with lactation and parturition.

FIGURE 8-6. Immunohistochemical staining for neurophysins associated with vasopressin and oxytocin reveals the extent of the neurohypophysial system at the supraoptic level of the rat hypothalamus. Dense staining is seen within neurons of the paraventricular (PVN) and supraoptic (SON) nuclei. A ventral flow of axons from each nucleus (arrows) represents the origin of the hypothalamo-neurohypophysial tract as it courses to the posterior pituitary. Stained neurons also are seen within the suprachiasmatic nucleus (SCN); however, this nucleus does not contribute fibers to the neural lobe. (OC, optic chiasm; V, third ventricle.) Original magnification ×30

FIGURE 8-7. Appearance of neurons of the paraventricular nucleus after immunohistochemical staining for vasopressin. This set of magnocellular neurons occupies a subnucleus of the paraventricular nucleus that contains primarily vasopressin neurons that project to the neural lobe.30 The neurons are large, multipolar, and possess beaded processes. The position of the third ventricle is indicated (V). Original magnification ×350

Other neurons of the paraventricular nucleus are global, with widespread connections to autonomic centers of the brainstem and spinal cord, and to the forebrain and cortical areas, including the septum, cingulum, and hippocampus.11a These far-reaching interconnections focus attention on oxytocin and vasopressin as more than simply neurohypophysial hormones. The demonstration of these peptides in presynaptic nerve terminals suggests that they may function as neurotransmitters.

CHEMICAL NEUROANATOMY HORMONES AS CHEMICAL NEUROTRANSMITTERS The hypothalamus is one of the most complicated areas of the brain with respect to chemical neurotransmitters because of the small amount of tissue occupied by the hypothalamus and the great number of transmitter substances located within hypothalamic nerve cells and associated fiber systems.12 The classic studies of Bargmann13 and of Scharrer and Scharrer,14 who are credited with describing the principles of neurosecretion, brought to light the unique chemical characteristics of neurons of the supraoptic and paraventricular nuclei. The introduction in the early 1960s of the Falck-Hillarp histofluorescence method15 allowed identification of the catecholamines dopamine and norepinephrine within the hypothalamus. Most notable were the dopaminergic neurons of the tuberoinfundibular system, which are involved in regulating the release of anterior pituitary substances.16 A decade later, immunohistochemical methods17 increased the list of known hypothalamic chemical neurotransmitters and modulators to include substances such as luteinizing hormone–releasing hormone, corticotropin-releasing hormone, vasoactive intestinal peptide, neurotensin, somatostatin, enkephalin, endorphin, cholecystokinin, galanin, and several others.18,19 Although the discovery of luteinizing hormone–releasing hormone20 and somatostatin within hypothalamic, preoptic, and adjacent regions of the endocrine hypothalamus was not surprising, the finding of “gut peptides” and the discovery of a complex system of opioid neurons21 have redefined the hypot halamus, based on the chemical cytoarchitecture of transmitters and hormones.18,21a The application of in situ hybridization techniques to localize messenger RNA for these peptides indicates that this wide array of peptides is synthesized in the

hypothalamus and that neurons have the capacity to synthesize multiple regulatory peptides simultaneously.22 The specific peptides produced by a given neuron are not static but depend on the stimuli received by that cell.22,23 For example, hypophysectomy dramatically increases the expression of galanin in the vasopressin neurons of the supraoptic nucleus and of cholecystokinin in the oxytocin neurons, but salt loading induces the expression of tyrosine hydroxylase in the vasopressin neurons and of corticotropin-releasing hormone in the oxytocin neurons.24 The role of these simultaneously released peptides is not completely defined, but in at least some instances, they interact to regulate hormone release from the anterior pituitary or to modulate hormone release from the posterior pituitary. VASOPRESSIN RELEASE AS A MODEL OF INTERACTIVE CHEMICAL CIRCUITRY NOREPINEPHRINE REGULATION OF VASOPRESSIN AND OXYTOCIN The role of afferents to the paraventricular and supraoptic nuclei is considered relative to the regulation of vasopressin release as exemplary of the kind of functionally interactive chemical circuitry that is being revealed with respect to hypothalamic neuroanatomy. The supraoptic and paraventricular nuclei receive dense, diverse afferent inputs, many arising from the brainstem reticular formation. One of these is a well-defined system of noradrenergic afferents to vasopressin neurons of the supraoptic and paraventricular nuclei (Fig. 8-8). The paraventricular nuclei in turn send reciprocal peptidergic fibers to the reticular core of the brainstem. Norepinephrine-containing perikaryal groups in the brainstem have been designated A1 to A7.25 These noradrenergic neurons originally were seen primarily within the reticular formation of the pons and medulla. Later, attention focused on groups A1, A2, A5, and A7 as projecting to the hypothalamus through a ventral pathway that ascends within the dorsal portion of the reticular formation of the brainstem, entering the medial forebrain bundle at hypothalamic levels in association with serotonergic and dopaminergic systems from the brainstem. 26 On reaching the hypothalamus, these fibers exit the medial forebrain bundle to supply the supraoptic and paraventricular nuclei. The densest patterns of noradrenergic fibers—perhaps the densest patterns in the entire brain—are seen in the mammalian hypothalamus27,28 (Fig. 8-9). These fibers appear in contact with the cell bodies of the magnocellular neurons, lending further support to the concept that norepinephrine plays a role in the regulation of neurohypophysial peptides.29

FIGURE 8-8. Ascending noradrenergic axons reach the magnocellular nuclei of the hypothalamus through the ventral norepinephrine pathway (VNE) of the brainstem, which continues rostrally in the medial forebrain bundle (MFB) of the diencephalon. The neurons of origin of this system (A1, A2) are located in the lateral reticular formation of the medulla near the lateral reticular nucleus (LRN) and in an area of the dorsomedial medulla that is important in cardiovascular regulation—the nucleus solitarius (SOL) and the dorsal motor vagal nucleus (DMX), respectively. This pathway, which is probably reciprocal from the paraventricular nucleus (PVN), supplies a dense noradrenergic innervation to the magnocellular nuclei, as shown for the monkey (also see Fig. 8-9). (CC, corpus callosum; MLF, medial longitudinal fasciculus; OC, optic chiasm; PY, pyramidal tract; SON, supraoptic nucleus.)

FIGURE 8-9. Patterns of catecholamine innervation of hypothalamic nuclei at a supraoptic level in the rhesus monkey. Exceptionally dense patterns exist in the paraventricular (Pa) and supraoptic (So) nuclei; less impressive patterns are seen in other nuclei at this level. Depictions such as this led to intense examination of the role of catecholamines in the release of neurohy-pophysial peptides. (An, anterior nucleus; op ch, optic chiasma; Prl, lateral preoptic nucleus; Sc, suprachiasmatic nucleus; v, third ventricle.) (From Hoffman GE, Felten DL, Sladek JR Jr. Monoamine distribution in primate brain. III. Catecholamine-containing varicosities in the hypothalamus of Macaca mulatta. Am J Anat 1976; 147:501.)

In subsequent studies, the simultaneous demonstration of catecholamine fluorescence and peptide immunohistochemistry30 revealed a consistent juxtaposition between catecholamine varicosities and magnocellular perikarya in the supraoptic and paraventricular nuclei in both rodents31 and primates.32 Specifically, vasopressin-containing neurons in ventral portions of the supraoptic nucleus and the major vasopressin subcomponent of the paraventricular nucleus were seen to be studded with brightly fluorescent catecholamine-containing varicosities. In contrast, oxytocin neurons received far fewer fluorescent fibers on their cell bodies and proximal dendrites. The preferential innervation of vasopressin neurons primarily reflects innervation by the A1 noradrenergic cells of the ventrolateral medulla.33,34 and 35 The sparse innervation of the oxytocin neurons reflects innervation of the nuclei by the A2 noradrenergic cell group that does not differentiate between oxytocin and vasopressin neurons.35 CATECHOLAMINES, VASOPRESSIN, AND AUTONOMIC REGULATION The noradrenergic cell groups in the brainstem that innervate the vasopressin neurons are intimately involved in autonomic regulation. The A2 catecholamine neurons are located in the nucleus of the tractus solitarius, which receives baroreceptor information from the carotid sinus and aortic arch. This information is also transmitted to the A1 cells in the ventrolateral medulla that project to the vasopressin neurons.36 The paraventricular nucleus also is innervated by the locus ceruleus (A6), an important autonomic relay center.37 Thus, the catecholamine input to the supraoptic and paraventricular nuclei represents a source of autonomic regulation of neurohypophysial function. These pathways are important because blood pressure, blood volume, and the partial pressure of oxygen are potent regulators of vasopressin release from the neural lobe. Stimulation of the A1 region results in an increase in blood pressure and enhanced vasopressin release, and inhibition of this region prevents baroreceptor-initiated secretion of vasopressin.38 This response is prevented by destruction of the catecholamine terminals in the supraoptic and paraventricular nuclei39 but is not prevented by administration of adrenergic antagonists.40 This can be explained by the finding that, in addition to norepinephrine, the A1 neurons produce neuropeptide Y, adenosine triphosphate (ATP), and substance P.41,42 Pharmacologic evidence indicates that these neuroactive substances participate in the excitation of vasopressin neurons by the A1 pathway.43,44 The central role of the paraventricular nucleus and vasopressin in blood pressure regulation and other autonomic functions was underscored by the finding of reciprocal pathways containing vasopressin and oxytocin in the same brainstem regions that give origin to the ascending noradrenergic fibers to the supraoptic and paraventricular nuclei. The finding of neurophysin-positive varicosities in juxtaposition to norepinephrine perikarya in the A1, A2, A5, and A7 brainstem groups suggests a functionally interactive reciprocal circuit.45 Moreover, microinjection of vasopressin into these regions has profound effects on cardiovascular function. Thus, a reciprocal pathway involving catecholamine afferents and vasopressin efferents is one mechanism involved in hypothalamic modulation of the autonomic nervous system.

OTHER CHEMICALLY DEFINED AFFERENTS AND VASOPRESSIN RELEASE The paraventricular and supraoptic nuclei receive chemically defined afferents from other central nervous system sites.46,47 These include serotonergic projections from the midbrain raphe nuclei, GABAergic (transmitting or secreting g-aminobutyric acid) projections from the region of the nucleus accumbens, and several hypothalamic projections. The hypothalamic afferents include cholinergic48 and GABAergic projections arising from cells near the paraventricular and supraoptic nuclei, b-endorphinergic afferents from the arcuate nucleus, histaminergic afferents from the tuberomammillary nuclei,49,50 dopaminergic afferents from neurons in the A11, A12, and A13 groups,51 and numerous projections from the preoptic region, including projections from the subfornical organ and the organum vasculosum of the lamina terminalis. The chemical nature of these last projections includes glutamatergic, GABAergic, angiotensinergic, and atriopeptidergic fibers. The functional role of many of these afferents is unknown, but the afferents from the preoptic region and the circumventricular organs clearly are intimately involved in the osmotic regulation of vasopressin release as part of their role in fluid and electrolyte balance. A particularly important role for the classic excitatory amino acid neurotransmitter glutamate in regulating hypothalamic activity has been suggested by the presence of numerous immunocytochemically identified glutamatergic synapses in the hypothalamus and the finding that glutamate antagonists virtually eliminate excitatory postsynaptic potentials in the paraventricular, supraoptic, and arcuate nuclei. Thus, the hypothalamus continues to emerge as a complex brain region in which a wide variety of classic and more recently discovered transmitters serve numerous endocrine and other functions related to autonomic, limbic, and pituitary activity.52 CHAPTER REFERENCES 1. Nauta WJH, Haymaker W. Hypothalamic nuclei and fiber connections. In: Haymaker W, Anderson E, Nauta WJH, eds. The hypothalamus. Springfield, IL: Charles C Thomas, 1969:136. 2. Crosby EC, Woodburne RT. The comparative anatomy of the preoptic area and the hypothalamus. Research Publications—Association for Research in Nervous and Mental Disease. New York: Raven Press, 1939; 20:52. 3. Brodal A. The autonomic nervous system: the hypothalamus. In: Brodal A, ed. Neurological anatomy in relation to clinical medicine. New York: Oxford University Press, 1981:698. 4. Palkovits M, Zaborszky L. Neural connections of the hypothalamus. In: Morgane PJ, Panksepp J, eds. Handbook of the hypothalamus, vol 1. Anatomy of the hypothalamus. New York: Marcel Dekker, 1979:379. 5. Swanson LW, Kuypers HBJM. The paraventricular nucleus of the hypothalamus: cytoarchitectonic subdivision and organization of projections to the pituitary, dorsal vagal complex, and spinal cord as demonstrated by retrograde fluorescence double-labeling methods. J Comp Neurol 1980; 194:555. 6. Nilaver G, Zimmerman EA, Wilkins J, et al. Magnocellular hypothalamic projections to the lower brainstem and spinal cord of the rat. Neuroendocrinology 1980; 30:150. 7. Swaab D, Pool C, Nijveldt F. Immunofluorescence of vasopressin and oxytocin in the rat hypothalamo-neurohypophyseal system. J Neural Transm 1975; 36:195. 8. Silverman AJ, Zimmerman EA. Magnocellular neurosecretory system. Annu Rev Neurosci 1983; 6:357. 9. Vanderhagen JJ, Lotstra F, Vandesand F, Dierickx K. Coexistence of cholecystokinin and oxytocin-neurophysin in some magnocellular hypothalamohypophyseal neurons. Cell Tissue Res 1981; 221:227. 10. Watson SJ, Akil H, Fischli W, et al. Dynorphin and vasopressin: common localization in magnocellular neurons. Science 1982; 216:85. 11. Buijs RA. Intra- and extrahypothalamic vasopressin and oxytocin pathways in the rat: pathways to the limbic system, medulla oblongata and spinal cord. Cell Tissue Res 1978; 192:423. 11a. Jordan J, Shannon JR, Black BK, et al. The pressor response to water drinking in humans: a sympathetic reflex? Circulation 2000; 101:504. 12. Hoffman GH, Phelps CJ, Khachaturian H, Sladek JR Jr. Neuroendocrine projections to the median eminence. In: Pfaff DW, Ganten D, eds. Current topics in neuroendocrinology, vol 7. Morphology of hypothalamus and its connections. New York: Springer-Verlag, 1986:161. 13. Bargmann W. über der neurosekretorische Vernupfung von Hypothalamus und Neurohypophyse. Z Zellforsch 1949; 34:610. 14. Scharrer E, Scharrer B. Hormones produced by neurosecretory cell. Recent Prog Horm Res 1954; 10:183. 15. Falck B, Hillarp N-A, Thieme G, Torp A. Fluorescence of catecholamines and related compounds condensed with formaldehyde. J Histochem Cytochem 1962; 10:348. 16. Fuxe K, Hökfelt T. The influence of central catecholamine neurons on the hormone secretion from the anterior and posterior pituitary. In: Stutinsky F, ed. Neurosecretion. Berlin: Springer-Verlag, 1967:166. 17. Sternberger L, Hardy P, Cuculis J, Meyer H. The unlabeled antibody enzyme method in immunohistochemistry: preparation and properties of soluble antigen-antibody complex (horseradish peroxidase-antihorseradish peroxidase) and its use in identification of spirochetes. J Histochem Cytochem 1969; 18:315. 18. Hökfelt T, Elde R, Fuxe K, et al. Aminergic and peptidergic pathways in the nervous system with special reference to the hypothalamus. In: Reichlin S, Baldessarini RJ, Martin BJ, eds. The hypothalamus. Research Publications—Association for Research in Nervous and Mental Disease. New York: Raven Press, 1978:69. 19. Gai WP, Geffen LB, Blessing WW. Galanin immunoreactive neurons in the human hypothalamus: colocalization with vasopressin-containing neurons. J Comp Neurol 1990; 298:265. 20. Hoffman GE, Melnyk V, Hayes T, et al. Immunocytology of LHRH neurons. In: Scott DE, Kizlowski GP, Weindl A, eds. Brain-endocrine interactions, vol III. Neural hormones and reproduction. Basel: S Karger, 1978:67. 21. Watson SJ, Khachaturian H, Akil H, et al. Comparison of the distribution of dynorphin systems and enkephalin systems in brain. Science 1982; 218:1134. 21a. Patrick RL. Synaptic clefts are made to be crossed: neurotransmitter signaling in the central nervous system. Toxicol Pathol 2000; 28:31. 22. Lightman SL, Young WS III. Vasopressin, oxytocin, dynorphin, enkephalin and corticotrophin releasing factor mRNA stimulation in the rat. J Physiol (Lond) 1987; 394:23. 23. Meister B, Villar MJ, Ceccatelli S, Hökfelt T. Localization of chemical messengers in magnocellular neurons of the hypothalamic supraoptic and paraventricular nuclei: an immunohistochemical study using experimental manipulations. Neuroscience 1990; 37:603. 24. Meister B. Gene expression and chemical diversity in hypothalamic neuro-secretory neurons. Mol Neurobiol 1993; 7:87. 25. Dahlstrüm A, Fuxe E. Evidence for the existence of monoamine-containing neurons in the central nervous system. I. Demonstration of monoamines in the cell bodies of brainstem neurons. Acta Physiol Scand 1965; 62:1. 26. Ungerstedt U. Stereotaxic mapping of the monoamine pathways in the rat brain. Acta Physiol Scand 1971; 367:1. 27. Cheung Y, Sladek JR Jr. Catecholamine distribution in feline hypothalamus. J Comp Neurol 1975; 164:339. 28. Hoffman GE, Felten DL, Sladek JR Jr. Monoamine distribution in primate brain. III. Catecholamine-containing varicosities in the hypothalamus of Macaca mulatta. Am J Anat 1976; 147:501. 29. Sladek JR, Sladek CD. Neurological control of vasopressin release. Fed Proc 1985; 44:66. 30. Sladek JR Jr, McNeill TH. Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. IV. Verification of catecholamine-neurophysin interactions through single section analysis. Cell Tissue Res 1980; 210:181. 31. McNeill TH, Sladek JR Jr. Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. II. Correlative distribution of cate-cholamine varicosities and magnocellular neurosecretory neurons in the rat supraoptic and paraventricular nuclei. J Comp Neurol 1980; 193:1023. 32. Sladek JR Jr, Zimmerman EA. Simultaneous monoamine histofluorescence and neuropeptide immunocytochemistry. VI. Catecholamine innervation of vasopressin and oxytocin neurons in the rhesus monkey hypothalamus. Brain Res Bull 1983; 9:431. 33. Sawchenko PE, Swanson LW. The organization of noradrenergic pathways from the brainstem to the paraventricular and supraoptic nuclei in the rat. Brain Res Rev 1982; 4:275. 34. McKellar S, Loewy AD. Organization of some brainstem afferents to the paraventricular nucleus of the hypothalamus in the rat. Brain Res 1981; 217:351. 35. Cunningham ET Jr, Sawchenko PE. Reflex control of magnocellular vasopressin and oxytocin secretion. Trends Neurosci 1991; 14:406. 36. Day TA, Sibbald JR. A1 cell group mediates solitary nucleus excitation of supraoptic vasopressin cells. Am J Physiol 1989; 257:R1020. 37. Sladek JR Jr. Central catecholamine pathways to vasopressin neurons. In: Schrier RW, ed. Vasopressin. New York: Raven Press, 1985:343. 38. Blessing WW, Willoughby JO. Inhibiting the rabbit caudal ventrolateral medulla prevents baroreceptor-initiated secretion of vasopressin. J Physiol (Lond) 1985; 367:253. 39. Lightman SL, Todd K, Everitt BJ. Ascending noradrenergic projections from the brainstem: evidence for a major role in the regulation of blood pressure and vasopressin secretion. Exp Brain Res 1984; 55: 145. 40. Day TA, Renaud LP, Sibbald JR. Excitation of supraoptic vasopressin cells by stimulation of the A1 noradrenaline cell group: failure to demonstrate role for established adrenergic or amino acid receptors. Brain Res 1990; 516:91. 41. Bittencourt JC, Benoit R, Sawchenko PE. Distribution and origins of substance P-immunoreactive projections to the paraventricular and supraoptic nuclei: partial overlap with ascending catecholaminergic projections. J Chem Neuroanat 1991; 4:63. 42. Lundberg JM, Pernow J, Lacroix JS. Neuropeptide Y: sympathetic cotransmitter and modulator? News Physiol Sci 1989; 4:13. 43. Day TA, Sibbald JR, Khanna S. ATP mediates an excitatory noradrenergic neuron input to supraoptic vasopressin cells. Brain Res 1993; 607:341. 44. Sibbald JR, Wilson BKJ, Day TA. Neuropeptide Y potentiates excitation of supraoptic neurosecretory cells by noradrenaline. Brain Res 1989;499: 164. 45. Sladek JR Jr, Sladek CD. Anatomical reciprocity between magnocellular peptides and noradrenaline in putative cardiovascular pathways. Prog Brain Res 1983; 60:437. 46. Sladek CD, Armstrong WE. Effect of neurotransmitters and neuropeptides on vasopressin release. In: Gash DM, Boer GJ, eds. Vasopressin. New York: Plenum Publishing, 1987:275. 47. Renaud LP, Bourque CW. Neurophysiology and neuropharmacology of hypothalamic magnocellular neurons secreting vasopressin and oxytocin. Prog Neurobiol 1991; 36:131. 48. Mason WT, Ho YW, Eckenstein F, Hatton GI. Mapping of cholinergic neurons associated with rat supraoptic nucleus: combined immunocytochemical and histochemical identification. Brain Res Bull 1983; 11:617. 49. Kjaer A, Larsen PJ, Knigge U, et al. Histamine stimulates c-fos expression in hypothalamic vasopressin, oxytocin, and corticotropin releasing hormone-containing neurons. Endocrinology 1994; 134:482. 50. Atkins VJ, Bealer SL. Hypothalamic histamine release, neuroendocrine and cardiovascular responses during tuberomammillary nucleus stimulation in the conscious rat. Neuroendocrinology 1993; 57:849. 51. Lindvall O, Bjorklund A, Skagerberg G. Selective histochemical demonstration of dopamine terminal systems in rat di- and telencephalon: new evidence for dopaminergic innervation of hypothalamic neurosecretory nuclei. Brain Res 1984; 306:19. 52. Levey AI. Molecules of the brain. Hosp Pract (Off Ed) 2000; 35(2):41 and 51.

CHAPTER 9 PHYSIOLOGY AND PATHOPHYSIOLOGY OF THE ENDOCRINE BRAIN AND HYPOTHALAMUS Principles and Practice of Endocrinology and Metabolism

CHAPTER 9 PHYSIOLOGY AND PATHOPHYSIOLOGY OF THE ENDOCRINE BRAIN AND HYPOTHALAMUS PAUL E. COOPER Hypothalamic-Pituitary Unit Embryology Functional Neuroanatomy Hypophysiotropic Hormones Functions of the Nonendocrine Hypothalamus Temperature Regulation Appetite Regulation Emotion and Libido Autonomic Functions Biologic Rhythms Memory Sleep Hypothalamic Syndromes Chapter References

The brain and endocrine system have been linked since the 1940s, when the hypothalamus was first assigned a central role in the control of anterior pituitary secretion.1,2 This chapter discusses the functional neuroanatomy of the hypothalamic-pituitary unit, as well as the important nonendocrine functions of the hypothalamus.

HYPOTHALAMIC-PITUITARY UNIT EMBRYOLOGY Between the third and fourth weeks of embryonic development, a longitudinal groove, the sulcus limitans, appears in the lumen of the neural tube. This sulcus divides the alar (dorsal) plate from the basal (ventral) plate. The basal plate plays no part in the development of the hypothalamus or pituitary, participating only in the formation of nervous tissue caudal to the diencephalon. At ~5.5 weeks, the alar plate, in the region that gives rise to the diencephalon, develops a longitudinal groove contiguous with the sulcus limitans. This groove, the hypothalamic sulcus, divides the alar plate into a dorsal portion, which gives rise to the thalamus, and a ventral portion, which gives rise to the hypothalamus3 (Fig. 9-1A and Fig. 9-1B).

FIGURE 9-1. A, Cross section of a human embryo (3.5 weeks) through the first cervical somite. B, Cross section of a human embryo (5.5 weeks) in the region of the diencephalon. C, Midsagittal section of the hypothalamus of a human embryo (11 weeks).

During the fifth week of embryonic life, the anterior pituitary begins to form as a diverticulum (Rathke pouch) of the buccal cavity. It expands dorsally to join the basal portion of the forebrain (see Fig. 9-1C) so that by the eighth week, the posterior pituitary, which begins as a diverticulum of the floor of the third ventricle, has contacted the Rathke pouch and been invested by it. By the 11th week, the cavity of the Rathke pouch becomes flattened and loses its connection with the buccal cavity.3 Remnants of its attachment to the buccal cavity form the craniopharyngeal duct, residual cells of which persist in the posterior lobe of the pituitary, the hypophysial stalk, and the basisphenoid. Traditionally, these residual cells are thought to be the origin of craniopharyngioma; however, because craniopharyngiomas occasionally develop in the sella or along the craniopharyngeal duct itself, this idea on the origin of these tumors has been questioned. Careful examination of the superior pharynx in infants and adults almost always reveals a pharyngeal hypophysis at the buccal end of the craniopharyngeal duct. It is composed mainly of undifferentiated epithelial cells with a few chromophobes and chromophils, although rarely this residual pharyngeal hypophysial tissue has been reported to produce excessive pituitary hormone secretion. FUNCTIONAL NEUROANATOMY Neuroanatomically, the traditional borders of the hypothalamus are the lamina terminalis (rostrally); the posterior edge of the mammillary body (caudally); the hypothalamic sulcus (dorsally); the floor of the third ventricle (ventrally); and the internal capsule, basis pedunculi, and subthalamus (laterally). The hypothalamus usually is divided into three regions: the chiasmatic (preoptic region), the tuberal region, and the mammillary complex. Of these three, the first two contain the neuronal groups and tracts that are of most significance in neuroendocrine regulation (Table 9-1).

TABLE 9-1. Hypothalamic Regions and Nuclei of Neuroendocrine Interest

Histologically, the hypothalamus contains two types of neurons: large (magnocellular) and small (parvicellular). Magno-cellular neurons are of two classes: those that secrete arginine vasopressin (AVP) and neurophysin II, and those that secrete oxytocin and neurophysin I. AVP often is colocalized with dynorphin or angiotensin II, whereas oxytocin frequently is colocalized with cholecystokinin (CCK), corticotropin-releasing hormone (CRH), metenkephalin, or proenkephalin. Parvicellular neurons

contain a variety of neuropeptides or biogenic amines, many of which are colocalized. The parvicellular neurons of the hypothalamus that are found in the immediate periventricular region of the third ventricle, from the preoptic area to the mammillary bodies, project to the median eminence, where their neurosecretory products are released to influence anterior pituitary function. Functionally discrete groups of parvicellular neurons have been located in the anterior preoptic region, the paraventricular nucleus, and the arcuate nucleus. The more medial area of the hypothalamus has collections of neurons that receive inputs from the brainstem and the limbic system. They, in turn, project to the periventricular zone, the limbic system, the brainstem, and the spinal cord in a circuit that is suited to the coordination of endocrine, autonomic nervous, and more complex behavioral responses to ensure homeostasis. In the lateral hypothalamic area is found the medial fore-brain bundle, a neural pathway that connects the brainstem reticular formation with the septum and the limbic system and also has input into the periventricular secretory neurons. HYPOPHYSIOTROPIC HORMONES The anterior pituitary is known to produce six peptide hormones (Table 9-2). The secretion of each of these is under the control of one or more hypothalamic neuropeptides and several other classic neurotransmitters. A comprehensive review of these intricate interactions has been published elsewhere.4

TABLE 9-2. Anterior Pituitary and Hypophysiotropic Hormones

In primates, a few neurons containing gonadotropin-releasing hormone (GnRH) are found in clusters above the optic chiasm, medial to the supraoptic nuclei, but most are found in the medial basal hypothalamus.5 Their axons project to the median eminence, where GnRH is released in a pulsatile fashion. In humans, GnRH-positive neurons are found in the arcuate nucleus.4 The GnRH-containing neurons of the medial basal hypothalamus have an intrinsic firing pattern that results in the pulsatile release of GnRH. Norepinephrine, epinephrine, serotonin, acetylcholine, and N-methyl D-aspartic acid all stimulate GnRH release, as do angiotensin II (probably through its action on norepinephrine) and neuropeptide Y.4 Opioid peptides and g-aminobutyric acid (GABA) inhibit GnRH release.4 CRH-containing neurons are found in the medial portion of the paraventricular nucleus. From there, they project to the median eminence, where they release CRH in a pulsatile fashion (see Chap. 8). CRH release is stimulated by norepinephrine, epinephrine, serotonin, glutamine, aspartamine, acetylcholine, angiotensin II, and neuropeptide Y, and is inhibited by GABA, AVP, opioid peptides, and substance P.4 Neurons containing thyrotropin-releasing hormone (TRH) are found in the periventricular portion of the paraventricular nucleus and in a similar location in the anterior hypothalamus. They project to the median eminence through the lateral retrochiasmatic area.5 TRH release is responsive to triiodothyronine (T3) and thyroxine (T4) levels. It also is stimulated by norepi-nephrine, dopamine, and serotonin, and is inhibited by GABA and opioid peptides.4 Prolactin release is inhibited by dopamine and stimulated by TRH and vasoactive intestinal peptide (VIP). Dopamine-containing neurons are found in the arcuate nucleus and the preoptic ventricular nucleus.6 They project into hypothalamic regions rich in TRH and somatostatin. VIP-containing neurons are found in the paraventricular nucleus and project to the median eminence. They also are found in the suprachiasmatic nucleus. Neurons containing growth hormone–releasing hormone are found in the arcuate nucleus, just lateral to the median eminence. Their nerve terminals project to the median eminence and pituitary stalk.7,8 The release of growth hormone–releasing hormone is stimulated by norepinephrine and serotonin.4 Neurons containing somatostatin (growth hormone release– inhibiting hormone) are distributed widely in the central nervous system (see Chap. 169). In the hypothalamus, they are localized to neurons in the paraventricular nucleus. Their axons travel laterally and ventrally toward the optic chiasm and then caudally to the median eminence. The arcuate and ventromedial nuclei receive somatostatinergic innervation from other sources. Many somatostatin-containing nerve terminals are found in the ventromedial arcuate complex, the suprachiasmatic nucleus, the ventral premammillary nuclei, and the organum vasculosum of the lamina terminalis.5,9 HYPOTHALAMIC NEURAL CONNECTIONS Hypothalamic afferent pathways carry important information on emotion and visceral function. Ascending visceral afferents convey data to the hypothalamus from baroreceptors, volume receptors, and taste receptors. These inputs reach the hypothalamus over poorly defined pathways that relay in the reticular formation and midline thalamic nuclei. Somatosensory information also reaches the hypothalamus. The limbic system has rich afferent connections to the hypothalamus. The medial fore-brain bundle serves as an important integrative pathway between the brainstem and the limbic system. The amygdala sends fibers to the hypothalamus through the stria terminalis, and the fornices relay information from the hippocampal formation to the mammillary bodies of the hypothalamus. The mammillary bodies also receive input from and provide input to the anterior thalamic nuclei through the mammillothalamic tract and, therefore, are indirectly connected to the cingulate cortex. Finally, afferent hypothalamic connections from the dorsomedial thalamic nucleus and direct corticohypothalamic connections from the orbital surface of the frontal lobe are found.10 Two major hypothalamic efferent pathways are the dorsal longitudinal fasciculus and the mammillotegmental fasciculus. The fibers of the dorsal longitudinal fasciculus terminate in the dorsal motor nucleus of the vagus, the salivatory and lacrimal nuclei, the intermediolateral cell column of the thoracolumbar cord, and the sacral autonomic area. In addition, efferents from this tract go to various brainstem motor nuclei connected with eating and drinking, and to spinal cord motor neurons that participate in the shivering that raises body temperature. The fibers of the mammillotegmental fasciculus end in the raphe nuclei of the midbrain and pons. Finally, the mammillothalamic fasciculus contains two-way connections with the anterior nuclei of the thalamus.10 Through these rich connections, the hypothalamus monitors and influences body functions, preserving the constancy of the internal milieu. SUPRACHIASMATIC NUCLEUS The suprachiasmatic nucleus is responsible for generating circadian rhythms in all mammals, including humans. It is composed of neurons that contain AVP, VIP, neuropeptide Y, and neurotensin. A marked seasonal variation exists in suprachiasmatic nucleus cell numbers and volumes in humans, with summer values being approximately half those found in the fall. Hypothalamic tumors that involve the anterior hypothalamus disturb circadian rhythms in humans.11,12 The volume of the suprachiasmatic nucleus and the number of cells is similar in men and women, although the shape of the nucleus is different.13 Homosexual males have been reported to have a suprachiasmatic nucleus that is 1.7 times larger than that in heterosexual males and contains 2.1 times as many cells.14 The functional implication of this finding, if confirmed, remains to be determined. SEXUALLY DIMORPHIC NUCLEUS The sexually dimorphic nucleus, or intermediate nucleus, is found in the preoptic area. It contains twice as many cells in young adult males as in comparable adult

females.15 Cell numbers in the nucleus are similar in males and females at birth and not until ~4 years of age can any difference be detected. No difference in cell number in this nucleus is observed when the brains of homosexual and heterosexual males are compared.14,16 The third nucleus of the interstitial nuclei of the anterior hypothalamus has been found to be larger in males than in females, and is approximately half as large in homosexual males as in heterosexual males.17 This finding, which has not been confirmed, suggests yet another hypothalamic area that is dimorphic with respect to gender as well as sexual orientation. BLOOD SUPPLY OF THE HYPOTHALAMIC-PITUITARY UNIT The hypothalamus receives its arterial blood supply from the circle of Willis, the internal carotid and posterior cerebral arteries. Apart from blood that drains into the pituitary gland, most blood from the hypothalamus enters the basal vein of Rosenthal through numerous small venous plexuses in and around the hypothalamus. The blood supply of the pituitary gland is more complex (see Chap. 11). The posterior pituitary (neurohypophysis) is supplied by blood directly from the inferior hypophysial artery and drains into the inferior hypophysial veins. At its most rostral portion, in the median eminence, the neurohypophysis is supplied by the superior hypophysial arteries; free communication exists between the blood supply of the median eminence and that of the posterior pituitary. Apart from the most superficial layers of the gland, which are supplied by small capsular arteries, the anterior pituitary gland lacks a direct arterial blood supply. Blood is supplied to the median eminence by the superior hypophysial arteries, and the major venous drainage of the median eminence is through the portal veins to the anterior pituitary. Blood then drains from the anterior pituitary into the posterior pituitary, from which it enters the cavernous sinus or returns to the median eminence. This anatomic configuration means that blood from the median eminence, rich in hypothalamic factors and neurotransmitters, is directed toward the anterior pituitary, where it influences the secretion of pituitary hormones. Whether blood from the anterior pituitary, rich in pituitary hormones, returns to the median eminence to participate in feedback control and other possible effects on the brain is debated.18,19 ANATOMY OF THE PARAPITUITARY AREA Disorders of the hypothalamic-pituitary unit may be seen clinically because of neuroendocrine dysfunction or because of symptoms caused by compression of local structures. To fully understand the symptoms produced by compression, one must understand the anatomy of the parapituitary area (Fig. 9-2).

FIGURE 9-2. Anatomic relationships of the pituitary fossa and cavernous sinus. The lateral wall of the sella turcica is formed by the cavernous sinus. The sinus contains the carotid artery, two branches of the fifth cranial nerve (ophthalmic [Ophth.] and maxillary [Max.]), the third nerve (oculomotor), the fourth nerve (trochlear), and the sixth nerve (abducens). The optic chiasm and optic tract are located superior and lateral, respectively, to the pituitary. (A, artery; N, nerve.) (From Martin JB, Reichlin S, Brown GM, eds. Clinical neuroendocrinology. Philadelphia: FA Davis, 1987:447.)

The pituitary gland rests in the sella turcica. Inferiorly, it is bounded by the sphenoid sinus. Superiorly, it is separated from the cranial cavity by a double layer of dura mater, the diaphragma sellae; through this passes the pituitary stalk. Normally, the dura tightly surrounds the stalk. If the opening is larger or if the intracranial pressure is increased, however, the arachnoid membrane may herniate into the sella, displacing the pituitary gland peripherally and enlarging the sella—the empty sella syndrome (see Chap. 11 and Chap. 17). The diaphragma sellae is pain sensitive; stretching of this tissue by pituitary enlargement may give rise to frontotemporal headaches. The optic chiasm lies 3 to 10 mm above the pituitary fossa. In 90% of persons, the chiasm is partly or completely above the diaphragma. Of the remaining 10% of persons, approximately half have an anteriorly placed chiasm (“prefixed”) and half have a posteriorly placed chiasm (“postfixed”).20 This anatomic variability accounts for the lack of visual field abnormalities in some patients with a large suprasellar extension of a pituitary tumor. Compression of the optic chiasm by a pituitary tumor usually affects the crossing fibers in the chiasm. These come from the nasal portions of the retina, which serve the temporal fields. Thus, the typical visual field abnormality that occurs in chiasmatic compression is a bitemporal hemianopsia. Because the most inferiorly placed fibers in the optic chiasm carry information from the superior visual fields, tumors pushing on the chiasm from below tend to cause superior quadrantanopsias of the bitemporal variety, whereas lesions pushing on the chiasm from above cause inferior bitemporal quadrantanopsias. Unfortunately, several exceptions to these rules do occur. Lesions in and around the optic chiasm tend to produce incongruous visual field defects (i.e., of a different configuration in each eye; see Chap. 19). Immediately lateral to the pituitary gland are the cavernous sinuses. Within each is found the carotid artery and its sympathetic plexus; cranial nerves III, IV, and VI; and the ophthalmic and maxillary divisions of the trigeminal nerve. Sudden expansion of a pituitary tumor into the cavernous sinus may produce cranial nerve palsy. Conversely, an aneurysm of the cavernous portion of the carotid artery can erode laterally and mimic the plain skull radiographic appearance of a pituitary tumor and even lead to mild hypopituitarism.

FUNCTIONS OF THE NONENDOCRINE HYPOTHALAMUS TEMPERATURE REGULATION For the body to function efficiently, its temperature must be maintained within narrow limits. Wide variations beyond this range can result in serious metabolic derangement and death. The hypothalamus ensures that the heat gained by the body from metabolic activity, and in some circumstances from the environment, is balanced by the heat lost.21,22 The preoptic anterior hypothalamus contains thermosensitive neurons that monitor the temperature of blood (Fig. 9-3). Experiments have shown that serotonin (5-hydroxytryptamine) released in this area stimulates hypothalamic heat production centers. This effect is blocked by norepinephrine or epinephrine.

FIGURE 9-3. Hypothalamic temperature regulation mechanisms. The preoptic anterior hypothalamic area functions as a thermostat and contains mechanisms for regulation of heat loss. The posterior hypothalamus integrates heat production mechanisms. Lesions of the preoptic anterior hypothalamic area cause hyperthermia; lesions of the posterior hypothalamus cause hypothermia or poikilothermia. (Modified from Myers RD. Ionic concepts of the set-point for body temperature. In: Lederis K, Cooper KE, eds. International symposium on recent studies of hypothalamic function, Calgary, 1973. Basel: S Karger, 1974:371; and from Cooper PE, Martin JB.

Neuroendocrinologic diseases. In: Rosenberg R, ed. Comprehensive neurology. New York: Raven Press, 1991:608.)

The caudolateral portion of the hypothalamus is insensitive to changes in body temperature, but the regulatory center that determines the normal setpoint of 37°C is located here. The injection of acetylcholine-like substances into this region causes profound and long-lasting hypothermia. It also is through this area that the main pathways controlling heat loss and heat conservation travel to the midbrain, pons, medulla, and spinal cord. Intraventricular injection of the neuropeptides mammalian bombesin, neurotensin, TRH, somatostatin, and b-endorphin decreases body temperature and thus implicates these substances in thermoregulation. CRH appears to be an important mediator of thermogenesis in response to serotonin and its agonists and to cytokines. No peptide or group of peptides, however, has been singled out as the physiologic regulator of body temperature.20 In some experiments, prostaglandin E2 has increased body temperature. Hypothalamic injury after head trauma or cerebral infarction can produce prolonged hypothermia resulting either from a change in the setpoint or from an impairment of heat production mechanisms. Paroxysmal hypothermia has been described in a few patients. These persons appear to have a temporary alteration in setpoint, with the body temperature falling to 32°C or lower. The hypothermia may last for minutes to days and is associated with fatigue, decreased alertness, hypoventilation, and even cardiac arrhythmia. The loss of body heat is caused by increased sweating and vasodilation. Although the paroxysmal nature of these episodes has suggested an epileptic etiology, the attacks are not prevented by use of anticonvulsants.21 Paroxysmal hyperthermia can occur in some conditions. Acute damage to the preoptic anterior hypothalamus from surgery, subarachnoid hemorrhage, or cerebral infarction can lead to profound impairment of heat loss mechanisms, and the resulting hyperthermia can be lethal. Cyclic hyperthermia has been seen in some patients, but the neuropathologic substrate for this condition is unknown. Some have responded to therapy with phenytoin. Sustained hyperthermia probably is not seen in hypothalamic dysfunction. Reported case studies of prolonged hyperthermia attributed to hypothalamic dysfunction have not excluded an underlying malignancy or unrecognized infection. Cyclic hypothermia, responsive to treatment with anticonvulsant agents, such as clonidine or cyproheptadine, has been described.23 Large lesions in the posterior hypothalamus or lesions in the brainstem that damage the hypothalamic outflow tracts may result in poikilothermia, a condition in which the body temperature varies with environmental temperature. Most affected patients have hypothermia, although in hot, humid conditions, hyperthermia may be a problem.21 During fever, the body's temperature setpoint is elevated, although the ability to regulate temperature around the new setpoint is normal.24 In response to an infection or other cause of inflammation, the body's inflammatory cells—primarily monocytes—release cytokines,25 which act at the hypothalamus to cause fever.24 Interleukin-1 (IL-1) releases phospholipases in the hypothalamus that, in turn, release arachidonic acid from plasma membranes. Arachidonic acid causes a rise in prostaglandin E, which raises the body temperature setpoint. Treatment with acetylsalicylic acid or acetaminophen to reduce fever probably affects this process. Animal studies suggest that the action of IL-1 occurs in the preoptic anterior hypothalamus through the reduction of the sensitivity of “warm-sensitive” neurons, allowing the body to tolerate a higher temperature. Once this new setpoint has been established, the hypothalamus uses normal physiologic mechanisms to maintain body temperature by peripheral vasoconstriction, reduced sweating, and, if necessary, increases in heat production through shivering. Tumor necrosis factor (TNF) is another cytokine that alters the setpoint in the hypothalamus and increases the production of IL-1 locally, in the hypothalamus. Most TNF seems to be produced in macrophages stimulated by bacterial endotoxin. Inter-leukin-6 (IL-6) and interferon-g also act directly at the hypothalamus to raise the setpoint. The exact role of the cytokines in regulating body temperature and in causing fever is unclear because complex interactions exist among these compounds. IL-1, for example, stimulates its own production and interferon-g stimulates IL-1 production, whereas interleukin-4 suppresses the production of IL-1, TNF, and IL-6. IL-1 production also is inhibited by glucocorticoids and prostaglandin E. The fulminant hyperthermia (malignant hyperthermia) that can occur during anesthesia is not hypothalamic in origin. Rather it results from excessive muscle contraction caused by an abnormality of the muscle membrane. Another syndrome of hyperthermia is the neuroleptic malignant syndrome. This condition, characterized clinically by hyperthermia, rigidity of skeletal muscles, autonomic instability, and fluctuating levels of consciousness, has been associated with the use of major tranquilizers, rapid withdrawal from treatment with dopaminergic agents (e.g., L -dopa or bromocriptine), and, less commonly, the use of tricyclic antidepressants. The common denominator in the syndrome seems to be an alteration in dopamine function in the hypothalamus. Treatment consists of discontinuing the use of neuroleptics, providing general support, and administering anticholinergics for mild cases, bromocriptine (5 mg orally or nasogastrically four times daily) for more severe cases, and benzodiazepines or dantrolene (2–3 mg/kg per day intravenously to a maximum of 10 mg/kg per day) for resistant cases.26,27 APPETITE REGULATION In most animals, the body is able to balance its intake of food and output of energy to maintain body weight. It is the hypothalamus that receives inputs from the periphery that either stimulate or inhibit the intake of food, and it is the hypothalamus that, likewise, sends signals to other parts of the brain to influence endocrine, autonomic, and motor nervous system function.28 The interaction between the limbic system and the hypothalamus is critical in translating the need for food into behaviors such as hunting and stalking. The destruction of both ventromedial nuclei in the rat or the cat markedly increases food intake for a few days. As the animal becomes obese, the overeating decreases; if it is then fasted back to ideal weight and given free access to food, the animal again increases its food intake until it becomes obese. Lesions of the ventromedial nuclei, however, do not produce a pure syndrome of obesity. During the phase of overeating, the animals often are irritable and aggressive, becoming lethargic and passive when the increased weight is achieved. The pituitary gland is not necessary for the weight gain because hypophysectomized animals also become obese. If hyperphagia is prevented by tube feeding, the animals still become obese. Hyperinsulinemia has been observed in lesioned animals within minutes of surgery; if this is prevented by lesioning the pancreatic B cells with streptozocin, the obesity and hyperphagia are prevented. Classic neurophysiology explained the hyperphagia in animals with a lesion of the ventromedial nuclei on the basis of unopposed activity of a hypothalamic “feeding center” in the lateral hypothalamus. Implantation of electrodes in this lateral area causes, after stimulation, marked increases in food intake, whereas destruction of this area causes aphagia, even in animals with concomitant ventromedial nuclei lesions. The limbic system plays an important role in appetite. One suggested interaction is that the urge to eat arises in the hypothalamus and the limbic structures modify food intake through a discriminative function. Bilateral lesions in the amygdala cause prolonged aphagia and adipsia. Stimulation of this same area in the fed animal does not cause increased eating, but intake increases if the stimulation is performed while the animal is eating. Cells in the ventromedial hypothalamus also monitor blood glucose levels and coordinate the hypothalamic response to hypoglycemia. Although hypothalamic lesion studies have shown that blood glucose is decreased on stimulation of the anterior and tuberal regions medially, acute hypothalamic damage tends to produce hyperglycemia by activation of the sympathetic nervous system and a resulting glycogenolysis in the liver. Another important effect is the elevation of growth hormone, a contrainsulin hormone, which causes glucose levels to rise. In patients deficient in growth hormone, on either a hypothalamic or a pituitary basis, profound hypoglycemia can occur and may even be a presenting symptom, especially in children. Many of the features of lateral hypothalamic lesions in animals have been thought to result from damage to the nigrostriatal bundle, a dopamine-containing tract connecting the substantia nigra to the basal ganglia. Lesions in this tract cause anorexia and weight loss, whereas lesions in the ventral adrenergic bundle, a norepinephrine-containing tract originating in the locus coeruleus, cause hyperphagia and obesity. Serotonin is thought to inhibit eating, whereas GABA agonists have the opposite effect. Of interest has been the putative role of neuropeptides in the regulation of appetite. The gastrointestinal peptides CCK, bombesin, and glucagon can inhibit feeding behavior through an action on the vagal nucleus. Neuropeptide Y can cause hypothalamic obesity by inhibiting sympathetic drive and stimulating insulin release.29 An important putative stimulator of appetite is b-endorphin. One potential mechanism for this action is its ability to stimulate the release of insulin. Other neurotransmitters have been assumed to play a role in the control of appetite. Experimentally, the following substances decrease appetite: norepinephrine, serotonin, CCK, neurotensin, bombesin, TRH, naloxone, somatostatin, and VIP. The following substances increase appetite: dopamine, GABA, b-endorphin, enkephalin, and neuropeptide Y.30 (Also see Chap. 125.)

Increased oxidation of fatty acids raises levels of 3-hydroxy-butyrate, which can reduce food intake through an action at the level of the hypothalamus. The ob, db, and fa genes have all been implicated in the regulation of feeding in animals through production of a circulating factor. In the case of the ob gene, this circulating factor is leptin. Genetically obese mice have a leptin deficiency that, when corrected, leads to a reduction in food intake with a resulting fall in body weight. Leptin levels have been found to be high in obese humans. This suggests an insensitivity to endogenous leptin in these individuals.31 A discussion of the role of other peptides such as uncoupling proteins, agouti protein, melanocortin receptor isoforms, melanin-concentrating hormone, and the proteins responsible for the tub and fat monogenic mouse models of obesity is beyond the scope of this chapter but has been reviewed.32 Hypothalamic tumors that cause hyperphagia, aggressive behavior, and the development of marked obesity have been described. Similar clinical syndromes can be seen in patients with hypothalamic damage caused by radiation therapy or encephalitis. For many years, the syndromes of anorexia nervosa and bulimia nervosa have been considered by many to be purely psychiatric; however, the finding of reduced serotonin levels in the cerebrospinal fluid of patients with bulimia nervosa, the low cerebrospinal fluid levels of norepinephrine in patients with anorexia nervosa, and the enhanced secretion of CCK-8-S in patients with anorexia nervosa suggest that neurotransmitter or neuropeptide abnormalities could be responsible for at least some of the clinical features of these syndromes.33 The question remains whether these changes are a result of the condition or its cause (see Chap. 128). EMOTION AND LIBIDO The anatomic substrate for emotion is widespread and not confined to any single area of the brain; the frontal and temporal lobes, the limbic system, and the hypothalamus all participate in emotion. The hypothalamus is thought to play an important role in the integration and expression of emotion, especially sexual and aggressive behaviors.34 The hypothalamus is necessary for angry behavior in cats. If all cerebral tissue rostral to the tuberal region is removed, angry behavior still can be induced by minor stimuli, but if the remaining hypothalamus then is removed, this activity no longer can be provoked. Electrical stimulation of the caudal hypothalamus in the cat can elicit rage reactions. Bilateral anteromedial hypothalamic lesions cause normally friendly cats to become aggressive, as do bilateral lesions in the ventromedial nuclei, or electrical stimulation in the perifornical region or the periaqueductal gray area between the third and fourth ventricles. The limbic system appears to exert a tonic inhibition on the perifornical region of the hypothalamus. Damage to these inhibitory pathways or stimulation in this area results in angry behavior. Ablation of certain areas in the hypothalamus produces fearful behavior, and stimulation of certain hypothalamic areas in the dorsal hippocampal formation and the septal region produces pleasure reactions. For such reactions to occur, the basal telencephalon and thalamus must be intact. Some humans report a pleasurable or glowing feeling with electrical stimulation in the septal area; others report a feeling of sexual gratification. Primates from which the amygdala, piriform cortex, and part of the hippocampal formation have been removed bilaterally exhibit several behavioral disturbances, including hyper-sexuality, loss of discrimination of taste for thirst-quenching liquids, hyperoral behavior, and loss of awareness of harmful or painful objects (Klüver-Bucy syndrome). Rage attacks have occurred in patients with lesions in the caudal hypothalamus, the subthalamus, and even the midbrain. They also have been induced by manipulation of the hypothalamus during surgery. Fear and rage in patients with hypothalamic disorders usually occur in situations that normally would be associated with these same emotions, but in lesser degrees. In those cases examined pathologically, lesions of the basal portion of the brain, especially the descending pathways to the hypothalamus from the cerebral cortex, or the ventromedial nuclei, usually are found. Stimulation of the posterior portion of the hypothalamus in humans can arouse feelings of fear and horror. Lesions of the mammillary bodies and vicinity are associated with drowsiness, somnolence, apathy, and indifference. Psychosurgery directed at the medial posterior hypothalamus or the caudolateral region has caused apathy in previously aggressive persons. Euphoria rarely is seen in adults with hypothalamic disease but is seen in children as part of the diencephalic syndrome. This syndrome most commonly manifests in infancy as failure to thrive; a glioma of the anterior hypothalamus usually is found. Children with this condition are emaciated despite normal or excessive food intake, and are described as jovial and as having excessive energy. Approximately 50% of affected children have nystagmus. A tendency to hypoglycemia also may bring these children to medical attention. Despite growth failure, growth hormone levels usually are elevated. Computed tomography demonstrates abnormalities in most cases (see Chap. 18). Normal libido is the product of an interaction between hypothalamic and extrahypothalamic sites. Usually, hypothalamic damage leads to decreased GnRH levels and reduced libido. Hypersexuality is a rare accompaniment of hypothalamic disease and may be seen with or without increased libido.35 Paroxysmal hypersexuality consisting of sexual urges, genital sensations, and orgasm has been observed in association with temporal lobe tumors or epilepsy but not with primary hypothalamic disease. Altered sexual preference has been described in association with hypothalamic lesions.36 Epileptiform activity in the temporal lobe has been linked to violent behavior. This is extremely rare, however. In a review of 5400 patients with epilepsy, violent behavior during a seizure was found in only 19 cases.37 Evidence is accumulating to suggest that the hypothalamus plays a role in the symptoms of fibromyalgia and chronic fatigue syndrome.38 AUTONOMIC FUNCTIONS The hypothalamus plays an important role in the integration of the functions of the autonomic nervous system.39 CARDIOVASCULAR FUNCTIONS Signals of cardiovascular status are sent to the brain from baroreceptors and chemoreceptors at the carotid sinus and aortic arch, and from pressure/volume receptors in the atrium. This information is fed into the nucleus of the tractus solitarius, as well as into cells of the paramedian reticular nuclei in the medulla. In the medulla, these afferent signals are modulated by inputs from higher centers, particularly the hypothalamus. The main sympathetic outflow to the heart begins in the paraventricular nucleus of the hypothalamus. Some of the paraventricular neurons project to the intermediolateral cell column of the spinal cord (the site of origin of preganglionic sympathetic neurons), whereas others project to the dorsal motor nucleus of the vagus where they can influence parasympathetic output to the heart.40 Stimulation of the anterior hypothalamus, particularly the preoptic area, causes bradycardia, hypotension, and decreased baroreflex activity. Such stimulation also can lower the threshold for ventricular fibrillation and cause a variety of electrocardiographic changes. Such changes are common in patients with subarachnoid hemorrhage. Hypothalamic disease also may cause hypertension, and the disrupted autonomic function that occurs in subarachnoid hemorrhage can cause myocardial necrosis, an effect blocked by sympathetic blockade. The hypothalamus receives input from the nucleus of the tractus solitarius and relays information from higher cortical centers to brainstem vasomotor centers. The many and varied cardiovascular phenomena that occur in response to emotion, especially hypertension and cardiac arrhythmias, probably are mediated by the hypothalamus.40 The anterior portion of the ventral third ventricle has been shown, by lesion studies, to be a site at which body fluid homeostasis and arterial blood pressure are regulated. Angiotensin II levels reflect changes in serum osmolality or sodium concentration. In addition, the area receives sensory neural input from the kidney, carotid sinus, and other baroreceptors. Output from this area can induce vasoconstriction (through the sympathetic nervous system), and abnormal function of this region has been implicated in the pathogenesis of hypertension.41

RESPIRATION Acute pulmonary edema has been described in association with numerous injuries to the central nervous system, such as increased intracranial pressure, epilepsy, and hypothalamic injury.41a Classically, neurogenic pulmonary edema was defined as normal pulmonary capillary wedge pressures with increased protein content of edema fluid in the lung. One hypothesis has been that noncardiac pulmonary edema can be caused by excessive sympathetic outflow from the hypothalamus. Virtually all cases of this type of pulmonary edema, however, are accompanied by system hypertension; therefore, when the experimental physiologic data are reviewed, separation of cardiac from noncardiac causes of edema is impossible. Most of the data support centrally induced systemic hypertension as the cause of the syndrome.42 GASTROINTESTINAL FUNCTIONS Under normal resting conditions, gastric motility is relatively autonomous. Higher cortical centers, including the neocortex and limbic lobe, may influence gastric activity through their connections with the hypothalamus. Afferent signals go to the gut through the vagus nerves. Electrical stimulation of areas of the rostral hypothalamus causes a prompt fall in gastric pH, an effect that is prevented by vagotomy. Stimulation of the tuberal and caudolateral areas of the hypothalamus causes reduced secretion of gastric acid that is of slower onset. This is unaffected by vagotomy but is prevented by bilateral adrenalectomy. Stimulation and ablation studies in the monkey have shown that cortical areas affect peristalsis, the volume of gastric secretion, and its enzyme and acid content. Similar effects can be seen with stimulation or lesioning of the amygdala. Bilateral lesions in the anterolateral hypothalamus increase basal gastric acid secretion, an effect that is abolished by interruption of the fibers of the fornix and medial forebrain bundle. This observation suggests that the limbic lobe and frontal neocortex normally exert a tonic, inhibitory effect on basal gastric secretion. Prolonged hypothalamic stimulation often results in hemorrhage and ulceration in the gastric mucosa of monkeys and dogs, whereas sympathectomy prevents the hemorrhage but not the ulceration. Lesions anywhere from the rostral hypothalamus to the region of the vagal nuclei in the medulla oblongata may cause acute hemorrhagic erosions of the gastric mucosa; extensive ulceration of the lower esophagus and stomach; and acute perforation of the esophagus, stomach, and duodenum. Such lesions are most likely to occur after damage to or pressure on the hypothalamus, particularly its tuberal region. BIOLOGIC RHYTHMS Biologic rhythms are ubiquitous in the animal and plant kingdoms, occurring in single cells, tissues, organs, individual animals, and populations (see Chap. 6). Their periods range from milliseconds to years. Many endocrine rhythms are circadian, having a period ~24 hours long.43 Ultradian rhythms, with periods shorter than 24 hours, and infradian rhythms, with periods longer than 24 hours, certainly exist, but the neural mechanisms responsible for these rhythms are debated. Circadian rhythms have been studied extensively. Integrity of the retinohypothalamic projection, which terminates in the suprachiasmatic nucleus, appears to be essential for the light entrainment of circadian rhythms, and the suprachiasmatic nucleus appears essential for circadian rhythmicity. Ablation of the suprachiasmatic nucleus is associated with the loss of all circadian rhythms. The exact functioning of the suprachiasmatic nucleus in the generation of circadian rhythms is unknown, but at least two coupled oscillators appear to exist. These oscillators cause the circadian variations seen in brain monoamines, plasma amino acids, corticotropin and cortisol, growth hormone, prolactin, vasopressin, aldosterone, insulin, glucose, and sex steroids, as well as pineal activity, body temperature, and autonomic function.44 MEMORY The hypothalamus is closely linked to nervous structures associated with memory function: the reticular formation, hippocampal formation, and other limbic structures.45,46 Although lesions in the mammillary bodies are found in virtually all cases of Wernicke-Korsakoff syndrome (retrograde amnesia, confabulation, apathy),47 the memory deficit in this condition has been thought to correlate best with lesions in the dorsomedial nucleus of the thalamus. Nonetheless, a case report of a brain-injured patient who was studied using magnetic resonance imaging suggests that, at least in trauma, hypothalamic injury alone, without accompanying thalamic injury, can cause marked, relatively focal, memory deficits.48 Many patients with hypothalamic lesions exhibit disturbances of short-term memory, with sparing of immediate recall and long-term memory. Bilateral hippocampal lesions are associated with severe memory disturbance, although some controversy exists about whether involvement of the fornices, the major outflow tract of the hippocampal formation, produces permanent memory loss. Evidence suggests that fornix transection can cause wide-ranging memory disturbance in humans (see Chap. 176).49 SLEEP The region of the hypothalamic-midbrain junction appears to play an important role in sleep and wakefulness. Two types of patients were seen in the encephalitis pandemics50 of 1917 and 1920: those with prolonged somnolence and ophthalmoplegia, and those with agitation and hyperkinesia. Pathologically, somnolence correlated with damage to the tegmentum of the midbrain, and agitation was seen in patients with anterior hypothalamic lesions. Damage to the hypothalamic-midbrain junction by multiple sclerosis, abscess, or infarction has been described as causing hypersomnia or inversion of the sleep-waking cycles. Some studies support the concept of a “sleep center” in the anterior hypothalamus that, if damaged, causes hyperactivity and sleeplessness, whereas the posterior hypothalamus appears to be involved in the production of rapid eye movement sleep. The ascending reticular formation participates in normal wakefulness. Damage to it results in coma (complete unresponsiveness). Alternatively, the hypothalamus seems to be the generator of normal sleep-wake cycles. Damage to the hypothalamus causes either excessive wakefulness or somnolence (i.e., an unresponsive state from which the patient can be aroused, at least temporarily).

HYPOTHALAMIC SYNDROMES Several clinical syndromes have been attributed to hypothalamic dysfunction.51 Previously, patients have been described who had autonomic overactivity associated with tumors in the region of the hypothalamus and third ventricle. This was termed diencephalic epilepsy, although its epileptic nature is in doubt because electroencephalograms do not show seizure activity during spells and the condition does not respond to administration of anticonvulsants. Glioma of the anterior hypothalamus in early childhood produces a clinical syndrome characterized by profound emaciation despite normal or excessive food intake, excess energy, and euphoria—the diencephalic syndrome of infancy (see Chap. 18). Kleine-Levin syndrome is characterized by episodes of somnolence followed by hyperactivity, irritability, and increased appetite. Adolescent boys are most commonly affected. The attacks occur every 3 to 6 months and last days to weeks. The cause is unknown, and the specific hypothalamic pathology has not been determined.52 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

Fortier C. Obituary: Geoffrey Wingfield Harris. Endocrinology 1972; 90:851. Bajusz E, Ernst A. Scharrer. Editorial. Neuroendocrinology 1965; 1:65. Hamilton WJ, Boyd JD, Mossman HW. Human embryology: prenatal development of form and function, 3rd ed. Cambridge: W Heffer & Sons, 1962:315. Palkovits M. Neuropeptides in the brain. In: Martini L, Ganong WF, eds. Frontiers in neuroendocrinology, vol 10. New York: Raven, 1988:1. Page RB. Neuropharmacology of anterior pituitary control. In: Barrow DL, Selman WR, eds. Neuroendocrinology, vol 5. Concepts in neurosurgery. Baltimore: Williams & Wilkins, 1992:31. Müller EE, NisticD G. Brain messengers and the pituitary. San Diego, CA: Academic Press, 1989:1. Jacobowitz DM, Schulte H, Chrousos GP, Loriaux DL. Localization of GRF-like immunoreactive neurons in rat brain. Peptides 1984; 4:521. Merchenthaler I, Thomas CR, Arimura A. Immunocytochemical localization of growth hormone releasing factor (GHRF)-containing structures in the rat brain using anti-rat GHRF serum. Peptides 1984; 5:1071. Beal MF, Mazurek MF, Martin JB. A comparison of somatostatin and neuropeptide Y distribution in monkey brain. Brain Res 1987; 10:405. Barr M, Kiernan JA. The human nervous system: an anatomical viewpoint, 6th ed. Philadelphia: JB Lippincott, 1993:181. Schwartz WJ, Bosis NA, Hedley-Whyte ET. A discrete lesion of ventral hypothalamus and optic chiasm that disturbed the daily temperature rhythm. J Neurol 1986; 233:1. Cohen RA, Albers HE. Disruption of human circadian and cognitive regulation following a discrete hypothalamic lesion: a case study. Neurology 1991; 41:726. Swaab DF, Hofman MA, Lucassen PJ, et al. Functional neuroanatomy and neuropathology of the human hypothalamus. Anat Embryol (Berl) 1993; 187:317. Swaab DF, Fliers E, Partiman TS. The suprachiasmatic nucleus of the human brain in relation to sex, age and senile dementia. Brain Res 1985; 342:37.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

Swaab DF, Hofman MA. An enlarged suprachiasmatic nucleus of the human brain in homosexual men. Brain Res 1990; 537:141. Swaab DF, Hofman MA. Sexual differentiation of the human hypothalamus: ontogeny of the sexually dimorphic nucleus of the preoptic area. Dev Brain Res 1988; 44:314. LeVay S. A difference in hypothalamic structure between heterosexual and homosexual men. Science 1991; 253:1034. Bergland RM, Page RB. Can the pituitary secrete directly to the brain? (Affirmative anatomical evidence.) Endocrinology 1978; 102:1325. Page RB. Pituitary blood flow. Am J Physiol 1982; 243:E427. Banna M. Terminology, embryology and anatomy. In: Hankinson J, Banna M, eds. Major problems in neurology, vol 6. Pituitary and parapituitary tumors. London: WB Saunders, 1976:1. Cooper PE. Disorders of the hypothalamus and pituitary gland. In: Joynt RJ, ed. Clinical neurology, vol 3. Philadelphia: JB Lippincott, 1993:1. Rothwell NJ. CNS regulation of thermogenesis. Crit Rev Neurobiol 1994; 8:1. Kloos RT. Spontaneous periodic hypothermia. Medicine 1995; 74:268. Cooper KE. The neurobiology of fever: thoughts on recent developments. Annu Rev Neurosci 1987; 10:297. Reichlin S. Neuroendocrine-immune interactions. N Engl J Med 1993; 329:1246. Rosenberg MR, Green M. Neuroleptic malignant syndrome—review of response to therapy. Arch Intern Med 1989; 149:1927. Gratz SS, Levinson DF, Simpson GM. The treatment and management of neuroleptic malignant syndrome. Prog Neuropsychopharmacol Biol Psychiatry 1992; 16:425. Bray GA, Fisler J, York DA. Neuroendocrine control of the development of obesity: understanding gained from studies of experimental animal models. Front Neuroendocrinol 1990; 11:128. York DA. Lessons from animal models of obesity. Endocrinol Metab Clin North Am 1996; 25:781. Cezayirli RC, Robertson JT. Pharmacologic modulation of hypothalamic control of appetite. In: Givens JR, ed. The hypothalamus. Chicago: Year Book Medical Publishers, 1984:115. Considine RV, Sinha MK, Heiman ML, et al. Serum immunoreactive-leptin concentrations in normal-weight and obese humans. New Engl J Med 1996; 334:292. Bessesen DH, Faggioni R. Recently identified peptides involved in the regulation of body weight. Semin Oncol 1998; 25(Suppl 6):28. Leibowitz SF. Brain monoamine projections and receptor systems in relation to food intake, diet preference, meal patterns, and body weight. In: Brown GM, Koslow SH, Reichlin S, eds. Neuroendocrinology and psychiatric disorders. New York: Raven Press, 1983:383. Van de Poll NE, Van Goozen SH. Hypothalamic involvement in sexuality and hostility: comparative psychological aspects. Prog Brain Res 1992; 93:343. Plum F, VanUitert R. Nonendocrine diseases and disorders of the hypothalamus. In: Reichlin S, Baldessarini RJ, Martin JB, eds. The hypothalamus. New York: Raven Press, 1978:415. Miller BL, Cummings JL, McIntyre H, et al. Hypersexuality or altered sexual preference following brain injury. J Neurol Neurosurg Psychiatry 1986; 49:867. Delgado-Escuetta AV, Mattson RH, King L, et al. The nature of aggression during epileptic seizures. N Engl J Med 1981; 305:711. Crofford LJ, Demitrack MA. Evidence that abnormalities of central neurohormonal systems are key to understanding fibromyalgia and chronic fatigue syndrome. Rheum Dis Clin North Am 1996; 22:267. Grossman A, ed. Neuroendocrinology of stress. Baillieres' clinical endocrinology and metabolism, vol 1. London: Baillieres Tindall, 1987:247. Valeriano J, Elson J. Electrocardiographic changes in central nervous system disease. Neurol Clin 1993; 11:257. Samuels MA. Neurally induced cardiac damage. Neurol Clin 1993; 11:273.

41a. Keegan MT, Lanier WL. Pulmonary edema after resection of a fourth ventricle tumor: possible evidence for a medulla-mediated mechanism. Mayo Clin Proc 1999; 74:264. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52.

Simon RP. Neurogenic pulmonary edema. Neurol Clin 1993; 11:309. Aschoff J. Circadian rhythms: general features and endocrinological aspects. In: Kreiger DT, ed. Endocrine rhythms. New York: Raven Press, 1979:1. Moore RY. The anatomy of central neural mechanisms regulating endocrine rhythms. In: Kreiger DT, ed. Endocrine rhythms. New York: Raven Press, 1979:63. Zola-Morgan S, Squire LR. Neuroanatomy of memory. Annu Rev Neurosci 1993; 16:547. Wilson MA, McNaughton BL. Reactivation of hippocampal ensemble memories during sleep. Science 1994; 265:676. Pitella JE, Giannetti AV. Morphometric study of the neurons in the medial mammillary nucleus in acute and chronic Wernicke's encephalopathy. Clin Neuropathol 1994; 13:26. Dusoir H, Kapur N, Byrnes DP, et al. The role of diencephalic pathology in human memory disorder: evidence from a penetrating paranasal brain injury. Brain 1990; 113:1695. Gaffan EA, Gaffan D, Hodges JR. Amnesia following damage to the left fornix and to other sites: a comparative study. Brain 1991; 114:1297. Creisler CA, Klerman EB. Circadian and sleep-dependent regulation of hormone release in humans. Recent Prog Horm Res 1999; 54:97. Martin JB. Neurologic manifestations of hypothalamic disease. Prog Brain Res 1992; 93:31. Mukaddes NM, Alyanak B, Kora ME, Polvan O. The psychiatric symptomatology in Kleine-Levin syndrome. Child Psychiatry Hum Dev 1999; 29:253.

CHAPTER 10 PINEAL GLAND Principles and Practice of Endocrinology and Metabolism

CHAPTER 10 PINEAL GLAND RUSSEL J. REITER Morphology and Innervation Gross Anatomy Innervation Histology Indoleamine Metabolism Melatonin Serotonin Influence of Hormones and Drugs on Melatonin Secretion Plasma Levels of Melatonin and Its Metabolism Factors Influencing Plasma Melatonin Rhythm Light–Dark Cycle Age and Sexual Maturation Aversive Stimuli Possible Sex Steroid Feedback Clinical Implications Sexual Maturation Ovulatory Relationships Psychological Interrelationships Sleep Jet Lag and other Biologic Rhythm Disorders Tumor Growth Free Radical Scavenging Conclusion Chapter References

MORPHOLOGY AND INNERVATION GROSS ANATOMY The pineal gland, so called because of its resemblance to a pine cone (Latin pineas), is a multifaceted endocrine organ whose secretory products directly or indirectly affect every organ and cell in the body. Embryologically, it is derived from neuroectoderm as an outgrowth of the diencephalon. In adults, it is attached to the posterodorsal aspect of the diencephalon, and its proximal portion is invaginated by the pineal recess of the third ventricle (Fig. 10-1). The gland weighs ~130 mg, with great individual variation seen, and is ~1 cm long. The blood supply to the gland is derived from the posterior choroidal branches of the posterior cerebral arteries. The pineal gland has a copious blood supply. In advanced age, it may acquire calcium deposits (corpora arenacea, acervuli), which are visible on skull radiographs.

FIGURE 10-1. Midsagittal view of the human brain showing the relationship of the pineal gland to other neural structures.

INNERVATION The pineal gland is innervated by sympathetic axons and by axons coming directly from the brain.1 Postganglionic sympathetic axons arrive in the pineal from perikarya located in the superior cervical ganglia. The cell bodies of the preganglionic fibers that end in the ganglia are located in the intermediolateral cell column of the upper thoracic cord. The neurons in the intermediolateral cell columns receive terminals from, among other sources, perikarya in the hypothalamus, possibly in the paraventricular nuclei; these nuclei are connected to the supra-chiasmatic nuclei, which receive a prominent input from the ganglion cells of the retinas through the retinohypothalamic tract. Through this complex series of neurons, the retinas are functionally related to the pineal gland. Sympathetic innervation of the pineal is essential for the rhythmic metabolism of indoleamines and for the pineal's endocrine functions. The sympathetic neurotransmitter mediating the cyclic production of pineal indoleamines is norepinephrine. Besides sympathetic innervation, the pineal gland receives axons directly from the brain that enter through the stalk. The function of these fibers relative to the physiology of the pineal is unknown. HISTOLOGY The major cellular elements within the pineal gland are the pinealocytes, which are arranged into cords or follicles separated by connective tissue septa. With advancing age, the septa become more prominent. Ultrastructurally, the pinealocytes have one to several cytoplasmic processes that typically terminate near the many capillaries perfusing the gland. The nerve endings in the pineal gland usually lie close to the pinealocyte processes in the perivascular spaces, but true morphologic synapses are not obvious.

INDOLEAMINE METABOLISM MELATONIN Within the pinealocyte, tryptophan, an amino acid taken up from the blood, is metabolized to a variety of potential hormones (Fig. 10-2).2 The compound that has been most thoroughly investigated is N-acetyl-5-methoxytryptamine, or melatonin. Tryptophan is initially metabolized to serotonin (5-hydroxytryptamine); the concentration of this monoamine in the pineal gland exceeds that in any other organ. Especially during the daily dark period, serotonin is N-acetylated by the enzyme serotonin N-acetyltransferase to form N-acetylserotonin.3 The activity of N-acetyltransferase increases many-fold during darkness (Fig. 10-3). N-acetylserotonin is converted to melatonin by the action of hydroxyindole-O-methyltransferase. Once formed, melatonin is rapidly secreted into the blood vascular system and, as a consequence, plasma melatonin levels typically are highest at night, when pineal melatonin production is greatest.

FIGURE 10-2. Interactions of postganglionic sympathetic neurons with the endocrine unit of the pineal gland, the pinealocyte, and the conversion of tryptophan to a variety of indole products. Within the pinealocyte, cyclic adenosine monophosphate (AMP) acts as a second messenger to stimulate the activity of serotonin N-acetyltransferase (NAT), primarily during darkness. Serotonin also is acted on by hydroxyindole-O-methyl-transferase (HIOMT) to form 5-methoxytryptamine, and by monoamine oxidase to form 5-hydroxyindole-3-acetylaldehyde. Melatonin is a major secretory product of the pineal gland. (Modified from Reiter RJ, MacLeod RM, Müller EE, eds. Neuroendocrine perspectives, vol 3. New York: Elsevier, 1984:350.)

FIGURE 10-3. Pineal, blood, and urinary rhythms of various constituents. Methods are available for estimating blood and urinary levels of melatonin and its metabolites. The shaded area in the right half of the figure represents the daily dark period. (Ac Co A, Acetyl coenzyme A; CoA, coenzyme A; S Ad M, S-adenosyl methionine; S Ad H, S-adenosyl homocysteine; CNS, central nervous system.) (From Reiter RJ. Pineal indoles: production, secretion and actions. In: Muller CE, MacLeod RM, eds. Neuroendocrine perspectives, vol 13. Amsterdam: Elsevier, 1984:358.).

SEROTONIN Hydroxyindole-O-methyltransferase also acts directly on serotonin, converting it to 5-methoxytryptamine (see Fig. 10-2), a compound that has been proposed as a pineal hormone. Serotonin also is metabolized by monoamine oxidase; the product is eventually converted to 5-methoxytryptophol and 5-methoxyindoleacetic acid, the former of which possibly is released from the pineal.

INFLUENCE OF HORMONES AND DRUGS ON MELATONIN SECRETION Most of what is known about the interactions of the postganglionic neurotransmitter norepinephrine with the pinealocyte comes from observations in animals, although the relationships probably are similar in humans. Norepinephrine is released into the synaptic cleft, primarily during the daily dark period, after which it acts primarily on b-adrenergic receptors on the pinealocyte membrane, where it stimulates cyclic adenosine monophosphate production and, eventually, N-acetyltrans-ferase activity and melatonin synthesis (see Fig. 10-2).4 Alpha receptors on the pinealocyte membrane also may be involved in mediating the nocturnal rise in melatonin. In humans, the peripheral infusion of isoproterenol, a b-receptor agonist, does not increase blood levels of melatonin as readily as it does in some animals. In addition, orciprenaline and L -dopa are ineffective in altering circulating melatonin in humans. The inability of isoproterenol to stimulate human pineal melatonin production may relate to the relative insensitivity of the pinealocyte b-receptors to the agonist throughout most of the day; animal studies have shown that either the number of receptors or their affinity for the ligand is greatly increased at night.5 Circulating norepinephrine may be relatively ineffective in promoting pineal melatonin production because the sympathetic nerve endings in the pineal gland have an active uptake mechanism for the catecholamine, thereby protecting the pinealocytes from circulating norepinephrine. Other drugs that seem incapable of altering blood melatonin levels in humans include scopolamine, amphetamine, thyrotropin-releasing hormone, luteinizing hormone–releasing hormone, and desaminocys-D-arg-vasopressin.5 Both clonidine and dexamethasone have been reported to lower plasma levels of melatonin. In general, pineal melatonin production seems to be out of the usual endocrine feedback loop.

PLASMA LEVELS OF MELATONIN AND ITS METABOLISM Plasma levels of melatonin seem to reflect closely the amount being synthesized and secreted by the pineal gland (see Fig. 10-3).6 When pineal melatonin production increases at night, plasma levels rise accordingly. In the few individuals in whom the pineal has been surgically removed because of tumors in the epithalamic region, plasma levels of melatonin either are severely depressed or are undetectable.7 In animals, pinealectomy is associated with very low or nonmeasurable amounts of melatonin in plasma. The nocturnal rise in plasma melatonin levels has been measured by bioassay, radioimmunoassay, and gas chromatography–mass spectrometry.7a Depending on the technique used, daytime levels of plasma melatonin range from undetectable to 20 pg/mL. At night, values may increase to 300 pg/mL; however, with the most specific methods, nighttime levels typically are 3 log standard deviations below the mean of normal euthyroid persons.46 More than 95% of sera from thyrotoxic persons would be expected to have TSH levels below the lower limit of normal in an assay meeting this criterion. Most commercial IMAs appear capable of meeting that standard.47 Clinical chemists have traditionally reported as the analytic sensitivity or detection limit of an assay the lowest TSH level statistically distinguishable from zero concentration by measurement of replicate samples in the same assay run. Such a definition of sensitivity may be clinically misleading, because a single measurement of a specimen containing TSH at the analytic threshold concentration would yield a result of zero 50% of the time. Furthermore, analytic sensitivity is a function of within-assay variance and does not assess the reliability of between-assay comparisons, which are more likely to be clinically useful in the diagnosis and treatment of an individual patient. As an alternative to analytic sensitivity, the proposal has been made that assay sensitivity be characterized by a criterion based on interassay variability. Specifically, the “lower limit of interassay quantitative measurement,”48 or functional sensitivity, of an assay is the TSH concentration for which the interassay coefficient of variation is less than some preestablished threshold (generally 20%) to permit reliable quantitative comparisons between specimens measured in different assay runs. A “generational” classification of TSH assays has been proposed49 (Table 15-2). Each generation is approximately an order of magnitude more sensitive than the previous one. Both second- and third-generation assays distinguish suppressed TSH levels in hyperthyroidism from normal values. However, the third-generation assay further differentiates partial suppression of basal TSH concentrations in some patients with subclinical hyperthyroidism, nonthyroid illness, glucocorticoid therapy, and other clinical states (Table 15-3) from the more complete suppression of basal TSH concentrations in overt hyperthyroidism.50

TABLE 15-2. Properties of Thyroid-Stimulating Hormone (TSH) Assays

TABLE 15-3. Clinical Influence on Basal Thyroid-Stimulating Hormone (TSH) and TSH Response to Thyrotropin-Releasing Hormone

Measurement of TSH is frequently used as the initial, and sometimes sole, thyroid function test.51 This approach is generally sensitive and specific in the ambulatory population, in which the finding of a normal TSH level is strong evidence that a patient is euthyroid, and an abnormal TSH has a high likelihood of being due to thyroid dysfunction. However, abnormally high or low TSH values (compared to those of an ambulatory euthyroid control population) are frequently noted in hospitalized patients as a result of the effects of nonthyroid illness, acute psychiatric illness, or glucocorticoid therapy.52,53 Therefore, diagnoses of hypothyroidism or hyperthyroidism in hospitalized patients should be based on clinical evaluation, measurement of free thyroid hormone levels, and other indices of thyroid function, rather than on TSH measurement alone. The TRH stimulation test has been used in the assessment of mild thyroid dysfunction and in the functional evaluation of the hypothalamic–pituitary–thyroid axis.54 The test entails measuring serum TSH levels at baseline and after the bolus intravenous administration of TRH. A dose-response relation between administered TRH and peak TSH levels is observed for TRH doses of 6.25 to 400 µg. In clinical practice, a TRH dose sufficient to produce a maximal TSH response is used. The peak TSH response occurs 20 to 40 minutes after TRH administration. If a primary thyroid disorder is suspected, measurement of TSH levels at baseline and at 20 or 30 minutes after TRH administration is sufficient (Fig. 15-4). If pituitary or hypothalamic dysfunction is suspected, TSH measurements should be continued for 2 to 3 hours at 30- to 60-minute intervals. During the first 5 minutes after TRH administration, side effects may include mild nausea, headache, a transient rise in blood pressure, light-headedness, a peculiar taste sensation, a flushed feeling, and urinary urgency.54

FIGURE 15-4. Typical thyroid-stimulating hormone (TSH) responses to the administration of thyrotropin-releasing hormone (TRH) under different conditions. Basal TSH is suppressed in overt thyrotoxicosis and does not respond to TRH. The blunted TSH response in patients with nonthyroid illness may be similar to the response in patients with subclinical hyperthyroidism. Patients with subclinical hyperthyroidism or nonthyroid illness may have a basal TSH below the detection threshold for second-generation TSH assays.

TRH testing may be viewed as a means of amplifying and detecting small differences in TSH secretion due, most importantly, to alterations in serum T4 or T3 concentrations. A slight excess of T4 or T3 blunts or completely blocks the TSH response to TRH, whereas small decrements in thyroid hormone levels enhance the response. The peak TSH response to TRH is proportional to the basal serum TSH level.55 Expressed as a multiple of the basal TSH, the peak TSH is a mean 8 to 9.5 times higher. However, considerable variability is seen in individual responses (range: 3- to 23-fold increment in euthyroid persons). In addition to thyroid hormone concentrations, other factors can alter the TSH response to TRH (see Table 15-3). In patients with severe illnesses, the TSH response to TRH is likely to be diminished. Cortisol and other neurohumoral factors secreted in response to stress, malnutrition, and the administration of glucocorticoids or dopamine all may contribute to the blunted TSH response. In patients with a subnormal basal TSH level, however, the magnitude of the TSH response to TRH does not distinguish those in whom TSH is suppressed as a result of intercurrent illness from those in whom it is suppressed as a result of partial suppression of the pituitary by slight excess of free thyroid hormone (i.e., patients with autonomous thyroid nodules or exogenous thyroid hormone suppression).55 In general, if the basal TSH level exceeds the functional sensitivity threshold of the assay system and, therefore, can be accurately measured, then measurement of the TRH-stimulated TSH level does not provide additional information regarding the cause of the suppressed TSH. With the improvement in sensitivity of TSH assays, the TRH-stimulation test is not generally required in the evaluation of primary hypothyroidism or hyperthyroidism with suppressed TSH. However, the TRH-stimulation test may be useful in the evaluation of central hypothyroidism, in the rare patient with TSH-dependent hyperthyroidism, and in some patients with functioning pituitary tumors that respond to TRH stimulation (e.g., acromegaly). THYROID-STIMULATING HORMONE IN PRIMARY HYPOTHYROIDISM The basal serum TSH concentration is increased in patients with intrinsic failure of the thyroid gland (primary hypothyroidism) of all causes. The magnitude of the increase is roughly proportional to the severity of disease.56 In general, basal TSH levels show a better inverse correlation with serum T4 levels than with serum T3 levels; this is because of the importance of the uptake of serum T4 and its intracellular deiodination as a source of T3 in the thyrotrope.29 In some persons, elevated TSH levels may be found with normal serum T3 concentrations but decreased serum T4 values. Such findings are common in patients with early thyroid gland failure, patients with mild iodine deficiency, and some patients with Graves disease who have been given long-term antithyroid drug treatment. The isolated elevation of serum TSH levels with normal serum T4 and T3 concentrations in the absence of clinical signs or symptoms of hypothyroidism has been termed subclinical hypothyroidism. This condition has an overall prevalence of 2% to 7% and is particularly common in older women. Overt hypothyroidism develops at a rate of 5% to 10% per year in persons with elevated TSH levels and positive antithyroid antibodies.57 Measurement of the serum TSH level remains a sensitive test for the diagnosis of primary hypothyroidism in severely ill patients because basal TSH levels, although sometimes partially attenuated, remain higher than normal during intercurrent illness in patients with moderate or severe hypothyroidism. The diagnosis of hypothyroidism should be confirmed by measurement of the free T4 value, however. Patients with mildly elevated TSH and normal free T4 concentrations generally should undergo repeated thyroid function testing after discharge from the hospital to confirm the diagnosis of hypothyroidism. The TSH response to TRH is exaggerated in patients with primary hypothyroidism. However, TRH testing should not be needed in the evaluation of suspected hypothyroidism if the basal serum TSH level is elevated. THYROID-STIMULATING HORMONE IN HYPERTHYROIDISM Circulating TSH is suppressed in hyperthyroidism of all causes except in the rare patients with TSH-dependent thyrotoxicosis. In clinically hyperthyroid patients, the basal TSH level measured with a third-generation assay (functional sensitivity of 40% of adenomas present as vision symptoms, and 70% show field defects. HEADACHE Chronic headaches, mild or severe, are noted by most patients with pituitary adenomas.13 Headache symptoms are variable but may be most marked where advancing pituitary adenomas are restrained by a taut diaphragma sellae. In acromegaly, chronic headaches may indicate paranasal sinus enlargement, with or without active sinusitis. In children, headaches usually are not thoroughly evaluated until nausea, vomiting, and behavioral changes occur, at which point an intracranial lesion should be suspected. Additionally, obesity, precocious or delayed sexual development, somnolence, and diabetes insipidus should alert the clinician to hypothalamic dysfunction (see Chap. 9, Chap. 18 and Chap. 26). SENSORY PHENOMENA Peculiar sensory phenomena may be noted by patients with bitemporal field defects, resulting in a nonparetic form of strabismus or diplopia and in difficulty with visual tasks requiring depth perception (e.g., use of a screwdriver, threading a needle, and the like). Loss of portions of normally superimposed binocular fields results in the absence of corresponding points in visual space (and on the retina) and subsequently diminished fusional capacity. In essence, the patient has two free-floating nasal hemi-fields with no interhemispheral linkage to keep them aligned. Vertical and horizontal slippage produces doubling of images, gaps in otherwise continuous visual panorama, and steps in horizontal lines. A series of 260 patients with pituitary adenoma included some degree of double vision preoperatively in 98 patients, but a demonstrable ocular palsy was present in only 14.13 Additionally, without temporal fields, objects beyond the point of binocular fixation fall on nonseeing nasal hemiretina, so that a blind area exists with extinction of objects beyond the fixation point. EYE MOVEMENT DISORDERS The association of extraocular muscle palsies with chiasmal field defects implies involvement of the structures in the cavernous sinus, usually a sign of rapid expansion of a pituitary adenoma. Only rarely is tumor diagnosis delayed sufficiently for obstruction of the ventricular system to occur, with elevation of intracranial pressure and unilateral or bilateral sixth nerve palsies (Fig. 19-5 and Fig. 19-6).

FIGURE 19-5. Diagram illustrating nerves and structures in and near the cavernous sinus. (a, artery; ant., anterior; n., nerve; post., posterior; sup., superior.) (From Glaser JS. Topical diagnosis: the optic chiasm. In: Glaser JS. Neuro-ophthalmology, 3rd ed. Philadelphia: Lippincott, 1999:408.)

FIGURE 19-6. A, Acromegalic patient with involvement of left oculomotor nerve. B, Patient had reported a gradual onset of left-sided ptosis and diplopia with a mild headache.

Patients with large parasellar masses may display “seesaw” nystagmus, with alternate depression and extorsion of one eye and elevation and intorsion of the other. This results from expansion of tumors within the third ventricle, with secondary midbrain compression, rather than from chiasmal involvement per se.14 PALLOR OF THE OPTIC DISCS Pallor of the optic discs, although an anticipated physical sign of chiasmal interference, is not a prerequisite for diagnosis. In a series of 156 cases of pituitary tumors, optic atrophy was found in only 155 of 312 eyes (50%)15, disc pallor was present in 34% of the adenomas studied at the Mayo Clinic.12 Optic atrophy may not be present even when vision symptoms have lasted as long as 2 years.16 Furthermore, extensive field loss in chiasmal syndromes may be associated with normal or

minimally pale discs. Therefore, it is unwise to rely on the presence of optic atrophy as an indication of chiasmal interference; such findings are corroborative evidence at best, after the visual fields have been carefully evaluated. It is somewhat risky to predict on the basis of disc appearance the ultimate level of vision anticipated after chiasmal decompression. As a rule, the more atrophic the disc, the less likely is the return of function in defective areas of field. However, vision recovery to surprisingly good levels may occur despite relatively advanced disc pallor. Papilledema is a less common finding and results from large tumors compressing the third ventricle, with resultant hydrocephalus. EVALUATION OF FIELD DEFECTS The importance of establishing that the vertical meridian forms the central border of the defect (see Fig. 19-4) is paramount in distinguishing chiasmal interference from deficits that mimic temporal hemianopia. Such mimicking conditions include tilted discs (congenital inferior scleral crescents); nasal sector retinitis pigmentosa; bilateral cecocentral scotomas; papilledema with greatly enlarged blind spots; nutritional optic neuropathy; retinal inflammatory disease; and redundant overhanging upper lid tissue.17 The endocrinologist should be aware of the general pattern of evolution of chiasmal field defects and of the useful screening procedures that are appropriate in the context of the general physical examination. Visual field defects caused by chiasmal interference may be characterized by the following17: depressions initially occur in the central 20-degree field (therefore, exploration of the periphery is time-consuming and insensitive) (see Fig. 19-4); the central edge of the defect is aligned along the vertical meridian that passes through the point of visual fixation, in one or both eyes; the defect is more readily apparent with red targets than, for example, white against black (Fig. 19-7); and the loss of monocular acuity (“reading vision”) that is not explained by uncorrected refractive error or ocular disease (cataract, macular degeneration, etc.) is evidence of prechiasmal optic nerve compression until proved otherwise. Three additional caveats are that normal visual fields do not preclude the presence of a parachiasmal lesion (e.g., a microadenoma does not cause field defects); the screening technique described does not replace other formal perimetric techniques, such as Goldmann or automated perimetry; and the assessment of visual fields in no way obviates the need for anatomic studies, that is, computed tomography (CT) or magnetic resonance imaging (MRI). Only limited information is available on plain films; subtle changes in the bony structures of the sella may be easily overlooked by the inexperienced. Radiologic measurements of the sella, whether linear or volumetric, are not intrinsically important. In marginal cases, such measurements are unreliable; in obvious cases, they are superfluous; and in neither case is the problem of suprasellar extension solved. Thus, plain films have little, if any, use today in the diagnostic evaluation of patients with parasellar disease.

FIGURE 19-7. A, Finger-counting fields in adults. Four quadrants of each eye should be tested. Patient may name or hold up same number of fingers. Simultaneous finger counting may bring out subtle hemian-opic defect. B, Hand comparison in hemianopic depression appears “darker,” “in shadow,” or “blurred.” (From Glaser JS. Topical diagnosis: the optic chiasm. In: Glaser JS. Neuro-ophthalmology, 3rd ed. Philadelphia: Lippincott, 1999:31–33.)

The advent of thin-section CT and of gadolinium-enhanced MRI has greatly simplified the diagnosis of intracranial mass lesions. Pneumoencephalography no longer is indicated, and cerebral angiography is reserved for those patients in whom the precise configuration of neighboring vessels may be relevant to transsphenoidal or transcranial surgical approaches to intrasellar or suprasellar tumors. The question of “full” or “empty” sella is clearly settled by coronal-section MRI (Fig. 19-8).

FIGURE 19-8. Empty sella syndrome. An MRI shows flattened remnant of pituitary (arrowheads) on sellar floor; chiasm above (arrow).

DEVELOPMENTAL CHIASMAL DYSPLASIAS The chiasm is infrequently the site of developmental anomalies; at times, it is related to malformation of other diencephalic mid-line structures, including the third ventricle. Embryonic dysgenesis may result in the abnormal development of the primitive optic vesicles with resulting unilateral or bilateral anophthalmos (absent globe) or useless microphthalmic cysts. Such gross ocular abnormalities may occur in isolation or may be associated with a spectrum of neural defects, including major malformations that preclude survival. Perhaps of greater clinical import are the more subtle ocular dysplasias that accompany those anterior forebrain malformations compatible with long life. Certain specific congenital anomalies of the optic disc may indicate the presence of otherwise occult forebrain malformations. DE MORSIER SYNDROME The clinical constellation of pituitary dysfunction and optic nerve hypoplasia (Fig. 19-9) is recognized as the de Morsier syndrome.18,19 Variable findings may include neonatal hypotonia; seizures; prolonged bilirubinemia; deficiencies of growth hormone, adrenocorticotropic hormone, or vasopressin; panhypo-pituitarism; mental retardation; or a combination of these.20 The association of this entity with midline cerebral anomalies (i.e., absence of septum pellucidum, agenesis of the corpus callosum) has been well documented.21 Additional neuropathologic findings have included extensive atrophy of the optic nerves and chiasm, aplasia of olfactory bulbs and tracts, and hypoplasia of the hypophysial stalk and infundibulum.22 Although the exact etiology of this condition has not been determined, it is well known to occur in association with maternal gestational diabetes and maternal use of phenytoin, quinine, alcohol, and lysergic acid.19 Genetic defects—including partial deletion of chromosome 1423 and mutations in the PAX324 and HESX125 genes—have been reported. Others have hypothesized that this group of defects represents a vascular disruption sequence.26

FIGURE 19-9. De Morsier syndrome. A, Hypoplastic optic discs. Note depigmented ring around discs. B, MRI, coronal view, demonstrates single ventricle with absence of septum pellucidum.

The wide clinical spectrum of patients with bilateral optic nerve hypoplasia is now well recognized. In one study, only 30 of 40 such children had coexistent central nervous system anomalies. Concomitant endocrinologic defects were associated with posterior pituitary ectopia rather than agenesis of the corpus callosum or other central nervous system abnormalities.27 Some investigators feel that endocrine function may be predicted on the basis of pituitary appearance on MRI,28 but others have documented normal-appearing glands in patients with central diabetes insipidus.29 In 35 patients with bilateral optic nerve hypoplasia, endocrine dysfunction was seen in only 27% (growth hormone deficiency and hypothyroidism being most common), and abnormal neuroimaging in 46%.30 Eighty percent of these patients were legally blind (34% had no light perception in either eye), but 10% of eyes had vision 20/60 or better. The “full-blown” syndrome (i.e., septo-optic dysplasia with panhypopituitarism) occurred in only 11.5% of these patients. An important clinical point is that children with such disorders are at risk for sudden death during febrile illnesses secondary to hypocortisolism, thermoregulatory disturbances, and dehydration.31 The occurrence of segmental superior optic nerve hypoplasia has been observed in infants born to mothers with gestational diabetes32; this, perhaps, represents a forme fruste of the de Morsier syndrome. FOREBRAIN BASAL ENCEPHALOCELES A variety of optic nerve abnormalities (hypoplasia, colobomata, peripapillary staphylomata, and the “morning glory” disc) may be associated with other developmental defects; principally, they take the form of midline craniofacial anomalies,33,34 including hypertelorism; defects of the lip, palate, and maxilla; anencephaly35 and prosencephaly36; skeletal malformations37; or combinations B thereof. Of special neuro-ophthalmologic interest is the association of optic disc malformation with forebrain basal encephaloceles. Herniated brain tissue may present as pulsating exophthalmos (spheno-orbital encephalocele usually associated with neurofibro-matosis), hypertelorism with a pulsatile nasopharyngeal mass (transsphenoidal encephalocele), or a frontonasal mass, with or without hypertelorism (frontoethmoidal encephalocele). Congenital disc anomalies, such as hypoplasia or the coloboma dysplasia variety38 (Fig. 19-10), associated with hypertelorism or other mid-facial malformation, are evidence of basal encephalocele until proved otherwise. There are also reports of isolated retinal colobomas, sparing the optic nerve, in association with sphenoethmoidal encephaloceles. The physical findings of transsphenoidal or transethmoidal basal encephalocele39 are listed as follows:

FIGURE 19-10. Transsphenoidal encephalocele. A, Large dysplastic optic disc (coloboma of nerve). B, Polytomogram demonstrates transsellar herniation of forebrain (arrows).

Midline facial anomalies: Broad nasal root, hypertelorism, mid-line lip defect, wide bitemporal skull diameter, cleft palate Nasopharyngeal mass: Midline epipharyngeal space in location pulsatile; often symptoms of nasal obstruction may be confused with a nasal polyp; rare in infancy Hypopituitarism/dwarfism, ocular: Congenital disc anomalies such as colobomatous dysplasias, chiasmal field defects, poor vision, exotropia In the evaluation of basal encephalocele, neuroradiologic imaging of the cranial base is indicated. Biopsy or attempted resection of posterior nasopharyngeal masses should be avoided because these masses invariably are encephalomeningoceles, and surgical manipulation may result in meningitis with tragic outcome. Because both optic disc hypoplasia and colobomatous dysplasia may reflect developmental defects of the chiasm, it is essential to recognize that such ophthalmologic anomalies may be a clue to underlying brain and endocrine defects. HYPOGONADISM SYNDROMES Hypogonadotropic hypogonadism (Kallmann syndrome) is an autosomal dominant disorder characterized by low levels of serum follicle-stimulating hormone and luteinizing hormone and by eunuchoid features40 (see Chap. 115). Other pituitary hormone levels are usually normal, but midline facial anomalies and unilateral renal agenesis may be present. This disorder can be associated with anosmia, hypoplasia of the nose and eyes, hypogeusia (taste deficiency),41 retinitis pigmentosa,42 and Möbius syndrome (congenital facial diplegia and horizontal gaze palsies with esotropia) with peripheral neuropathy.43,44 Hypoplastic development of the olfactory bulbs and the hypothalamus has been documented. PITUITARY DWARFISM A large series45 of 101 pituitary dwarfs revealed MRI evidence of posterior pituitary ectopia in 59. Pituitary volume did not change with hormonal therapy. There was a 3:1 male predominance, and a higher incidence of breech delivery (32% versus 7%) and congenital brain anomalies (12% versus 7%) in children with posterior pituitary ectopia. Defective induction of mediobasal brain structures in early embryonic life is the presumed cause, and mutation in the PIT-1 gene may be implicated.

PARACHIASMAL LESIONS Most chiasmal syndromes are caused by extrinsic masses: classically, pituitary adenomas, suprasellar meningiomas, craniopharyngiomas, and internal carotid artery aneurysms. Although certain patterns of vision failure may suggest the location and type of lesion, such clinical impressions often prove fallible in the face of neuroradiologic procedures (and at craniotomy). At any rate, the diagnostic evaluation of all nontraumatic chiasmal syndromes is stereotyped: to rule in or out the presence of a potentially treatable mass lesion. Inflammatory or infectious causes are extremely rare, and chiasmal involvement by head trauma or radionecrosis is uncommon. The distinction is made by age-related incidence; accompanying signs and symptoms; and typical, if not diagnostic, radiologic appearance.

OPTIC AND HYPOTHALAMIC GLIOMAS Primary astrocytic tumors of the anterior visual pathways assume two major clinical forms: the relatively benign glioma (juvenile pilocytic astrocytoma) of childhood and the rare malignant glioblastoma of adulthood. With the exception of vision loss and anatomic location, these two groups have little in common, and the assumption that the progressive malignant form stems from the static childhood form is untenable. For the indolent glioma of childhood, clinical presentation is predicated on the location and extent of the tumor. Gliomas may be separated into roughly three topographic groups: unilateral optic nerve (orbital, intracranial, or both, but not involving the chiasm); principally chiasmal mass; or simultaneous infiltration of the hypothalamus. Strictly intraorbital gliomas present as insidious proptosis of variable degree, and although vision is usually diminished, remarkably good vision function is not uncommon. Strabismus, disc pallor, or disc swelling may be observed. Progressive proptosis, even if abrupt, or increased visual deficit does not imply an aggressive activity of the tumor, a hemorrhage, or necrosis. These tumors may rarely extend intracranially and have a low (~10%) mortality rate.46 Chiasmal gliomas are more common than the isolated orbital type. These tumors present with unilateral or bilateral vision loss, strabismus, optic atrophy, and/or infantile nystagmus. The nystagmus may mimic spasmus nutans (usually unilateral or asymmetric nystagmus), complete with head nodding and torticollis, or may show a coarse, conjugate mixed horizontal–rotary pattern, especially when vision is severely defective. “Seesaw” nystagmus may also rarely be seen in these patients. Children with extensive basal tumors also show hydrocephalus and signs and symptoms of increased intracranial pressure. Hypothalamic signs include precocious puberty, obesity, dwarfism, hypersomnolence, and diabetes insipidus. Usually, the non-vision complications of extensive gliomas occur in infancy or early childhood, and onset of obstructive signs or hypothalamic involvement much beyond the age of 5 years is uncommon. Gliomas of the optic nerves and chiasm may be associated with neurofibromatosis in 20% to 40% of cases; the patients either show other characteristic stigmata or have affected relatives.47, 48 and 49 Indeed, there is good evidence that children with neurofibromatosis type 1 and chiasmal gliomas have a better long-term prognosis than those with chiasmal gliomas alone.46 Absence of neurofibromatosis, electrolyte abnormalities, and intracranial hypertension are all indicators of a poor prognosis.50,51 Treatment of benign optic gliomas in childhood is somewhat controversial, with most pediatric and neuro-ophthalmologists favoring a conservative approach. When there is documented progressive vision loss, significant tumor growth, or obstructive signs, intervention is obviously indicated. Debulking surgery, radiation therapy, and chemotherapy have all been advocated. Newer radiation techniques may allow a higher percentage of treated patients to grow normally,52 and, when chemotherapy is indicated, reports suggest success with a combined carboplatin/vincristine regimen,53 as well as oral etoposide (VP-16).54 DIENCEPHALIC SYNDROME When the hypothalamus is the site of childhood glioma, a dien-cephalic syndrome evolves.55 Findings consist of emaciation, despite adequate food intake, that develops after a period of normal growth; hyperactivity and euphoria; skin pallor (without anemia); hypotension; and hypoglycemia. Other notable signs include nystagmus and disc pallor, to which may be added sexual precocity and laughing seizures.56 Twelve cases of histologically proven opticochiasmatic glioma with diencephalic syndrome were culled from a 22-year review.57 There were 6 men and 6 women, all with “failure to thrive” but with normal linear growth; none had stigmata of neurofibromatosis. Two patients died in the immediate postoperative period, and 10 patients received radiotherapy with “reversal of their diencephalic syndrome” (weight gain, deposition of subcutaneous fat, normal development). Six of 10 are alive, 3 being considered normal, and 3 are blind, retarded, or both. Clinical evidence of bilateral optic nerve involvement was seen in 10 of these 12 cases, but it was not possible during surgery to determine the origin of these tumors. The diencephalic syndrome appears to be related to the age at which the hypothalamus becomes compressed; none of the 12 patients (and only 4% of all published cases) had onset of symptoms after 2 years of age. Craniotomy with biopsy and radiotherapy is often indicated, as tumors involving the hypothalamus appear to be larger and more aggressive than other astrocytomas arising in this region.58 RADIOLOGIC INVESTIGATION OF GLIOMAS The radiologic investigation of suspected gliomas is now sophisticated to the extent that “neuroradiologic biopsy” may obviate tissue diagnosis. The typical, but variable, findings include CT and MRI evidence of enlarged orbital or intracranial optic nerves; enlarged chiasm; a homogeneous hypothalamic mass; enlarged optic canals; and J-shaped or gourd-shaped sellae. The demonstration of such typical dysplastic changes of the sella turcica and optic canals, coupled with CT or MRI evidence of intrinsic chiasmal mass, so strongly suggests the diagnosis of glioma that histopathologic affirmation probably is superfluous and hazardous to vision. Analysis of visual fields in patients with chiasmatic gliomas has shown no consistent relationship between the pattern of field defects and the location, size, or extent of tumor; in 12 of 20 patients, the putative bitemporal pattern of chiasmal involvement was absent.59 Central scotomas or measurable depression of the central field occurred in 70% of the eyes; therefore, the absence of bitemporal hemianopia, or one of its variants, cannot be interpreted as a sign that the glioma does not involve the chiasm. CRANIOPHARYNGIOMAS Craniopharyngiomas (see Chap. 11) are developmental tumors that arise from vestigial epidermoid remnants of Rathke pouch, scattered as cell nests in the infundibulohypophysial region. These tumors are usually admixtures of solid cellular components and variable-sized cysts containing oily mixtures of degenerated blood and desquamated epithelium or of necrotic tissue with cholesterol crystals. Calcification of such debris may be radiologically detectable, a helpful diagnostic sign. In rare instances, these tumors may present in the neonate, attesting to their congenital origin.60 Craniopharyngiomas constitute 3% of all intracranial tumors (~15% in children), with two distinctive modes of presentation—in childhood and in adulthood—with a peak in the 40- to 70-year-old age range (Fig. 19-11).

FIGURE 19-11. Age distribution of craniopharyngioma. (Data from Svolos D. Craniopharyngiomas: a study based on 108 verified cases. Acta Chir Scand Suppl 1969; 403:1; Matson DD, Crigler JF. Management of craniopharyngioma in childhood. J Neurosurg 1969; 30:377; Bartlett JR. Craniopharyngiomas: an analysis of some aspects of symptomatology, radiology and histology. Brain 1971; 94:725. Note: Matson and Crigler series limited to children younger than 16 years of age.)

The symptomatology of childhood craniopharyngioma is variable, depending on the position and mass of the tumor.61,62 Frequently, progressive vision loss goes unnoticed until a level of severe impairment is reached, or unless headache, vomiting, or behavioral changes occur because of hydrocephalus. Obesity, delayed sexual development, somnolence, and diabetes insipidus attest to hypothalamic dysfunction, and other endocrinopathies may be present. Increased intracranial pressure is not uncommon, and papilledema may be observed. Some optic atrophy is usually present, but its absence does not conflict with the presence of chronic compression of the anterior visual pathways, even with severe vision loss. Suprasellar or intrasellar calcification is a rather constant radiologic finding in childhood craniopharyngiomas, occurring in 80% to >90% of affected children. Cystic areas frequently occur in craniopharyngioma but rarely in opticochiasmatic gliomas. CT scanning retains special sensitivity in diagnosis, being superior to MRI in detecting calcifications and cyst formations. However, involvement of adjacent structures is more clearly defined by MRI. In adults, defects in visual field or acuity are the initial symptoms, although increased intracranial pressure or endocrine dysfunction less frequently occur. Visual field defects often take the form of asymmetric bitemporal hemianopia or homonymous patterns, indicating optic tract involvement. Intracranial calcification is seen much

less regularly in adults than in children. The surgical therapy for craniopharyngiomas ranges from total (or at least radical) excision63 or postoperative radiotherapy after partial removal of tumor, to radiation therapy administered after simple biopsy or cyst decompression.64 Transsphenoidal decompression may be indicated for large tumors filling the sella. When possible, total removal of the tumor is ideal, but radical manipulations should not be attempted when adhesions to the optic nerves, chiasm, carotid arteries, or hypothalamus are present. The more conservative approach of simple decompression of the anterior visual pathways and relief of third-ventricle obstruction appears judicious, and postoperative radiation therapy has established efficacy. Endocrine replacement therapy is anticipated in the vast majority of cases, often for life. As with pituitary adenoma and meningioma, craniopharyngiomas may enlarge abruptly during pregnancy.65 RATHKE CLEFT CYSTS Although previously regarded as a rare lesion in the sellar area, these cysts derive from Rathke cleft, an embryonic vestige of Rathke pouch. In a series of 18 patients with this lesion,66 7 presented with visual disturbance or bitemporal hemianopia, and 7 presented with a variety of endocrine dysfunctions. Unlike craniopharyngiomas, partial removal or decompression of these cysts with one procedure is usually sufficient, and regrowth is less common. ARACHNOID CYSTS Enlarging loculations of cerebrospinal fluid (CSF) contained in arachnoidal cysts infrequently present as a chiasmal syndrome. These may arise, for example, in the floor of the third ventricle, causing chiasmal compression, a J-shaped sella, and occasional precocious puberty.67 Women with benign intrasellar cysts have been reported,68 showing bitemporal hemianopia, headache, optic atrophy, and panhypopituitarism. Another patient presented with obesity and amenorrhea but without visual defects.69 SUPRASELLAR DYSGERMINOMA Primary suprasellar dysgerminomas (atypical teratoma, “ectopic pinealoma”) are rare causes of chiasmal interference, but they constitute a more or less distinguishable clinical syndrome. These tumors likely arise from cell rests in the anterior portion of the third ventricle and are not directly related to the pineal itself, although histologically, they resemble atypical pineal teratomas. A review of 64 cases67 revealed that the classic triad consists of early diabetes insipidus; visual field loss, not necessarily of a clearly chiasmal pattern (owing to infiltration of the anterior visual pathways); and hypopituitarism. Symptoms commence at the end of the first or during the second decade of life. Girls are affected more frequently, with a peak incidence at 10 to 20 years of age. Usually, plain film radiology of the sella is normal, but MRI readily reveals the lesion. Frequently, there is growth retardation. The diagnosis is confirmed by CSF cytology, measurement of human chorionic gonadotropin, or both, but often biopsy is necessary.70 The radical excision of tumor invading the optic nerves and chiasm, infundibulum, and floor of third ventricle is not possible, but radiotherapy offers excellent palliation, if not a cure. Because subarachnoid seeding of the neuraxis is a distinct possibility, more extensive radiation may be indicated. Long-range endocrine replacement is critical. PITUITARY ADENOMAS Asymptomatic pituitary adenomas occur in >20% of pituitary glands, and some degree of adenomatous hyperplasia can be found in almost every pituitary gland.71 A postmortem study72 of pituitaries removed from 120 patients without clinical evidence of pituitary tumors revealed a 27% incidence of microadenomas, of which 41% stained for prolactin, without gender difference. To generalize, >1 in 10 people in the general population dies harboring a prolactinoma. The incessant parade of this clinical syndrome is, therefore, not surprising. Tumor of the pituitary gland is the single most common intracranial neoplasm that produces neuro-ophthalmologic symptomatology, and chiasmal interference is overwhelmingly the most frequent presentation (see Chap. 11). Strictly speaking, a microadenoma refers to a tumor that is 10 mm or less in diameter and confined to the sella. Symptomatic adenomas occur infrequently before 20 years of age but are common from the fourth through seventh decades of life. When these tumors do occur in childhood, most are asymptomatic. When symptoms are present, headache, visual field loss, and endocrinopathies are the most common. Dissimilar to adults, in children there is a definite male predominance, and many tumors are hemorrhagic.73 Histologic staining characteristics alone do not correlate well with patterns of growth or clinical symptomatology. A functional classification of pituitary adenomas, as elaborated by electron microscopy and immunohistochemistry, has replaced the previous simplistic classification of “eosinophilic, basophilic, and nonfunctioning” (see Chap. 11). SYMPTOMATOLOGY Nonocular symptoms include chronic headaches (severe or mild) in more than two-thirds of patients, fatigue, impotence or amenorrhea, sexual hair change, or other signs of gonadal, thyroid, or adrenal insufficiency (see Chap. 17). Prediagnostic signs and symptoms, affecting vision or otherwise, may exist for months to years before diagnosis is established. VISION CHANGES With pituitary tumors, vision failure may take the form of a rather limited number of field patterns. As suprasellar extension evolves, a single optic nerve may be compromised, with resultant progressive monocular vision loss in the form of a central scotoma. More frequently, as the tumor splays apart the anterior chiasmal notch, superotemporal hemianopic defects occur (Wilbrand knee, as discussed previously). However, this well-touted superior bitemporal hemianopia is almost always accompanied by minor or major hemianopic scotomas approaching the fixational area along the vertical meridian (see Fig. 19-4). Asymmetry of field defects is common, the eye with the greater field deficit also being likely to show diminished central vision. Marked asymmetry is not uncommon, such that one eye may be blind and the other may show a temporal hemianopic defect, the so-called junctional scotoma; this combination is as exquisitely localizing to the chiasm as is the classic bitemporal hemianopia. Adenomas extending posteriorly produce incongruous homonymous hemianopias by optic tract involvement; central vision usually is diminished, at least in the ipsilateral eye. In late stages, the only suggestion of the chiasmal character of field defects may be minimal preservation of the nasal field of one eye. The absence of field defects in patients undergoing evaluation for amenorrhea or a sella enlargement that is incidentally discovered does not imply the absence of an adenoma. For example, many patients with acromegaly do not show field defects, and microadenomas by definition do not escape the confines of the sella. EFFECT OF PREGNANCY The effect of pregnancy on pituitary adenomas is of interest diagnostically and therapeutically. Enlargement of preexisting pituitary tumors during the third trimester of pregnancy may occur,74 with reduction in size postpartum. That an otherwise normal pituitary gland may enlarge owing to the changes of pregnancy alone, causing symptoms affecting vision, is controversial.75 Nevertheless, a 30-week pregnant woman with an enlarged pituitary and bitemporal hemianopia that regressed spontaneously postpartum was reported76; a retrospective diagnosis of lymphocytic hypophysitis was made. DIAGNOSIS Many pituitary tumors deform the sella turcica sufficiently to be detected by plain film techniques, but, normal or otherwise, such procedures must be considered preliminary or superfluous. CT with contrast or gadolinium-enhanced MRI is mandatory when chiasmal lesions are suspected (see Chap. 20). THERAPY The rational approach to treatment of pituitary adenomas has evolved radically over the past 2 decades with the advent of thin-section CT; MRI; transsphenoidal microsurgery; hormonal assays; and dopamine agonists (e.g., bromocriptine), potent inhibitors of pituitary synthesis and release of prolactin. The choice of treatment is open to discussion, with enthusiastic advocates in each camp, but the prime consideration is the ultimate well-being of the patient. Patients with high surgical risk, especially the elderly, should not be subjected to frontal craniotomy. After uncomplicated transsphenoidal surgery alone, vision recovery approaches 90%.77 Radiation therapy, used either primarily or postoperatively, has great efficacy,78 and stereotactic radiosurgery has been shown to be effective for select patient groups.79

The administration of bromocriptine may rapidly improve vision function when prolactinomas compress the chiasm. In a study80 of 10 men with field defects caused by prolactinomas (initial prolactin level range 1535–14,200 ng/mL) who were treated with 7.5 to 30 mg per day bromocriptine, an increase in vision usually began within days of commencing therapy, and CT evidence of a decrease in tumor volume was documented somewhat later. Pregnancy apparently is not a contraindication for bromocriptine therapy.81 An extraordinary, rare complication of chiasmal herniation from shrinkage of a pituitary tumor treated with bromocriptine has been reported; recovery of vision ensued after a decrease in the dosage.82 With the advent of pergolide, another ergot-derived dopamine agonist, comes another viable treatment alternative, with apparently fewer frequent side effects of hypotension, nausea, and headache. Also, cabergoline has a very long duration of action, as well as fewer adverse effects. Quinagolide, a non-ergot long-acting prolactin inhibitor, a pure D2 agonist, is also useful.83 Finally, the long-acting somatostatin analog octreotide may be effective in the treatment of somatotropic, thyrotropic, gonadotropic, and nonfunctioning adenomas.84 In many cases, hormonal therapy of prolactinomas results in rapid improvement in vision function, often independently of decrease in tumor size. ACROMEGALY Acromegaly is the relatively rare clinical condition related to adenomatous secretion of growth hormone, with resultant hypertrophy of bones, soft tissues, and viscera (see Chap. 12). Sellar changes, when present, are indistinguishable from those caused by other adenomas. Of 1000 pituitary adenomas, 144 of 228 acromegalic patients had visual field defects.96 Possibly, this relatively high incidence of visual defects reflects delay in diagnosis in a series commenced 5 decades ago. Diabetes mellitus in acromegaly may be associated with typical retinopathy.85 Increase in corneal thickness and elevated ocular tension (glaucoma) has been reported,86 and CT scan has revealed thickened extraocular muscles.87 An unusual developmental condition with a dominant inheritance pattern, the so-called ACL (acromegaly, cutis verticis, leukoma) syndrome, has been described.88 This syndrome consists of acromegaloid features combined with severe ridging of the skin of the scalp (cutis verticis gyrata) and corneal whitening (leukoma). Pathologic examination of corneal leukoma has demonstrated a propensity for the nasal limbus, with whorl-like accumulations of disorganized collagen material and mucinous deposits.89 Signs tend to increase with age, with variable family penetrance. PITUITARY APOPLEXY Pituitary apoplexy—an acute change in adenoma volume resulting from hemorrhage, edematous swelling, or necrosis—is not rare, although the appropriate diagnosis may be elusive (see Chap. 17). Perhaps some 10% of pituitary adenomas undergo such acute or subacute changes,90 with clinical signs and symptoms including change in headache pattern (often severe frontal cephalgia), rapid drop in visual function, unilateral or bilateral ophthalmoplegia, epistaxis or CSF rhinorrhea, and other complications of blood or necrotic debris in the CSF. In a review of 320 verified pituitary adenomas,91 evidence of hemorrhage was found in 98 cases (18.1%). There was a high incidence of giant or large recurrent adenomas (41%). The mean age was 50 years (range, 17–71 years). The clinical course included acute apoplexy (7 cases); subacute apoplexy (11 cases); recent silent hemorrhages (13 cases); and old silent hemorrhages (27 cases). Sella enlargement was present in all patients. These patients need not be stuporous, but rapid deterioration and obtundation are highly suggestive. There appears to be a tendency for such events to take place in intrasellar secretory adenomas confined by a competent diaphragma sellae. Ischemic necrosis causes sudden expansion of the tumor with acute compression of neighboring structures, including the optic nerves and chiasm and the ocular motor nerves in the cavernous sinus. Although this syndrome should now be well known, delay in diagnosis is frequent. Common misdiagnoses usually include meningitis, ruptured intracerebral aneurysm, or sphenoidal mucocele. Almost all cases show abnormal sellae on plain skull series. The CT and MRI scans are typical, if not diagnostic.92 MRI and CT scans distinguish between many tissue densities, and MRI can detect the presence of blood; the finding of acute or subacute bleeding within a tumor based in an enlarged sella is highly suggestive of pituitary apoplexy. Although in a few cases (limited suprasellar extension and intact or improving vision) corticosteroid replacement and other expectant medical management may suffice, as a rule, rapid transsphenoidal decompression of the often hemorrhagic tumor should be accomplished without delay to minimize devastating visual consequences; final endocrine status is less likely to be affected. VISION ASPECTS OF THERAPY The medical, surgical, and radiation therapies of pituitary adenomas are covered elsewhere (see Chap. 21, Chap. 22, Chap 23 and Chap. 24). The present role of irradiation of pituitary adenomas is problematic, considering the palpable failure rate and question of untoward side effects. Radiation therapy does indeed appear to reduce the rate of recurrence of pituitary adenomas.93 However, optic nerve and chiasm damage have occurred secondary to radiation necrosis anywhere from 2 months to 6 years after treatment.94 Radiation retinopathy, empty sella syndrome, cranial neuropathies, and further pituitary–hypothalamic disturbances may result from radiation therapy of pituitary lesions. There are also anecdotal reports of sarcomas, gliomas, and meningiomas occurring after radiation treatment of pituitary adenomas.95 The addition of bromocriptine before, during, or after radiotherapy may be helpful in controlling tumor secretion and size until the radiation treatment reaches its maximal effect. After uncomplicated surgical decompression, visual acuity and fields may return rapidly within 24 to 48 hours or improve weekly (Fig. 19-12). Such restoration is dependent on the duration of visual morbidity and the degree of pallor of the optic discs. After surgery, if careful ophthalmoscopy reveals attrition of the retinal nerve fiber layer, corresponding field defects are permanent. For the most part, what vision returns does so by 3 to 4 months, although continued improvement to 1 year postoperatively is possible. Although fortunately the exception rather than the rule, vision loss is a well-known complication of both transsphenoidal surgery and craniotomy. Failure of vision recovery within the 24-48 hour postoperative interval is highly suggestive of occult hemorrhage in the tumor bed or from related vessels. MRI is essential and decompression may be necessary.

FIGURE 19-12. A, Preoperative automated visual fields from a 68-year-old man with a nonfunctioning pituitary adenoma. Note dense bitemporal defect. B, Same patient, 10 weeks after transsphenoidal decompression, enjoys dramatic recovery and near-normalization of visual fields. (Left, left eye; right, right eye.)

FOLLOW-UP OF TREATED PITUITARY ADENOMAS From the standpoint of detecting recurrence, the follow-up of treated adenomas has been problematic. Even as adenomas must be large initially to cause visual defects, so must recurrences be substantial before defects again evolve. Although progressive vision failure may be the incontestable impetus for reoperation, consecutive perimetry may not be counted on to reveal early tumor recurrence. One should obtain an anatomic assessment, as provided by CT scanning or MRI. Recurrence of vision failure may be caused by regrowth of tumor, arachnoidal adhesions associated with progressive empty sella syndrome, or delayed radionecrosis. Tumor recurrence is, by far, the most common mechanism of vision deterioration, but field examination alone may not make this distinction.

EMPTY SELLA SYNDROME Extension of the subarachnoid space into the sella turcica through a deficient sellar diaphragm may manifest itself clinically and radiologically as a syndrome mimicking pituitary adenoma. The empty sella may be defined as nontumorous remodeling that results from a combination of incomplete diaphragma sellae and CSF fluid pressure.96 Diaphragmal openings are common; in one study, defects >5 mm were found in 39% of normal autopsy cases.97 The sella is characteristically enlarged, but an empty sella may be of normal size. Primary empty sella occurs spontaneously and may be associated with arachnoidal cysts or, possibly, infarction of the diaphragma and pituitary. Secondary empty sella follows pituitary surgery or radiotherapy (see Chap. 11 and Chap. 17) and may also be seen in cases of elevated intracranial pressure (e.g., pseudotumor cerebri or hydrocephalus). Neuroradiographic evidence of a reversible empty sella syndrome after therapy for idiopathic intracranial hypertension has been reported.98 Visual field defects, hypopituitarism, headaches, and spinal fluid rhinorrhea occasionally occur. A thorough review of the clinical and radiographic characteristics of primary empty sella99 has revealed the following features: obese women predominate, ranging in age from 27 to 72 years, with a mean age of 49 years; headache is a common symptom; there is no vision impairment because of chiasmal interference; usually, an enlarged sella turcica is found serendipitously on radiologic studies obtained for evaluation of headaches, syncope, or other symptoms; pseudo-tumor cerebri was present in 13% of patients; approximately two-thirds of the patients had normal pituitary function; and the remaining one-third demonstrated endocrine disturbances, including panhypopituitarism and growth hormone, gonadotropin, and thyrotropin deficiency. In another series of patients with primary empty sella,100 the following features are noteworthy: all 19 were female; 12 patients initially reported headache; in 7, vision disturbances were prominent subjective symptoms (blurred vision, diplopia, micropsia); 3 patients had bilateral papilledema, and pseudotumor cerebri was diagnosed; and 2 patients demonstrated minimal, relative hemianopias without obvious cause. Additionally, visual field defects typical of those seen in glaucoma are well documented in patients with empty sella syndrome; the normal intraocular pressures implicate the empty sella syndrome as a potential cause of so-called low-tension glaucoma.101 Secondary empty sella occurs after pituitary surgery or radiotherapy, wherein adhesions form between the tumor “capsule” (or sellar diaphragm) and the nerves and chiasm. Retraction of these adhesions into the empty sella draws the chiasm and nerves downward, with resulting visual defects. Packing the sellar cavity to elevate the diaphragma (chiasmapexy) has been suggested for prophylactic purposes102 or after the fact. Primary empty sella may rarely occur in children in association with multiple congenital anomalies, including the de Morsier syndrome.103

MISCELLANEOUS LESIONS OF THE OPTIC CHIASM TRAUMATIC CHIASMAL SYNDROME Vision loss that follows closed-head trauma usually is attributed to contusion or laceration of the optic nerves occurring abruptly at the time of impact. Much less frequently, a chiasmal syndrome may be identified by the pattern of field loss and associated deficits, including diabetes insipidus, anosmia, CSF rhinorrhea, and fractures of the sphenoid bone. From a report of several such patients,104 it was clear that neither the degree of vision loss nor the extent of diabetes insipidus was necessarily related to the severity of craniocerebral trauma. Transient diabetes insipidus was present in approximately one-half of these patients. Rarely, panhypopituitarism may occur.105 The traumatic chiasmal syndrome may occur more commonly than recognized because of its frequent association with extensive basilar skull fractures and its concomitant altered level of consciousness and high mortality rate.106 Lesions of the hypophysial stalk and, more frequently, of the hypothalamus may follow blunt head trauma. Hypothalamic lesions have been noted in 42% of patients who died after head trauma.107 Ischemic lesions and microhemorrhages were attributed to shearing of small perforating vessels. METASTATIC LESIONS Pituitary metastases are uncommon manifestations of systemic cancer and, initially, may be difficult to distinguish from simple adenomas. To ascertain the incidence of pituitary tumors in cancer patients and to characterize the clinical presentations of metastases, the experience at Memorial Sloan-Kettering Cancer Center was reviewed.108 Also, a series of 500 consecutive autopsies was analyzed, with inclusion of examination of the pituitary gland. In the clinical series, histologic diagnosis was made in 60% of patients. Radiologic evaluation, including polytomography and CT, did not reliably distinguish metastasis from adenoma, but the clinical syndromes were distinctive. In the metastasis group, the review108 revealed an 82% incidence of diabetes insipidus but vision loss in only 11%. In the autopsy series, metastases were found in 36% of cases and adenomas in 1.8%. Two other reported cases109 of sellar metastases showed diplopia resulting from palsies of the third, fourth, and sixth cranial nerves and eventually diabetes insipidus in one patient. In another report, a man with known colon carcinoma developed panhypopituitarism, hyperprolactinemia, chiasmal field loss, and a right third nerve palsy but no diabetes insipidus.110 Several generalizations emerge from these reports of meta-static involvement of the pituitary gland: either the anterior or posterior lobe may be involved; diabetes insipidus is more common than with simple adenoma; cranial nerve palsies are more common than simple adenomas; hyperprolactinemia may be seen, but the serum prolactin level usually is 1 cm, with the superior margin (solid white arrow) protruding into the suprasellar cistern without coming in contact with the optic chiasm (open white arrow). The carotid arteries (solid black arrows) are displaced slightly laterally in the cavernous sinus regions. There does not appear to be clear-cut involvement of the cavernous sinus. C, Sagittal T1-weighted non–contrast-enhanced magnetic resonance image. Again, the small macroadenoma is seen filling the sella and remodeling the sella (solid white arrows). The clivus is slightly remodeled, and the posterior clinoids are not well defined. There is an incidental infarct in the pons (open white arrow).

FIGURE 20-15. Invasive pituitary macroadenoma. This coronal contrast-enhanced 3D gradient echo magnetic resonance image through the sella demonstrates a large irregular aggressive skull-base mass (white arrow) with the pituitary gland enhancing diffusely. The pituitary stalk is displaced to the right. The low-signal-intensity left carotid artery (black arrow) is enveloped by the mass and is displaced inferiorly. The mass protrudes into the left suprasellar cistern. It is in contact with the left medial temporal lobe after breaking through the left cavernous sinus. The mass is extending through the left foramen ovale into the masticator space.

The imaging appearance of pituitary adenomas is nonspecific, and no inference to histology can be made from the sellar patterns. However, additional clues may be present, related to other secondary endocrine changes. For instance, with GH-secreting tumors, acromegaly occurs, and one may visualize thickening of the scalp or enlargement of the mandible on radiograph or physical examination (see Fig. 20-13); these tumors tend to be larger than 5 mm. Cushing adenomas usually are microadenomas, but compression vertebral fractures and a “buffalo hump” deformity may be clues. Prolactinomas are more variable in size; they usually are microadenomas, but may be macroadenomas. Nonfunctioning tumors tend to be large. Enlargement of the pituitary gland may result from many etiologies, not just neoplasia. End-organ failure is a cause of gland enlargement, such as is seen with primary hypothyroidism or surgical removal of the adrenals (Nelson syndrome; see later in this chapter and Chap. 75). If the functional status of a suspected adenoma or pituitary mass is in question, venous sampling of the petrosal sinuses can be performed by means of a catheter placed from the femoral vein into the internal jugular vein and then advanced into the greater petrosal veins.6 Analysis of blood samples can help determine the type of adenoma and the location of a lesion not detected by other imaging modalities. Sampling is also of value to demonstrate that the hormone originated from the gland rather than from an ectopic site. Although inferior petrosal sinus sampling is usually performed, it has been shown that bilateral, simultaneous cavernous sinus sampling, using corticotropin-releasing hormone, is as accurate as inferior petrosal sinus sampling in detecting Cushing disease and is perhaps more accurate in lateralizing the abnormality within the pituitary gland.24 The sella is usually normal in size with microadenomas and CT usually demonstrates no bony expansion, although there may be some asymmetry in the shape of the pituitary gland (see Fig. 20-12). CT, using thin-section coronal images and intravenous contrast, has been used successfully to detect microadenomas. The adenoma is identified as either a hypodense or hyperdense region in the gland after contrast enhancement. Cushing disease adenomas are more difficult to detect by CT, possibly because of their relative enhancement with respect to the normal gland.25 The recommended modality for examining a pituitary adenoma is MRI, with coronal and sagittal imaging. Detection is best with high-resolution techniques, such as three-dimensional imaging. The coronal plane is the most sensitive imaging plane, and T1-weighted spin-echo and three-dimensional imaging sequences are the best pulse sequences. The use of gadolinium enhancement is somewhat controversial,26,27 although the vast majority of radiologists believe that contrast is essential in the evaluation of the sella and parasellar regions. Usually, the tumors enhance less-than-normal tissue. Dynamic imaging can be of value in defining the abnormal segment of the gland.13 Of the pituitary macroadenomas, a higher percentage of these are nonfunctioning adenomas. Plain radiographs of the skull may demonstrate bony expansion or

erosion of the sella; at times, the masses can be huge, with wide destruction of the skull base (to the extent that the site of origin is not clear). Calcifications are rare. The sensitivity for detecting macroadenomas by CT is higher than for microadenomas; the CT examination should use thin-section coronal and axial imaging with intravenous contrast. Generally, the margins of the macroadenomas are more readily defined by MRI than by CT. Involvement of the optic chiasm, cavernous sinus, sphenoid sinus, orbit, temporal lobes, and carotid arteries can all be seen using MRI. In prolactinomas, MRI is used to evaluate the patient's response to bromocriptine therapy. A decrease in tumor size can be seen as early as 1 week after the start of therapy. Additionally, MRI can detect posttherapy hemorrhage into macroadenomas and mass effect or inferior herniation of the chiasm as a result of a decrease in the tumor size.28 In macroadenomas, subacute hemorrhage is readily detected by MRI because the breakdown products of hemoglobin have paramagnetic or diamagnetic effects, depending on their chemical composition. Moreover, MRI is good for evaluating invasion into the adjacent cavernous sinus and for documenting the patency of the carotid arteries (see Fig. 20-15). INFUNDIBULAR MASSES The thickness of the normal pituitary stalk averages 3.5 mm at the median eminence and 2.8 mm near its midpoint. The normal stalk enhances markedly on CT with contrast and on MRI with gadolinium. The most common clinical problem associated with disease of the pituitary stalk is diabetes insipidus. When this is present, there usually is absence of the normal hyperintensity of the posterior pituitary. On T1-weighted MRI, diabetes insipidus may be found to occur as a result of transection of the pituitary stalk. The differential diagnosis of a thickened stalk includes sarcoidosis, tuberculosis, histiocytosis X, and ectopic posterior pituitary as well as germinoma. A thickened stalk can also be due to an extension of a glioma within the hypothalamus. In patients with neurosarcoidosis and tuberculous infiltration of the stalk, the chest radiograph is generally abnormal and may be helpful in the differentiation from histiocytosis X. Clinically, patients with histiocytosis X may have skin lesions, otitis media, or bone lesions in addition to interstitial lung disease.12 HYPOTHALAMIC HAMARTOMAS A hamartoma of the tuber cinereum usually presents as precocious puberty in a young child.29 It is important to differentiate this lesion from a hypothalamic glioma because the prognosis for hamartoma is much more favorable. Imaging is best with MRI thin-section coronal and sagittal planes (Fig. 20-16). The findings are usually characteristic: The mass arises from the undersurface of the hypothalamus and is exophytic. The nodular mass (90% of patients. Surgery and radiation therapy are not indicated for this condition. PROLACTIN HYPERSECRETION ASSOCIATED WITH OTHER LESIONS Non–prolactin-secreting adenomas and other central nervous system lesions may increase prolactin levels by interfering with normal hypothalamic inhibition. Bromocriptine, 2.5 to 7.5 mg per day, usually normalizes serum prolactin levels in these patients. The normalization of prolactin is usually associated with the resolution of galactorrhea. Amenorrhea and infertility may also resolve with this mode of therapy; however, when the underlying lesion has disrupted the normal hypothalamic pituitary axis or has destroyed the gonadotropes, use of bromocriptine does not restore menses or fertility. DETAILS OF BROMOCRIPTINE THERAPY

Bromocriptine (2-Br-a-ergocryptine mesylate) is a semisynthetic ergot alkaloid. It specifically binds to and stimulates dopamine receptors. One-third of an oral dose is absorbed, and peak serum levels are reached 1 to 3 hours after oral administration. It is extensively metabolized by the liver, with the metabolites being excreted almost entirely by biliary secretion.17 Less than 5% of the drug is excreted in the urine. Maximal suppression of prolactin occurs 6 to 8 hours after a single dose, and suppression may be maintained for 12 to 14 hours. Treatment with bromocriptine should be initiated with a dose of one-half of a 2.5-mg tablet taken with food just before bedtime, followed by a regimen of 1.25 mg given with food every 8 to 12 hours. Less than 1% of treated patients experience a first-dose phenomenon, characterized by marked faintness or dizziness. This is observed most commonly in elderly patients and in those with a previous history of fainting, peripheral vascular disease, or use of vasodilators. Increases in dosage should be gradual, no more than 2.5 to 5 mg within a period of a few days to 1 week. The total daily dose is usually divided and administered every 8 to 12 hours. Side effects are usually dose related, with a rapid development of tolerance. Many side effects are potentiated by alcohol, the use of which should be avoided in sensitive patients. To tolerate bromocriptine therapy, some patients may need to begin with a dosage of 0.625 mg per day (one-fourth tablet), thereafter increasing the dosage at 1-week intervals. Nausea is the most common side effect and occurs in up to 25% of treated patients. The nausea is usually mild, may be minimized by administration of the drug with food and by the initial use of low doses, and generally improves with time.18 Constipation is also frequently reported, and some patients experience abdominal cramps. Seven patients receiving high doses of bromocriptine for the treatment of acromegaly were reported to have had major gastrointestinal hemorrhage associated with peptic ulcer disease (three of these episodes were fatal).18 However, bromocriptine has not been associated with an increased incidence of peptic ulcer disease. A slight decline in blood pressure is commonly observed in treated patients; however, patients usually remain asymptomatic. Mild orthostatic hypotension has also been noted.18 The decrease in blood pressure is probably related to both a relaxation of vascular smooth muscle and central inhibition of sympathetic tone. As with the gastrointestinal side effects, symptomatic hypotension usually improves with time. Vascular side effects, including digital vasospasm, livedo reticularis, and erythromelalgia, occur infrequently and are usually associated with bromocriptine doses that exceed those used in the treatment of hyperprolactinemia. Significant mental changes, including hallucinations, have been noted, most commonly in elderly patients receiving large doses of bromocriptine. In two patients, a dose of 5 to 7.5 mg of bromocriptine, administered for treatment of hyperprolactinemia, was reported to have caused psychotic delusions. However, one of these patients had a known history of schizophrenia in remission, and the other was under severe emotional stress. Other side effects of bromocriptine include nasal stuffiness, headache, and fatigue. Women taking bromocriptine should be advised to use mechanical contraception and, if pregnancy is desired or suspected, to discontinue bromocriptine whenever expected menses are >2 days late. Visual fields should be evaluated regularly during pregnancy. If evidence of tumor enlargement is found, a choice is made between continued observation, treatment with bromocriptine, or transsphenoidal surgery, depending on the status of the individual patient. In the United States, women are usually advised to discontinue bromocriptine therapy during pregnancy; in Europe, however, treatment is commonly continued. A review of 1410 pregnancies in 1335 women who received bromocriptine while pregnant revealed that the incidence of spontaneous abortions (11.1%) and congenital anomalies (3.5%) was no higher than that seen in the general population.19 In women not taking other fertility agents, a slightly increased incidence of twin pregnancies (1.8%) was seen. A retrospective study of 64 children born to 53 mothers who took bromocriptine while pregnant revealed no evidence of adverse effects on motor or psychological development.20 THERAPY WITH OTHER DOPAMINERGIC AGONISTS Several other dopamine agonists have been developed that may be useful in the treatment of hyperprolactinemia. A parenteral formulation of long-acting bromocriptine has been effective, with intramuscular injections given every 4 weeks. Pergolide is an ergoline derivative that can be given once daily in a dose of 50 to 100 µg.21 Although it is similar to bromocriptine in its effectiveness and side effects, some patients who do not tolerate bromocriptine may tolerate pergolide.22 The nonergot dopamine agonist quinagolide (CV 205-502) can be administered in dosages of 0.1 to 0.5 mg per day, with fewer side effects than bromocriptine or pergolide. Quinagolide was effective in patients who were unable to tolerate bromocriptine and in some patients who failed to respond adequately to bromocriptine.23 Cabergoline is a long-acting ergoline derivative that can be effective when given weekly or biweekly in doses of 0.5 to 2.0 mg. Its efficacy and side effects profile are similar to or better than those of bromocriptine.24 In several studies, tumor shrinkage and normalization of prolactin levels have occurred in patients who could not tolerate bromocriptine or failed to respond adequately.25,26 and 27

ADRENOCORTICOTROPIC HORMONE HYPERSECRETION When Cushing syndrome is caused by a pituitary tumor (Cushing disease), transsphenoidal surgery is the treatment of choice.3,28 Radiation therapy, by comparison, is less often successful and may take 1 to 2 years to be effective3 (see Chap. 22). Drug treatment is generally not used as a primary mode of therapy except in patients who refuse surgery or irradiation. However, drug treatment may be appropriate in severely ill patients with marked hypokalemia, psychiatric disturbances, infection, or poor wound healing or in patients awaiting transsphenoidal surgery. Medical therapy is also useful in reducing cortisol levels and ameliorating symptoms until pituitary irradiation is fully effective. Finally, drug therapy may be useful in patients in whom surgery and radiation therapy have failed. Patients with Cushing disease who are treated by adrenalectomy may develop large, ACTH-secreting, pituitary macroadenomas (Nelson syndrome). The response of such lesions to both surgery and irradiation has been disappointing. Agents used in the treatment of ACTH hypersecretion can be divided into two classes—those that act centrally to reduce ACTH release and those that act peripherally to reduce cortisol production or block its effect (Table 21-1; see Chap. 1). Centrally acting agents are preferred if a drug is to be used for primary therapy; moreover, they are the only agents appropriate for the treatment of Nelson syndrome. Peripherally acting drugs are the preferred agents for rapid preoperative treatment of severely ill patients awaiting surgery. When the treatment regimen involves the chronic use of peripherally acting drugs, the resultant reduction in cortisol and in negative feedback may be followed by an increase in ACTH hypersecretion, thereby necessitating increased dosages of the drug.

TABLE 21-1. Treatment of Adrenocorticotropic Hormone Hypersecretion

CENTRALLY ACTING DRUGS BROMOCRIPTINE Unlike the excellent results achieved with bromocriptine therapy in patients with hyperprolactinemia, long-term administration of the drug, even at dosages of 20 to 30 mg per day, effectively reduces ACTH hypersecretion in only a few patients.29 Although a single 2.5-mg dose of bromocriptine reduces ACTH levels in ~40% of patients, many of these short-term responders fail to improve significantly with long-term treatment. Conversely, some patients who fail to respond to a single dose of bromocriptine demonstrate marked improvement in symptoms and in ACTH hypersecretion with prolonged therapy.30 Neither the pretreatment ACTH and cortisol

levels nor the tumor size can be used to predict accurately the response to therapy. CYPROHEPTADINE The antiserotoninergic effect of cyproheptadine hydrochloride is thought to be the mechanism whereby ACTH secretion is reduced; however, this drug also has anticholinergic, antihistaminic, and antidopaminergic effects. Thirty percent to 50% of patients with Cushing disease achieve an initial clinical remission with this agent.31 Usually, when the drug is discontinued, elevated cortisol levels and symptomatic disease promptly return. No clinical features can predict which patients will respond to cyproheptadine. Importantly, many authors report poor efficacy and significant side effects with this drug. Occasionally, patients with Nelson syndrome have been reported to improve with administration of cyproheptadine. VALPROIC ACID The anticonvulsant agent valproic acid (and its derivatives) is a g-aminobutyric acid transaminase inhibitor that decreases ACTH hypersecretion in some patients with Cushing disease or Nelson syndrome. Reduction of tumor size with valproate sodium has been reported in a single instance.32 The drug is highly protein bound and has a serum half-life of 6 to 16 hours. Capsules should be swallowed whole and not chewed to avoid local irritation to the mouth and pharynx. Nausea and vomiting are commonly experienced at the time therapy is initiated. Tolerance to these side effects develops rapidly, and symptoms may be reduced by administering the drug with meals. Fatal hepatic failure has occurred in several patients receiving this drug as an anticonvulsant agent. Liver function tests should be performed before the initiation of therapy and at regular intervals during the first year. The drug should not be used in patients with a history of liver disease and should be discontinued if evidence of hepatic dysfunction is found. However, hepatic dysfunction has been known to progress even after discontinuation of the drug. An increased incidence of neural tube defects has been reported in children whose mothers received this agent during the first trimester of pregnancy. PERIPHERALLY ACTING DRUGS METYRAPONE Metyrapone reduces the production of cortisol by inhibiting 11-b-hydroxylation in the adrenal gland. The dosage is titrated to maintain normal serum cortisol levels (which should be evaluated at multiple intervals throughout the day) or titrated to keep the 24-hour urine free cortisol level within the physiologic range. The maintenance dosage varies from 250 mg three times a day to 1000 mg four times a day.30,33 The metabolism of metyrapone is accelerated by administration of phenytoin (Dilantin). The most common side effect is gastrointestinal irritation, which can be avoided by administering the drug with food. Despite improvement in serum cortisol levels, some women note worsening of hirsutism and acne during therapy.33 Cost and side effects may be reduced and efficacy enhanced by combining metyrapone with aminoglutethimide, with 1 g per day of each administered in divided doses. Although the manufacture of metyrapone tablets has been discontinued, capsules remain available from the manufacturer. MITOTANE Mitotane (1,1-dichloro-2-[o-chlorophenyl]-2-[p-chlorophenyl]-ethane or o,p'-DDD) suppresses the function of the zona fasciculata and zona reticularis of the adrenal cortex. The drug has been known to cause necrosis of the adrenal gland, producing acute adrenal insufficiency. Mitotane is inappropriate for rapid treatment because control of cortisol secretion requires 2 to 4 months of therapy.34 It may be useful in the treatment of patients awaiting the full effect of radiation therapy or in those in whom surgery and irradiation have failed.3,30 AMINOGLUTETHIMIDE Aminoglutethimide reduces cortisol production by inhibiting the conversion of cholesterol to D5-pregnenolone. During short-term therapy, serum cortisol levels usually are suppressed to less than one-half of pretreatment values. In some patients, glucocorticoid insufficiency occurs, necessitating concurrent glucocorticoid replacement therapy. When aminoglutethimide is used to treat patients with Cushing disease, a secondary increase in ACTH levels frequently leads to escape from acceptable control.30 Few patients have been treated for >3 months. Therapy is begun with administration of one 250-mg tablet every 6 hours. This dosage is then increased by 250 mg per day every 1 to 2 weeks until a total daily dose of 2 g is reached. Significant side effects occur in two-thirds of patients treated with this agent. The most frequent effects of the drug include drowsiness, which occurs in 33% of patients; skin rashes, which affect 16%; and nausea and vomiting, which occur in 13%. Other significant side effects include vertigo and depression. In general, side effects decrease with smaller doses and often improve or disappear after 1 to 2 weeks of continued therapy. Skin rashes may represent allergic or hypersensitivity reactions; if these are severe or persistent, the drug should be discontinued. Interference with thyroid hormone synthesis may produce hypothyroidism. Decreased estrogen synthesis may produce menstrual irregularities and increased hirsutism and acne in some women. Two cases of pseudohermaphroditism were reported in female infants of mothers who took this drug while pregnant. Because aminoglutethimide increases dexamethasone metabolism, hydrocortisone or cortisone acetate is preferred if glucocorticoid replacement therapy is needed. Inhibition of aldosterone synthesis may produce mineralocorticoid deficiency, presenting with orthostatic or persistent hypotension, which may require therapy with fludrocortisone acetate (Florinef). TRILOSTANE Trilostane is an inhibitor of the 3-b-hydroxysteroid dehydrogenase: D4,D5-isomerase enzyme system. It is generally less effective than the agents described earlier, and results are highly variable.35 Therapy is initiated with 30 mg of trilostane four times a day. This dosage is then increased as required to control serum cortisol and urinary cortisol levels, with an increase every 3 to 4 days until a total dose of 480 mg per day is reached. Significant side effects occur in half of treated patients. Gastrointestinal symptoms are the most common of these, with abdominal pain and discomfort being reported in 16% of patients, diarrhea in 17%, and nausea and vomiting in 5%. Trilostane has been reported to decrease progesterone levels, which has led to cervical dilation and termination of pregnancy in some women. KETOCONAZOLE Ketoconazole is an antimycotic agent that decreases serum cortisol by inhibiting cholesterol synthesis through blockade of the 14-demethylation of lanosterol. Ketoconazole may also inhibit 11-hydroxylation and may decrease the binding of glucocorticoid to its receptor. This drug has been reported to be effective in the treatment of patients with Cushing disease in whom surgery and other drug therapy have proved unsuccessful.3,30,36 After oral administration, the drug is rapidly absorbed. An acid pH is required for absorption; therefore, in patients who are also taking antacids or antihistaminic H2-inhibitors, the drug should be administered 2 hours after such therapy. Patients with achlorhydria may need to dissolve the tablets in aqueous hydrochloric acid. In serum, the drug is 99% protein bound. In patients with Cushing disease, therapy is initiated with 400 mg of ketoconazole administered every 12 hours for 1 month; this dosage is then decreased to 400 to 600 mg per day. Urinary cortisol levels were reported to decline significantly within 1 day after onset of therapy. In patients receiving conventional antifungal doses (200–400 mg per day), the most common side effects are nausea and vomiting, occurring in 3%, and abdominal pain, occurring in 1.5%. Hepatotoxicity has been reported to occur in 1 in 10,000 treated patients; this condition usually resolves on discontinuation of the drug. However, one fatal case of hepatic necrosis that progressed despite discontinuation of the drug was reported. GLUCOCORTICOID RECEPTOR ANTAGONIST Mifepristone (RU 486) is a synthetic steroid agonist antagonist that blocks the binding of glucocorticoids to their receptor. It is under investigation as a potential therapeutic agent in the treatment of Cushing disease.37

GROWTH HORMONE HYPERSECRETION Transsphenoidal surgery remains the treatment of choice for growth hormone–secreting adenomas (see Chap. 23). The overall rate of cure (defined as serum growth hormone levels of 7 mm); lateral deviation of the pituitary stalk; and a focal area of altered attenuation relative to the normal gland, on either contrast or noncontrast studies. Usually, precise delineation between tumor and other important structures in the area can be accomplished (Fig. 23-1).

FIGURE 23-1. Magnetic resonance image without contrast (A) and with contrast (B) with direct coronal scans for a young woman with a pituitary macroprolactinoma. Note the low-density areas in the lesion on both scans. The surrounding tissue enhances after the administration of intravenous contrast, correlating well with the surgical finding of normal glandular tissue, rather than tumor, surrounding the low-density center. The tumor was precisely confined to the low-density area.

High-resolution MRI scanning with and without gadolinium enhancement is recommended for assessment of all suspected pituitary and hypothalamic lesions. Carotid angiography is reserved for those patients in whom an intrasellar aneurysm is suspected after high-quality MRI scans, including MRI angiography, have been performed. The author has not found it necessary to perform carotid angiography on the last 200 patients treated, now that high-resolution MRI angiography is available. CT scanning is no longer routinely performed in these patients. It can occasionally be helpful for patients in whom areas of hemorrhage, bony erosion, or calcification are being assessed or for patients with unusual bony sphenoid sinus anatomy, particularly those who have undergone previous transsphenoidal exploration. SURGICAL APPROACHES Surgical approaches to pituitary adenomas have been described in detail.5,6 The specific morphologic configuration of the neoplasm, rather than the endocrinologic syndrome, determines the choice between the transcranial and the transsphenoidal approach. The transsphenoidal approach is the technique of choice for tumors that occupy the sella, whether or not any extension has occurred into the sphenoid sinus (Fig. 23-2, Fig. 23-3 and Fig. 23-4). Tumors with vertical suprasellar extension without significant lateral extension are also well treated with this approach. The advantage of the transsphenoidal approach is that it usually allows selective excision of tumor with preservation of remaining normal pituitary gland, even when most of the sella is occupied by tumor. The approach involves no retraction of the cortex whatsoever, as opposed to the transcranial approach, in which, at times, considerable brain retraction may be necessary. In addition, the morbidity of the procedure is exceedingly low, and it is well tolerated even by patients who would be considered unacceptable surgical candidates for craniotomy. In experienced hands, only 1% of patients with pituitary tumors require a transcranial operation.

FIGURE 23-2. Bony and cartilaginous anatomy of the base of the skull, sphenoid sinus, and nasal areas. Note that the posterior wall of the sphenoid sinus is the floor of the sella turcica, making the transsphenoidal route uniquely suited for the removal of sellar lesions. (From Tindall GT, Barrow DL. Disorders of the pituitary. St. Louis: CV Mosby, 1986.)

FIGURE 23-3. Diagrammatic summary of the transsphenoidal surgical approach. A, A linear incision is made from canine fossa to canine fossa. The entire surgical field lies within this incision. This provides a cosmetically favorable result because the scar is never visible externally. The nasal mucosa is dissected away from the cartilaginous and bony nasal septum. B, A speculum is then placed to expose the sphenoid sinus, and the posterior wall of the sphenoid sinus (the floor of the sella) is removed. Note the adenoma in the anterior aspect of the gland, where most of these lesions are located. This procedure is performed with the aid of an operating microscope, using a C-arm fluoroscope, which facilitates visualization of the area. C, Removal of the microadenoma. Using magnification and microdissection technique, the adenoma can be removed from the gland, sparing the normal gland tissue. D, Reconstitution of the sella after removal of the tumor or the gland. Fat is placed in the sella to prevent downward migration and herniation of the optic chiasm. A piece of nasal bone is then used to reconstitute the sellar floor, which later calcifies and forms new bone. (A, C, and D from Hardy J. Transsphenoidal operations on the pituitary. Codman and Shurtless, Inc. A division of Johnson and Johnson. 1983; B from Tindall GT, Barrow DL. Disorders of the pituitary. St. Louis, CV Mosby, 1986.)

FIGURE 23-4. Technical details relating to removal of a microadenoma. A, Basic principles of tumor removal. Note the development of a plane between the tumor located laterally and the normal gland located medially. B, Dissection of the pseudocapsule, or the fibrous tissue surrounding the outer aspects of the tumor, which ensures a clean removal. C, Two important principles of microsurgical removal are illustrated. The first is to carefully inspect, or at least palpate, all hidden pockets. One can see tumor hidden in the anterior corner of the sella, which is extracted by the inserted curette. In addition, extracting the surrounding tissue for biopsy to

confirm that it contains only normal gland and, therefore, that all tumor has been removed is usually advisable. (From Hardy J. Transsphenoidal approach to the sella. In: Wilson CB, ed. Neurosurgical procedures: personal approaches to classical operations. Philadelphia: Williams & Wilkins, 1992:30.)

An advance in surgical technology is the introduction of endoscopy (Fig. 23-5) into neurosurgical procedures.6a In selected cases, a transsphenoidal resection can now be performed via one nostril using the endoscope, so that the degree of invasiveness of the operation is even further reduced.7 Moreover, the endoscope now permits the surgeon to “look around the corner” at angles that were not possible using conventional microsurgical techniques, thereby improving surgical outcome. Patients do not have nasal packing placed, which thus avoids both the numbness in the upper teeth that persists for at least several months and the nasal congestion and stuffiness that often occurs for several weeks after a standard transsphenoidal operation. In many cases, the patient can be discharged as early as the first postoperative day. A few patients have been given surgery on an outpatient basis with excellent outcomes.

FIGURE 23-5. The use of the endoscope for transsphenoidal surgery. A, Diagram demonstrating an endoscopic endonasal approach to a sellar tumor. No septal, alar, or gingival incision is used, and no speculum or retractor is necessary. B, The endoscope is held in the surgeon's hand until an opening is made into the sphenoid sinus. C, The endoscope is mounted on a special holder, which provides the surgeon with a steady video image and frees both hands to use surgical instruments simultaneously. (From Jho HD, Carrau RL. Endoscopic endonasal transsphenoidal surgery: experience with 50 patients. J Neurosurg 1997; 87:44.)

Another advance is the refinement of intraoperative technology to permit the use of ultrasonography in the operating room to localize small tumors that might be otherwise difficult to visualize.8 This is particularly useful for patients with Cushing disease for whom imaging studies have been normal or equivocal. The ultrasonic probe developed for this procedure is pencil thin and can therefore be used in the very small area that constitutes the surgical field.

PROLACTIN-SECRETING ADENOMAS Prolactin-secreting adenomas comprise the largest group of pituitary tumors. The behavior and relatively benign clinical manifestations of small prolactinomas distinguish them from the tumors that produce Cushing disease and acromegaly, two distinct endocrinopathies that are usually life-threatening. Whereas the clinical necessity of treating patients with either Cushing disease or acromegaly is clear, the indications for immediate treatment of patients harboring a small prolactin-secreting adenoma are less so (see Chap. 13 and Chap. 21). CLINICAL MANIFESTATIONS In 1954, Forbes and colleagues9 first reported that pituitary adenomas could produce amenorrhea and galactorrhea. However, only recently have these tumors been recognized as a frequent cause of secondary amenorrhea and galactorrhea. Among the women in one surgical series5 (Table 23-2), 80% presented with secondary amenorrhea or galactorrhea, 10% with primary amenorrhea, and 10% with either oligomenorrhea and galactorrhea, secondary amenorrhea without galactorrhea, or secondary amenorrhea only. Among men, prolactinomas usually remain undetected until a large tumor produces either significant panhypopituitarism or compression and invasion of the parasellar structures. In the previously mentioned series,5 only seven men had symptomatic hyperprolactinemia without abnormalities of additional pituitary hormones; either thyroid, adrenal, or gonadotropic function, or a combination of the three, was usually impaired as well. Many of these patients experience impotence early in the course of their disease, but this problem often does not lead to an investigation of the prolactin level. The author can recall seeing in his practice a 35-year-old man who received electroshock therapy for 10 years as “treatment” for his impotence; the patient presented with a prolactin level of >1000 ng/mL.

TABLE 23-2. Clinical Features of 121 Patients with Prolactinoma Treated by Transsphenoidal Surgery*

Hyperprolactinemia secondary to a pituitary adenoma has extragonadal manifestations. Recent rapid weight gain is a frequent complaint of hyperprolactinemic women and occurs with a frequency that suggests a correlation. Correction of hyperprolactinemia, either by surgery or by medical therapy, has been followed by impressive weight loss in many cases, despite no apparent change in dietary habits. Equally impressive is the incidence of emotional lability, which is often dramatically reversed after the correction of hyperprolactinemia. Studies demonstrate that the estrogen deficiency secondary to hyperprolactinemia causes bone demineralization, sometimes producing secondary complications.10 LABORATORY EVALUATION The first step in the evaluation of a patient with suspected hyperprolactinemia is to obtain a fasting serum prolactin level. The administration of thyrotropin-releasing hormone (TRH) does not consistently distinguish between functional hyperpro-lactinemia and actual prolactinoma11 (see Chap. 13). In men whose basal prolactin values exceed 100 ng/mL, establishing a prolactinoma as the cause of the hyperprolactinemia is not difficult. In women, hyperprolactinemia (>200 ng/mL) almost invariably indicates a tumor. Caution must be exercised, because prolactin levels as high as 662 ng/mL have been observed to occur in nonsecreting tumors, presumably due to pronounced pressure on the pituitary stalk, which inhibits the transport of prolactin inhibitory factor to the pituitary gland.12 In the author's experience, the diagnosis of prolactinoma in a patient with basal prolactin levels 600 ng/mL. (For such lesions, the cure rate with surgery, even in the most experienced hands, is only 10%.10) Medical therapy is usually effective for long-term control, with normalization of serum prolactin levels. On the other hand, the presence of vision loss complicates the management of such patients. This is because of the concern that such therapy either may fail or may take too long to produce sufficient reduction of tumor volume to relieve the compression of the vision system, which could result in further irreversible vision damage during the trial of medical treatment. Because vision compromise can reverse after surgical treatment, even when the compression is longstanding, some believe that vision compromise is not a contraindication to a trial of medical therapy. Substantial tumor shrinkage can occur within days, leading to improved vision.13,14 Others,5,6 including the author, believe that surgical intervention is indicated in these patients if they are otherwise healthy, because a risk exists of further permanent vision damage with the less rapid decompression provided by medical therapy. If medical therapy is selected for patients with vision compromise, careful monitoring of vision is essential.13,14 PROLACTIN-PRODUCING MACROADENOMAS For macroadenomas, operative removal is recommended if visual compromise is present and if the patient's overall medical condition justifies the small risks of surgical intervention. For macroadenomas without compression of the optic apparatus, surgery may be considered if the tumor is 600 ng/mL, medical therapy is recommended initially. A desire for pregnancy complicates matters, because pregnant patients with macroadenomas may develop complications related to accelerated tumor growth. Because of this concern, such patients may be candidates for surgery, even if no visual compromise is present. PROLACTIN-PRODUCING MICROADENOMAS For microadenomas, opinions differ among surgeons concerning initial treatment. Some believe that all patients should be treated medically, except for those who develop unacceptable side effects to medical therapy or whose tumors are resistant to dopamine agonists.15 Others believe that surgery should be the initial treatment for healthy patients with microadenomas and that bromocriptine, cabergoline, and irradiation should be reserved for cases of surgical failure or for those in whom the risk of surgery is high.5,6 Surgery does not always cure prolactin-producing microadenomas; in particular, tumors with higher prolactin levels have a greater likelihood of surgical treatment failure. Serious surgical complications can occur, although, in experienced hands, the complication rate is 300 mOsm/kg and >145 mEq/L (145 mmol/L), respectively. Although rare, they constitute a major management challenge. ETIOLOGY Transient hyperosmolality may occur after the ingestion of large amounts of salt,34 but most hypernatremic states occur after inadequate water intake. This can occur in any healthy individual in whom the combination of excess fluid loss—from skin, gastrointestinal tract, lungs, or kidneys—and inadequate access to water is found. This occurs most commonly in acute illness in which water intake is compromised by vomiting or impaired consciousness and most vividly in patients with diabetes insipidus, before treatment, or when access to water is denied. In other cases, however, hypernatremia reflects a primary disorder of thirst deficiency (hypodipsia). A number of conditions are associated with hypodipsia (Table 26-2). One of the more common causes of hypodipsic hypernatremia that the authors have seen is ligation of the anterior communicating artery, after subarachnoid hemorrhage from a berry aneurysm. Other centers have reported that neoplasms account for 50% of such cases.35 Craniopharyngiomas are particularly associated with hypodipsic diabetes insipidus, sometimes in conjunction with other hypothalamus-related disorders, such as polyphagia, weight gain, and abnormal thermoregulation. Survivors of diabetic hyperosmolar coma have been shown to have impaired osmoregulated thirst,35 which suggests that hypodipsia contributes to the development of the hypernatremia, which is characteristic of the condition. In almost every case of hypodipsia, associated abnormalities of vasopressin secretion are seen, a finding that reflects the close anatomic proximity of the osmoreceptors for vasopressin secretion and thirst.

TABLE 26-2. Specific Causes of Hypodipsic Hypernatremia

CLINICAL FEATURES In young children and the elderly, hypernatremia may be associated with significant degrees of dehydration.36 Infants are at particular risk, and the mortality is high. In this clinical situation, signs are seen of extracellular fluid loss, decreased skin turgor and elasticity, dry and shrunken tongue, tachycardia, and orthostatic hypotension. Affected infants have depressed fontanelles and tachypnea, and their respirations are deep and rapid. Fever is often present, and the temperature may be as high as 40.5°C (105°F). Adults with mild hypernatremia may have no symptoms, but as plasma sodium levels rise above 160 mEq/L, neurologic signs become apparent.36,37 Early symptoms include lethargy, nausea, and tremor, which progress to irritability, drowsiness, and confusion. Later features of muscular rigidity, opisthotonus, seizures, and coma reflect generalized cerebral and neuromuscular dysfunction. The most severe neurologic disturbances are seen at both ends of the age spectrum. The severity of such disturbances is also related to the rate at which hypernatremia develops, as well as to the absolute degree of hyperosmolality. Intracerebral vascular lesions are often the cause of death. In contrast to patients with the life-threatening clinical features of hypernatremic dehydration, patients with long-standing, moderate hypernatremia (plasma sodium concentrations of 145 to 160 mEq/L) may have few manifestations of the disorder other than lack of thirst. Hypodipsia is the crucial symptom, but it is often overlooked in the clinical setting because patients fail to complain of lack of thirst. However, careful evaluation of these patients reveals that some have no desire to drink any fluid under any circumstances, which suggests a total loss of the thirst osmoreceptor function. Others have only minimal thirst with marked hypertonicity, whereas a third group eventually experiences a normal thirst sensation, but only at high plasma osmolality levels. The key to recognizing subtle differences in thirst appreciation rests with a satisfactory measure of thirst. Visual analog scales for measuring thirst during dynamic tests of osmoregulation38,39 have been shown to produce highly reproducible results.40 When these scales are used in evaluating healthy persons, a linear increase is noted in the degree of thirst and fluid intake with increase in plasma osmolality, and an osmolar threshold for thirst is seen that is a few milliosmoles per kilogram higher than the osmolar threshold for vasopressin secretion.39 The application of these techniques to patients with chronic hypernatremia has disclosed numerous disorders of osmoregulation. OSMOREGULATORY DEFECTS IN CHRONIC HYPERNATREMIA Chronic hypernatremia is characterized by inappropriate lack of thirst despite increased plasma osmolality and mild hypovolemia. Plasma sodium concentrations are typically elevated (150–160 mmol/L) and may reach extremely high concentrations during intercurrent illnesses (e.g., gastroenteritis) in which body water deficits increase. Although adipsic hyper-natremia is uncommon, four distinct patterns of abnormal osmoregulatory function have been described. Type 1 Adipsia. The characteristic abnormalities in type 1 adipsia are subnormal vasopressin levels and thirst responses to osmotic stimulation (Fig. 26-4). The sensitivity of the osmoreceptors is decreased, producing partial diabetes insipidus and relative hypodipsia. Because some capacity remains to secrete vasopressin and experience thirst, such patients are protected from extremes of hypernatremia, as they can produce near-maximal antidiuresis as plasma osmolality increases. Patients with this type of adipsia usually have normal vasopressin responses to hypotension and hypoglycemia, and show suppression of vasopressin secretion, with the development of hypotonic diuresis in response to water loading.

FIGURE 26-4. Thirst and vasopressin responses to osmotic stimulation in adipsic hypernatremia. Type 1: subnormal response of both thirst and vasopressin secretion. Type 2: total lack of response of thirst and vasopressin secretion. Type 3: reset of osmostat for vasopressin release and thirst to the right of normal. Shaded areas

indicate the response ranges in healthy control subjects; the dotted lines are the mean regression lines. (pAVP, plasma arginine vasopressin; LD, limit of detection of the pAVP assay [0.3 pmol/L]).

Type 2 Adipsia. Total ablation of the osmoreceptors produces complete diabetes insipidus and absence of thirst in response to hyperosmolality. This is the pattern of osmoregulatory abnormality seen after surgical clipping of aneurysms of the anterior communicating artery,41,42 and despite the complete absence of osmoregulated thirst and vasopressin release, thirst and vasopressin responses to hypotension and apomor-phine are preserved.41,43 Some patients also develop this type of osmoregulatory dysfunction after surgery for large, suprasellar craniopharyngiomas. Interestingly, these patients also have absent baroregulated thirst and vasopressin secretion—presumably because the extent of surgical injury is such that both the osmoreceptors and the paraventricular and supraoptic nuclei are damaged. Patients with complete adipsic diabetes insipidus have no defense against dehydration, and unless they are closely supervised and trained to drink even in the absence of thirst, they can develop profound hypernatremic dehydration, even in the absence of intercurrent illness. Interest has been shown in the concept that osmoreceptor activity is under bimodal control; that is, a specific stimulus is required to switch off vasopressin secretion in the same way that elevation of plasma osmolality stimulates vasopressin secretion. Patients with complete osmoreceptor ablation clearly are unable to respond to inhibitory inputs; this has been demonstrated in clinical studies in which complete suppression of the secretion of the small quantities of radioimmunoassayable vasopressin or the achievement of maximal free water clearance during water loading was impossible in a patient with this type of osmoregulatory dysfunction.44 Therefore, in some patients vasopressin secretion may not be entirely suppressed during fluid loads, resulting in significant hyponatremia. Type 3 Adipsia. The osmostats for thirst and vasopressin release may be reset to the right of normal (type 3 in Fig. 26-4), such that vasopressin secretion and thirst do not occur until higher plasma osmolalities are reached. Thereafter, the slope of the osmoregulatory lines are normal. This pattern is found in conjunction with a number of cases of “essential” hypernatremia, although type 1 defects have also been reported.45,46 and 47 Patients also have intact nonosmotic release of vasopressin and increased renal sensitivity to vasopressin, so that renal concentrating ability may be reasonably well maintained. Miscellaneous Causes of Adipsia. Osmoregulatory dysfunction has also been reported in elderly patients, who have diminished thirst in response to hypernatremia.38 Although the defect in thirst appreciation is similar to that in type 1 dysfunction, vasopressin responses have variously been reported as being subnormal, normal, or enhanced. Survivors of diabetic hyper-osmolar, nonketotic coma have also been reported to have hypodipsia with exaggerated vasopressin secretion.35 In addition, a single case has been reported of a young patient who had hypodipsia but a normal osmotically regulated vasopressin release.48 All of these reports lend support to the hypothesis that the osmoreceptors subserving vasopressin release are anatomically and functionally distinct from those controlling thirst. TREATMENT Water replacement is the basic therapy for patients with hyper-osmolar states associated with dehydration. The oral route is preferred, but if the clinical situation warrants urgent treatment, the infusion of hypotonic solutions may be necessary. However, overzealous rehydration with hypotonic fluids may result in seizures, neurologic deterioration, coma, and even death secondary to cerebral edema.34,37 Therefore, the decision to treat with hypotonic intravenous fluids should not be made lightly, and rehydration to a euosmolar state should proceed cautiously over at least 72 hours. As plasma osmolality falls, polyuria indicative of hypothalamic diabetes insipidus may develop; this responds to administration of desmopressin. For patients with chronic hyperosmolar syndromes (see Fig. 26-4), longer-term therapy must be considered. Patients with type 3 defects rarely need specific therapy because their osmoregulatory system is essentially intact but operates around a higher than normal plasma osmolality. Patients with type 1 defects (involving partial destruction of the osmoreceptor) should be treated with a regimen of increased water intake (2–4 L every 24 hours). If this leads to persistent polyuria, a small dose of desmopressin can be administered, but plasma osmolality or sodium levels must then be monitored regularly. Considerable difficulties arise in treating patients who have complete destruction of their osmoreceptors (type 2 defect), because these patients cannot protect themselves from extremes of dehydration and overhydration. Most patients need between 2 and 4 L of fluid per day, but the precise amount varies according to seasonal climatic changes, and the body weight must be monitored daily to provide an index of fluid balance.44 Regular (usually weekly) measurements of plasma osmolality or sodium are needed to ensure that no significant fluctuations occur in body water, and constant supervision is required to make certain the requisite volume of water is consumed. Despite the most vigorous supervision, such patients are extremely vulnerable to swings in plasma osmolality and are particularly prone to severe hypernatremic dehydration. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

De Wardener HE, Herxheimer A. The effect of high water intake on the kidneys' ability to concentrate urine in man. J Physiol (Lond) 1957; 139:42. Robertson GL. Diagnosis of diabetes insipidus. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:176. Maffly RH. Diabetes insipidus. In: Andreoli TE, Grantham JJ, Rector FCJ, eds. Disturbances in body fluid osmolality. Bethesda, MD: American Physiology Society, 1977:285. Robertson GL. Diabetes insipidus. Endocrinol Metab Clin North Am 1995; 24:549. Ito M, Mori Y, Oiso Y, Saito H. A single base substitution in the coding region for neurophysin II associated with familial central diabetes insipidus. J Clin Invest 1991; 87:725. Krishnamani MRS, Philips PA III, Copeland KC. Detection of a novel arginine vasopressin defect by dideoxy fingerprinting. J Clin Endocrinol Metab 1993; 77:596. McLeod JF, Kovacs L, Gaskill MB, et al. Familial neurohypophyseal diabetes insipidus associated with a signal peptide mutation. J Clin Endocrinol Metab 1993; 77:599A. Heppner C, Kotzka J, Bullmann C, et al. Identification of mutations of the arginine vasopressin-neurophsia II gene in two kindreds with familial central diabetes insipidus. J Clin Endocrinol Metab 1998; 83:693. Rotig A, Cormier V, Chatelain P. Deletion of the mitochondrial DNA in a case of early-onset diabetes mellitus, optic atrophy and deafness, Wolfram syndrome (MIM 222300). J Clin Invest 1993; 91:1095. Barrett TG, Bundey SE. Wolfram (DIDMOAD) syndrome. J Med Genet 1997; 34:838. Verbalis JG, Robinson AG, Moses AM. Postoperative and post-traumatic diabetes insipidus. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:247. Moses AM. Clinical and laboratory observations in the adult with diabetes insipidus and related syndromes. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers of hormone research, vol 13. Basel: S Karger, 1985:156. Baylis PH, Cheetham T. Diabetes insipidus. Arch Dis Child 1998; 79:84. Scherbaum WA, Bottazzo GF. Autoantibodies to vasopressin cells in idiopathic diabetes insipidus: evidence for an autoimmune variant. Lancet 1983; 1:897. Imura H, Nakao K, Shimatsu A, et al. Lymphocytic infundibuloneurohypo-physitis as a cause of central diabetes insipidus. N Engl J Med 1993; 329:683.

15a.Knoers NV, Monnens LL. Nephrogenic diabetes insipidus. Semin Nephrol 1999; 19:344. 15b.Bendz H, Aurell M. Drug-induced diabetes insipidus. Drug Saf 1999; 21:449. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42.

Bichet DG, Birnbaumer M, Louergan M, et al. Nature and recurrence of AVPR2 mutations in X-linked nephrogenic diabetes insipidus. Am J Hum Genet 1994; 55:278. Hochberg Z, van Lieburg A, Even L, et al. Autosomal recessive nephrogenic diabetes insipidus caused by an aquaporin-2 mutation. J Clin Endocrinol Metab 1997; 82:686. Sato N, Ishizaka H, Yagi H, et al. Posterior lobe of the pituitary in diabetes insipidus: dynamic MR imaging. Radiology 1993; 186:357. Dashe AM, Cramm RE, Crist CA, et al. A water deprivation test for the differential diagnosis of polyuria. JAMA 1963; 185:699. Miller MT, Dalakos T, Moses AM, et al. Recognition of partial defects in antidiuretic hormone secretion. Ann Intern Med 1970; 73:721. Baylis PH. Diabetes insipidus. Medicine 1997; 25:9. Robertson GL. The regulation of vasopressin function in health and disease. Recent Prog Horm Res 1977; 33:333. Moses A, Streeten D. Differentiation of polyuric states by measurement of responses to changes in plasma osmolality induced by hypertonic saline infusions. Am J Med 1967; 42:368. Baylis PH, Robertson GL. Vasopressin response to hypertonic saline infusion to assess posterior pituitary function. J R Soc Med 1980; 73:255. Zerbe RL, Robertson GL. A comparison of plasma vasopressin measurement with a standard indirect test in the differential diagnosis of polyuria. N Engl J Med 1981; 305:1539. Thompson CJ, Baylis PH. Thirst in diabetes insipidus: clinical relevance of quantitative assessment. Q J Med 1987; 65:853. Davison JM, Gilmore EA, Dürr J, et al. Altered osmotic thresholds for vasopressin secretion and thirst in human pregnancy. Am J Physiol 1984; 246:F105. Amico J. Diabetes insipidus in pregnancy. In: Czernichow P, Robinson AG, eds. Diabetes insipidus in man. Frontiers in hormone research, vol 13. Basel: S Karger, 1985:266. Hime MC, Richardson JA. Diabetes insipidus and pregnancy: case report, incidence and review of the literature. Obstet Gynecol Surv 1978; 33:375. Barron WM, Cohen LH, Ulland LA, et al. Transient vasopressin-resistant diabetes insipidus of pregnancy. N Engl J Med 1984; 310:442. Hughes JM, Barron WM, Vance ML. Recurrent diabetes insipidus associated with pregnancy: pathophysiology and therapy. Obstet Gynecol 1989; 73:462. Cobb WE, Spare S, Reichlin S. Diabetes insipidus: management with DDAVP (1-desamino-8- D-arginine vasopressin). Ann Intern Med 1978; 88:183. Williams TDM, Dungar DB, Lyon CC, et al. Antidiuretic effect and pharma-cokinetics of oral 1-desamino-8- D-arginine vasopressin. 1. Studies in adults and children. J Clin Endocrinol Metab 1986; 63:129. Ross EJ, Christie SBM. Hypernatremia. Medicine (Baltimore) 1969; 48:441. McKenna K, Morris AM, Azam H, et al. Subnormal osmotically stimulated thirst and exaggerated vasopressin release in human survivors of hyperosmolar coma. Diabetologia May 1999; 42:538. Robertson GL, Aycinena P, Zerbe RL. Neurogenic disorders of osmoregulation. Am J Med 1982; 72:339. Arieff AL, Guisado R. Effects on the central nervous system of hypernatremic and hyponatremic states. Kidney Int 1976; 10:104. Phillips PA, Rolls BJ, Ledingham JGG, et al. Reduced thirst after water deprivation in healthy elderly men. N Engl J Med 1984; 311:753. Thompson CJ, Thompson J, Burd J, Baylis PH. The osmotic threshold for thirst and vasopressin release are similar in healthy men. Clin Sci 1986; 71:651. Thompson CJ, Selby P, Baylis PH. Reproducibility of osmotic and nonosmotic tests of vasopressin secretion in men. Am J Physiol 1991; 260:R533. Pearce SHS, Argent NB, Baylis PH. Chronic hypernatremia due to impaired osmoregulated thirst and vasopressin secretion. Acta Endocrinol (Copenh) 1991; 125:234. McIver B, Connacher A, Whittle A, et al. Adipsic diabetes insipidus after clipping of anterior communicating artery aneurysm. BMJ 1991; 303:1465.

43. 44. 45. 46. 47. 48.

Teelucksingh S, Steer CR, Thompson CJ, et al. Hypothalamic syndrome and central sleep apnea associated with toluene exposure. Q J Med 1991; 286:185. Ball SG, Vaidja B, Baylis PH. Hypothalamic adipsic syndrome: diagnosis and management. Clin Endocrinol 1997; 47:405. De Rubertis FR, Michelis MF, Beck N, et al. “Essential” hypernatremia due to ineffective osmotic and intact volume regulation of vasopressin secretion. J Clin Invest 1971; 50:97. Dunger DB, Seckl JR, Lightman SL. Increased renal sensitivity to vasopressin in two patients with essential hypernatremia. J Clin Endocrinol Metab 1987; 64:185. Gill G, Baylis PH, Burn J. A case of “essential” hypernatremia due to resetting of the osmostat. Clin Endocrinol (Oxf) 1985; 22:545. Hammond DN, Moll GW, Robertson GL, Chelmicks-Schorr E. Hypodipsic hypernatremia with normal osmoregulation of vasopressin. N Engl J Med 1986; 315:433.

CHAPTER 27 INAPPROPRIATE ANTIDIURESIS AND OTHER HYPOOSMOLAR STATES Principles and Practice of Endocrinology and Metabolism

CHAPTER 27 INAPPROPRIATE ANTIDIURESIS AND OTHER HYPOOSMOLAR STATES JOSEPH G. VERBALIS Frequency and Significance of Hypoosmolality Definition of Hypoosmolality Situations in Which Hyponatremia does not Reflect True Hypoosmolality Influence of Unmeasured Solutes Pathogenesis of Hypoosmolality Solute Depletion Water Retention Cellular Inactivation of Solute Differential Diagnosis of Hypoosmolality Decreased Extracellular Fluid Volume (Hypovolemia) Increased Extracellular Fluid Volume (Edema, Ascites) Normal Extracellular Fluid Volume (Euvolemia) Syndrome of Inappropriate Antidiuresis Diagnostic Criteria Etiology Pathophysiology Clinical Manifestations of Hypoosmolality Therapeutic Approach to Hypoosmolality Acute Treatment of Hypoosmolality Long-Term Treatment of Hypoosmolality Chapter References

FREQUENCY AND SIGNIFICANCE OF HYPOOSMOLALITY Hypoosmolality of plasma is relatively common in hospitalized patients. The incidence and prevalence of hypoosmolar disorders depend on the nature of the patient population being studied and on the laboratory methods and diagnostic criteria used. Most investigators have used the serum sodium concentration ([Na+]) to determine the clinical incidence of hypoosmolality. When hyponatremia is defined as a serum [Na+] of 200 cases have been described. Of 154 cases reviewed, euthyroid goiter accounted for 39%, hypothyroidism without goiter for 17%, and hypothyroidism with goiter for the remaining cases.11 Such goiters, often associated with hypothyroidism, have been reported sporadically in the United States for many years and more commonly in Europe. In the past, the problem was seen most often after the use of

iodide-containing expectorants. Amiodarone use has been associated with hypothyroidism and, in several cases, goiter has developed with benziodarone use. The high incidence of antithyroidal antibodies in patients with goiter development suggests that these patients have underlying Hashimoto thyroiditis with abnormal hormone synthesis and a unique sensitivity to the effects of iodine.29 It may be that iodide-induced goiter or hypothyroidism occurs only in the presence of thyroid dysfunction. The true incidence of goiter in patients exposed to iodides is unknown. The incidence is higher in women than in men, and most patients have received large amounts of iodide (several milligrams to >1 g) daily for long periods.3 Goiter also has been seen after therapy with small amounts of iodide, however.12 The interval from the institution of therapy with iodine-containing compounds to the appearance of goiter or hypothyroidism varies widely, from a few months to several years. EFFECTS DURING PREGNANCY Iodides cross the placenta easily; therefore, one should expect that goiter may be seen in neonates whose mothers received iodide-containing drugs during pregnancy. Congenital goiter and neonatal hypothyroidism resulting from long-term maternal iodide use are rare and have similar manifestations as in adults, but the consequences for infants are severe and include asphyxiation from tracheal compression and potential mental retardation if the condition is not treated. Death from asphyxiation was reported in 8 of 22 infants born with goiter.11 Surgical intervention may be warranted in selected cases, although the goiters resolve over a few months in asymptomatic infants. Generally, the mothers have had no manifestations of goiter or hypothyroidism. NEONATAL PERIOD The frequency or magnitude of the problem of neonatal iodide-induced goiter or hypothyroidism is unknown. Iodides are secreted into breast milk, and hypothyroidism or goiter may be seen in breast-fed infants of mothers chronically exposed to high doses of iodides.11,30 Prenatal vitamins usually contain 150 µg of iodine, which should not constitute a hazard but does pose an additional risk in pregnant women who already are ingesting iodides from other sources. In an intensive care unit setting, within 1 month after the extensive topical use of 1% iodine and 3.9% iodine solutions (Polyvidone) for 4 to 8 days, goiter and hypothyroidism developed in 5 of 30 neonates.31 The affected infants had significantly increased ioduria, whereas the other infants did not, which suggests a difference in skin permeability. In all cases, the goiter resolved after brief therapy with thyroid hormone, and neither hypothyroidism nor goiter recurred subsequently. These observations suggest inadequate or undeveloped ability of the neonatal thyroid gland to “autoregulate” or control iodide uptake and binding, resulting in saturation and suppression of hormone synthesis and secretion. “COAST GOITER” “Coast goiter,” the endemic iodide goiter seen in the northern Japanese island of Hokkaido, is the most common form of iodide goiter.12,32 It occurs in 6% to 12% of that population, most commonly among the seaweed harvesters and their families, who may consume 50 to 200 mg of iodide daily. In a survey of schoolchildren, 508 goiters were found, 330 in girls and 178 in boys. Nearly all the glands (97.4%) were diffusely enlarged, and a histologic diagnosis of colloid goiter was rendered in the seven glands that underwent biopsy. Despite having extremely large goiters, these patients all have been euthyroid. When seaweed was withdrawn from the diet, the elevated plasma and urinary iodide levels returned toward normal, resulting in a rebound increase of 131I uptake 1 to 2 weeks later.32 In the few patients studied, the goiters disappeared or significantly decreased in size after seaweed withdrawal (Fig. 37-1). In 75% of 50 patients treated with thyroid hormone who maintained their usual dietary habits, the goiters decreased in size or resolved.33 The mechanism underlying the absence of hypothyroidism in these patients is not well understood, but they appear to reach a new steady state, manifested by a normal serum T4 level and goiter, presumably maintained by TSH. The erratic availability of seaweed also may be important because seaweed is a significant part of the diet only during good harvesting weather.12 A similar incidence of hyperthyroidism in five coastal regions of Japan has been reported, along with a widely varying incidence of hypothyroidism (0–9.7%). High urinary iodine concentration correlated with antibody-negative hypothyroidism.34

FIGURE 37-1. A 14-year-old patient from Hok-kaido, Japan, with usual daily diet of 50 to 80 mg of iodine derived from seaweed (kelp). A, Appearance at time of first presentation with “coast goiter.” B, Appearance 3 weeks after removal of seaweed from the diet. C, Appearance after 2 months of treatment with thyroid hormone. The goiter has regressed further. (From Suzuki H, Higuchi T, Sawa K, et al. Endemic coast goiter in Hokkaido, Japan. Acta Endocrinol [Copenh] 1965; 50:161.)

PREDISPOSITION TO INDUCED HYPOTHYROIDISM Congenital and acquired defects in intrathyroidal organification increase the risk of induced hypothyroidism with iodine exposure. Patients with simple goiter resulting from organification defects appear to be highly susceptible to iodide-induced hypothyroidism. A prospective study of biochemically euthyroid patients with Hashimoto thyroiditis who were treated with iodides revealed a high incidence of iodide-related hypothyroidism that was reversible on cessation of iodides.29 These thyroid glands probably fail to escape from the acute inhibitory effect of iodide on the organification process and manifest a persistence of the Wolff-Chaikoff effect. The increasing incidence of Hashimoto thyroiditis with hypothyroidism or goiter in the United States may be the result of an unmasking of the organification defect with a chronically high dietary iodide intake. Hypothyroidism also has been seen in euthyroid patients given iodides who previously underwent 131I or surgical therapy for Graves disease.11,35 With withdrawal of iodides, the euthyroid state was restored, and TSH levels normalized. The speculation is that this phenomenon is the result of an organification defect caused by the radiation therapy. Although hypothyroidism after iodide exposure was seen in all patients treated with 131I, it also was seen in 40% of surgically treated patients.35 Thus, an underlying defect related to Graves disease that can be compounded by a radiation-induced sensitivity of the thyroid gland to an iodide load appears to exist. CLINICAL FEATURES OF IODIDE-INDUCED HYPOTHYROIDISM The clinical presentation of iodide-induced hypothyroidism is variable. Goiter alone is much more common than is hypothyroidism, and the thyroid gland may be modestly or significantly enlarged. Generally, the goiter is diffuse, but in the context of antithyroidal antibodies, a nodular gland may be present. A history of previous Hashimoto disease, Graves disease, or goiter should raise the suspicion of an iodide-induced abnormality of thyroid function. The symptoms and signs of hypothyroidism may be subtle or absent.10,34,36 In the presence of normal to marginal values for serum T4 and basal TSH, the serum TSH response to thyrotropin-releasing hormone affords the most sensitive indicator of hypothyroidism. Iodine-131 uptake may not be a helpful measure, depending on the amount of exogenous iodide ingested and the functional capabilities of the thyroid gland. In a setting of high iodide intake, a low uptake of radioiodine is expected. Low, normal, and elevated uptakes have all been seen, however, with iodide-induced goiter. An early peak in 131I uptake has been described, with the 4-hour or 6-hour uptake higher than the 24-hour uptake, a typical finding in patients with organification defects. The best evidence that goiter or hypothyroidism is related to excess iodides is return to normal function and size of the gland on withdrawal of exogenous iodide. Patients at highest risk are those with underlying thyroid abnormalities, Hashimoto disease, or previously treated Graves disease. A Japanese study of iodine-induced hypothyroidism indicated that patients with hypothyroidism that was reversible after iodine withdrawal had underlying focal lymphocytic thyroiditis, whereas those with irreversible hypothyroidism had more severe destruction of the thyroid gland.37

IODIDE-INDUCED THYROTOXICOSIS The occurrence of hyperthyroidism after iodide administration has been reported throughout the world since this element was first used to treat endemic iodine-deficient goiter. The first case report of thyrotoxicosis after iodine supplementation was published in 1821. Iodide-induced hyperthyroidism subsequently has been referred to as

the jodbasedow phenomenon. Thyrotoxicosis occurring after iodide supplementation for endemic goiter in iodine-deficient areas is frequently reported. The mechanism underlying this phenomenon may relate to the rapid iodination and proteolysis of previously iodine-poor thyroglobulin, or to the presence of a subpopulation of patients with goiters who have gross or microscopic autonomous areas of functioning tissue. The latter hypothesis is supported by the observation that hyperthyroidism may develop after an iodine load in euthyroid patients with solitary autonomous nodules.38 Four populations—in Holland, Tasmania, Yugoslavia, and Australia—were observed closely for the prevalence of thyrotoxicosis before and after iodide supplementation. The incidence of thyrotoxicosis after iodide supplementation ranged from .01% to .04%.14 The incidence of thyrotoxicosis in a population rises 6 months after long-term iodine exposure begins, peaks at 1 to 3 years, and returns to baseline by 6 to 10 years. Although several reports have indicated no increased incidence of thyrotoxicosis in some populations given iodine prophylactically, these follow-up studies may have been conducted too late to detect transient hyperthyroidism. Reviews have addressed the prevention of iodine-induced thyrotoxicosis39 and its epidemiology in Europe and worldwide.40 Iodine-induced thyrotoxicosis has been reported from areas of endemic iodine deficiency in Germany after the use of iodine-containing drugs or contrast media, in much the same manner as was seen when iodine was added as a dietary supplement.41 The reported 5% to 30% incidence in this area of endemic goiter in Germany is much higher than that seen in iodine-sufficient areas. Follow-up studies on patients exposed to iodine-containing drugs or dyes indicate that serum iodothyronine concentrations increase rapidly in relation to iodine dose, with a doubling of free T4 index and a 53% increase in serum T3 at a time when urinary iodine excretion is increased 5-fold to 10-fold over baseline.42 Additional exposure to excess iodine intake further increased serum iodothyronine concentrations, albeit less dramatically.42 In addition, the incidence of T3 toxicosis relative to that of T4 toxicosis declined with increasing iodine intake. In the United States, concern exists that iodine-induced thyrotoxicosis may occur in persons without any underlying goiter or thyroid function abnormality. The incidence of goiter is ~3% in this country, and no areas of iodine deficiency are found. Dozens of cases of drug-induced thyrotoxicosis have been reported in the United States. Far more cases have come from Europe, although these may be from areas of relative dietary iodine deficiency. Several of the patients described in the United States have had preexisting thyroid disorders, most have had multinodular goiters, and most have been women. Drug-induced thyrotoxicosis has been seen after the use of potassium iodide, oral and intravenous contrast agents, the iodine-containing antiarrhythmic amiodarone (see preceding), and the topical antiinfective iodochlorhydroxyquin. Hyperthyroidism has been observed to develop 4 days to 4 months after iodine exposure. The thyrotoxicosis may persist for 1 to 6 months and usually is self-limited. In one study, thyrotoxicosis developed in four of eight patients with goiter after a 180-mg daily dose of potassium iodide, and the suggestion was made that thyrotoxicosis may be a common outcome in patients from nonendemic areas with goiter who are exposed to iodides.43 The authors recommended that caution be exercised in administering iodine-containing drugs or performing radiographic contrast procedures in persons with goiter. Patients with multinodular goiter are at some increased risk for iodine-induced thyrotoxicosis. The risk may be overstated, however; despite the high prevalence of goiter and the several million dye studies that are done yearly in this country, an epidemic of thyrotoxicosis has not occurred. Thus, most patients with goiter are not clinically affected by iodine exposure, and goiter should not be considered a contraindication to the use of necessary iodine-containing drugs or contrast media. In a study of 10 women who were in remission after a prior history of Graves thyrotoxicosis, the administration of SSKI caused recurrence of thyrotoxicosis in two patients and blunting of the thyrotropin-releasing hormone stimulation test in two others.44 Iodine-induced thyrotoxicosis also has occurred with metastatic thyroid carcinoma after earlier thyroidectomy.45 CLINICAL FEATURES OF IODIDE-INDUCED THYROTOXICOSIS The clinical features of iodide-induced thyrotoxicosis in endemic goiter are similar to those seen in Graves disease with thyrotoxicosis. In the former case patients tend to be older, however, and to have long histories of iodide deprivation preceding the addition of iodine. Some predisposition may be present with advancing age, because thyrotoxicosis occurred in none of 50,000 children in this country who were treated with iodides in the 1920s. The ratio of men to women affected may also be different from the ratio in Graves disease. In Yugoslavia, the ratio of men to women with thyrotoxicosis increased from 1:10 to 1:6 after iodine supplementation. In Holland, 15% of the cases of hyperthyroidism occurred in men before iodine supplementation, compared to 31% after iodine supplementation. Exophthalmos is not encountered, and thyroid antibodies are not detected. Among patients with iodide-induced thyrotoxicosis from iodine-deficient areas, most patients had long-standing nodular goiter; however, 15% to 30% of patients had minimal to no goiter, and many persons with goiter did not have nodules. The clinical manifestations of drug-induced thyrotoxicosis are similar to those of Graves hyperthyroidism and include tremor, tachycardia, palpitations, weight loss, heat intolerance, concentration difficulties, and mental status changes.19 There is no tenderness of the thyroid gland such as is seen in granulomatous thyroiditis; there is no exophthalmos such as is seen in Graves disease. The serum T4 and free T4 values and the free T4 index are elevated. Typically, the 24-hour 131I uptake is reduced, and the technetium thyroid scan may show reduced or patchy uptake. Generally, the T3 level also is elevated; however, with amiodarone and the oral cholecystographic dye ipodate, the peripheral conversion of T4 to T3 is blocked, and the serum T3 level may be disproportionately low compared to the elevated serum T4 level. TREATMENT OF IODIDE-INDUCED THYROTOXICOSIS The treatment of thyrotoxicosis is discussed in Chapter 42. The major difference in the approach to treating iodide-induced hyperthyroidism as opposed to Graves disease relates to the self-limited nature of the former, with resolution occurring after removal of the source of excess iodine and clearance of iodine from the circulation. The thyrotoxicosis, although usually self-limited, may persist for weeks to months because of the increased thyroidal iodine stores and increased plasma iodide. Mildly symptomatic patients may be treated with b-blockade therapy alone; more specific antithyroid drugs may be necessary in severely affected patients, but ablative therapy should not be required. Amiodarone iodine-induced thyrotoxicosis may be resistant to the usual antithyroidal therapy.46 A study of 58 patients with amiodarone-induced thyrotoxicosis showed persistent thyrotoxicosis at 6 to 9 months in more than one-half of those who were not treated. The use of methimazole in high doses proved ineffective in most patients. Euthyroidism was restored, however, in all patients who were treated with a combination of methimazole and potassium perchlorate (1 g per day). Other authors have found prolonged persistent thyrotoxicosis even with the addition of potassium perchlorate to propylthiouracil.47 Successful resolution of the hyperthyroidism may require prolonged therapy. Of note for those cases in which the underlying cardiac dysrhythmia precludes safe discontinuation of the drug, the thyrotoxicosis was treated successfully in five patients by adding only thiourea.48 Although most cases of amiodarone-induced thyrotoxicosis are related to the provision of its rich iodine content to patients who have thyroid glands with underlying autonomous function (e.g., hyperfunctioning nodules), a variant of the syndrome that is akin to destructive thyroiditis exists.49 Patients with this condition, who may have low radioiodine uptake and elevated serum interleukin-6 levels,50 should be treated with glucocorticoids rather than methimazole.51

IODINE-INDUCED AUTOIMMUNITY Investigators have speculated that increased iodine intake may contribute to the prevalence of autoimmune thyroid disease. Individuals with Hashimoto disease are well known to be sensitive to the effects of iodine and vulnerable to iodine-induced hypothyroidism. Less well known is that iodine exposure may induce a lymphocytic thyroiditis in a variety of experimental animals (rat, hamster, dog) as well as in humans.9,52 A few studies have examined the greater prevalence of thyroid antibodies after increases in iodine exposure.53 Specific histologic features of iodine-induced goiter have been described in one study of 28 patients, with half the biopsies also showing typical lymphocytic infiltration.54 Persons with preexisting thyroid disease who live in areas of iodide deficiency may be predisposed to the development of iodine-induced thyrotoxicosis.52 Iodine-rich thyroglobulin has been shown to be more immunogenic,55 and a greater proliferation of T cells is seen in patients with chronic thyroiditis after exposure to normally iodinated thyroglobulin, whereas noniodinated thyroglobulin elicits no response.56 In some unknown manner, iodine may relate to the development of thyroid growth-stimulating immunoglobulins.57 In this regard, iodide has been shown to enhance immunoglobulin G release into the media of cultured human lymphocytes,58 but other mechanisms of immune stimulation may exist.56 Controversy exists as to whether amiodarone may induce autoimmunity and, if so, whether the drug itself or its iodine content is responsible. Thirty percent to 50% of patients who become hypothyroid during amiodarone treatment have antithyroid antibodies in their serum. In one prospective study, antibodies developed in 6 of 13 patients given amiodarone, compared to none of 22 control subjects.59 The proposal has been made that amiodarone alters T-cell subsets with enhanced T-cell expression of Ia antigen, consistent with precipitation of organ-specific autoimmunity in susceptible persons.60 CHAPTER REFERENCES 1. 2. 3. 4.

Braverman LE. Iodine and the thyroid: 33 years of study. Thyroid 1994; 4:351. Koutras DA. Control of efficiency and results and adverse effects of excess iodine administration on thyroid function. Ann Endocrinol (Paris) 1996; 57:463. Khan LK, Li R, Gootnick D, et al. Thyroid abnormalities related to iodine excess from water purification units. Lancet 1998; 352:1519. Hollowell JG, Staehling NW, Hannon WH, et al. Iodine nutrition in the United States. Trends and public health implications: iodine excretion data from national health and nutrition examination surveys I and III (1971– 1974 and 1988–1994). J Clin Endocrinol Metab 1998; 83:3401. 5. Cavalieri RR. Iodine metabolism and thyroid physiology: current concepts. Thyroid 1997; 7:177. 6. Ermans AM. Disorders of iodine deficiency. In: Ingbar SH, Braverman LE, eds. The thyroid: a fundamental and clinical text. Philadelphia: JB Lippincott, 1986:707. 7. Peacock I, Davison H. Observations of iodide sensitivity. Ann Allergy 1958; 16:158.

8. 9. 10. 11. 12. 13. 14. 15.

Edmunds HT. Acute thyroiditis from potassium iodide. BMJ 1955; 1:354. Bagchi N, Brown TR, Urdanivia E, Sundick RS. Induction of autoimmune thyroiditis in chickens by dietary iodine. Science 1985; 230:325. Sato K, Okamura K, Hirata T, et al. Immunological and chemical types of reversible hypothyroidism; clinical characteristics and long-term prognosis. Clin Endocrinol 1996; 45:519. Vagenakis AG, Braverman LE. Adverse effects of iodides on thyroid function. Med Clin North Am 1975; 59:1075. Wolff J. Iodide goiter and the pharmacologic effects of excess iodide. Am J Med 1969; 47:101. Namba H, Yamashita S, Kimura H, et al. Evidence of thyroid volume increase in normal subjects receiving excess iodide. J Clin Endocrinol Metab 1993; 76:605. Fradkin JE, Wolff J. Iodide induced thyrotoxicosis. Medicine (Baltimore) 1983; 62:1. Savoie J-C, Massin JP, Thomopoulos P, Leger F. Iodine induced thyrotoxicosis in apparently normal glands. J Clin Endocrinol Metab 1975; 4:685.

15a. Jooste PL, Weight MJ, Lombard CJ. Short-term effectiveness of mandatory iodization of table salt, at an elevated iodine concentration, on the iodine and goiter status of school children with endemic goiter. Am J Clin Nutr 2000; 71:75. 16. Park YK, Harland BF, Vandervein JE, et al. Estimation of dietary iodine intake of Americans in recent years. J Am Diet Assoc 1981; 79:17. 17. Burgi H, Wimpfheimer C, Burger A, et al. Changes of circulating thyroxine, triiodothyronine, and reverse triiodothyronine after radiographic contrast agents. J Clin Endocrinol Metab 1976; 43:1203. 18. Nagataki S. Effect of excess quantities of iodide. In: Greer MA, Solomon DH, eds. Handbook of physiology, section 7, Endocrinology, VIII. Baltimore: Williams & Wilkins, 1974:329. 19. Safran M, Braverman LE. Effect of chronic douching with polyvinyl-pyrrolidone-iodine on iodine absorption and thyroid function. Obstet Gynecol 1982; 60:35. 20. Rajatanavin R, Safran M, Stoller WA, et al. Five patients with iodine-induced hyperthyroidism. Am J Med 1984; 77:378. 21. Loh KC. Amiodarone-induced thyroid disorders: a clinical review. Post-grad Med J 2000; 76:133. 22. Harjai KJ, Licata AA. Effects of amiodarone on thyroid function. Ann Intern Med 1997; 126:63. 23. Iudica-Souza C, Burch HB. Amiodarone-induced thyroid dysfunction. The Endocrinologist 1999; in press. 24. Leger AF, Massin JP, Laurent ME, et al. Iodine-induced thyrotoxicosis: analysis of eighty-five consecutive cases. Eur J Clin Invest 1984; 14:449. 25. Paul T, Myers B, Witorsch RJ, et al. The effect of small increases in dietary iodine on thyroid function in euthyroid subjects. Metabolism 1988; 37:121. 26. Wolff J, Chaikoff IL. Plasma inorganic iodide as a homeostatic regulator of thyroid function. J Biol Chem 1948; 174:555. 27. Gardner DF, Centor RM, Utiger RD. Effects of low dose oral iodide supplementation on thyroid function in normal men. Clin Endocrinol (Oxf) 1988; 28:283. 28. Wartofsky L, Ransil BJ, Ingbar SH. Inhibition by iodine of the release of thyroxine from the thyroid glands of normal subjects and patients with thyrotoxicosis. J Clin Invest 1970; 49:78. 29. Braverman LE, Ingbar SH, Vagenakis AG. Enhanced susceptibility to iodide myxedema in patients with Hashimoto's disease. J Clin Endocrinol Metab 1971; 32:515. 30. DeLange F, Chanoine JP, Abrassart C, Bourdoux P. Topical iodine, breast-feeding, and neonatal hypothyroidism. Arch Dis Child 1988; 63:106. 31. Chabrolle JP, Rossier A. Goiter and hypothyroidism in the newborn after cutaneous absorption of iodine. Arch Dis Child 1978; 53:495. 32. Suzuki H, Higuchi T, Sawa K, et al. Endemic coast goiter in Hokkaido, Japan. Acta Endocrinol (Copenh) 1965; 50:161. 33. Higuchi T. The study of endemic seashore goiter in Hokkaido. Folia Endo-crinol Jpn 1964; 40:982. 34. Konno N, Makita H, Yuri K, et al. Association between dietary iodine intake and prevalence of subclinical hypothyroidism in the coastal regions of Japan. J Clin Endocrinol Metab 1994; 78:393. 35. Braverman LE, Walker KA, Ingbar SH. Induction of myxedema by iodide in patients euthyroid after radioiodine or surgical treatment of diffuse toxic goiter. N Engl J Med 1969; 281:816. 36. Lesher JL Jr, Fitch MH, Dunlap DB. Subclinical hypothyroidism during potassium iodide therapy for lymphocutaneous sporotrichosis. Cutis 1994; 53:128. 37. Tajiri J, Higashi K, Morita M, et al. Studies of hypothyroidism in patients with high iodine intake. J Clin Endocrinol Metab 1986; 63:412. 38. Ermans AM, Camus M. Modification of thyroid function induced by chronic administration of iodide in the presence of “autonomous” thyroid tissue. Acta Endocrinol (Copenh) 1972; 70:463. 39. Dunn JT, Semigran MJ, Delange F. The prevention and management of iodine-induced hyperthyroidism and its cardiac features. Thyroid 1998; 8:101. 40. Stanbury JB, Ermans AE, Bourdoux P, et al. Iodine-induced hyperthyroidism: occurrence and epidemiology. Thyroid 1998; 8:83. 41. Kallee E, Wahl R, Bohner J, et al. Thyrotoxicosis induced by iodine-containing drugs. J Mol Med 1980; 4:221. 42. Emrich D, Karkavitsas N, Facorro U, et al. Influence of increasing iodine intake on thyroid function in euthyroid and hyperthyroid states. J Clin Endocrinol Metab 1982; 54:1236. 43. Vagenakis AG, Wang C, Burger A, et al. Iodide-induced thyrotoxicosis in Boston. N Engl J Med 1972; 287:524. 44. Roti E, Gardini E, Minelli R, et al. Effects of chronic iodine administration on thyroid status in euthyroid subjects previously treated with antithyroid drugs for Graves' hyperthyroidism. J Clin Endocrinol Metab 1993; 76:928. 45. Yoshinari M, Tokuyama T, Okamura K, et al. Iodide-induced thyrotoxicosis in a thyroidectomized patient with metastatic thyroid carcinoma. Cancer 1988; 61:1674. 46. Martino E, Aghini-Lombardi F, Mariotti S, et al. Amiodarone: a common source of iodine induced thyrotoxicosis. Horm Res 1987; 26:158. 47. Newnham HH, Topliss BJ, LeGrand BA, et al. Amiodarone-induced hyperthyroidism: assessment of the predictive value of biochemical testing and response to combined therapy using propylthiouracil and potassium perchlorate. Aust N Z J Med 1988; 18:37. 48. Davies PH, Franklyn JA, Sheppard MC. Treatment of amiodarone induced thyrotoxicosis with carbimazole alone and continuation of amiodarone. BMJ 1992; 305:224 49. Smyrk TC, Goellner JR, Brennan MD, Carnei JA. Pathology of the thyroid in amiodarone-associated thyrotoxicosis. Am J Surg Pathol 1987; 11:197. 50. Bartalena L, Grasso L, Brogioni S, et al. Serum interleukin-6 in amiodarone-induced thyrotoxicosis. J Clin Endocrinol Metab 1994; 78:423. 51. Bartalena L, Brogioni S, Grasso L, et al. Treatment of amiodarone-induced thyrotoxicosis, a difficult challenge: results of a prospective study. J Clin Endocrinol Metab 1996; 81:2930. 52. Harach HR, Escalante DA, Onativia A, et al. Thyroid carcinoma and thyroiditis in an endemic goitre region before and after iodine prophylaxis. Acta Endocrinol (Copenh) 1985; 108:55 53. Koutras DA, Karaiskos KS, Evangelopoulou K, et al. Thyroid antibodies after iodine supplementation. In: Drexhage HA, Wiersinga WM, eds. The thyroid and autoimmunity. Amsterdam: Excerpta Medica, 1986:211. 54. Mizukami Y, Michigishi T, Nonomura A, et al. Iodine-induced hypothyroidism: a clinical and histological study of 28 patients. J Clin Endocrinol Metab 1993; 76:466. 55. Sundick RS, Herdegen DM, Brown TR, Bagchi N. The incorporation of dietary iodine into thyroglobulin increases its immunogenicity. Endocrinology 1987; 120:2078. 56. Rose NR, Saboori AM, Rasooly L, Burek CL. The role of iodine in autoimmune thyroiditis. Crit Rev Immunol 1997; 17:511. 57. Medeiros-Neto GA, Halpern A, Cozzi ZS, et al. Thyroid growth immunoglobulins in large multinodular endemic goiters: effect of iodized oil. J Clin Endocrinol Metab 1986; 63:644. 58. Weetman AP, McGregor AM, Campbell H, et al. Iodide enhances IgG synthesis by human peripheral blood lymphocytes in vitro. Acta Endocrinol (Copenh) 1983; 103:210. 59. Monteiro E, Galvao Teles A, Santos ML, et al. Antithyroid antibodies as an early marker for thyroid disease induced by amiodarone. BMJ 1986; 292:227. 60. Rabinowe SL, Larsen PR, Antman EM, et al. Amiodarone therapy and autoimmune thyroid disease. Am J Med 1986; 81:53.

CHAPTER 38 NONTOXIC GOITER Principles and Practice of Endocrinology and Metabolism

CHAPTER 38 NONTOXIC GOITER PAUL J. DAVIS AND FAITH B. DAVIS Pathogenesis of Nontoxic Goiter Specific Etiologies of Goiter Congenital Goiter Endemic Goiter Sporadic Goiter Clinical Features of Nontoxic Goiter History and Physical Findings Nonthyroidal Anterior Cervical Neck Masses Laboratory Evaluation of the Patient with Goiter Establishment of the Cause of Nontoxic Goiter Prognosis of Nontoxic Goiter Treatment of Nontoxic Goiter Substernal Goiter Chapter References

Goiter (L. guttur, throat) is a clinical term denoting thyroid gland enlargement to twice normal size or larger (see also ref. 1). The term implies no specific etiology, although the mechanisms of certain toxic and nontoxic (eumetabolic) goitrous states are understood. The prevalence of nontoxic goiter in the United States is ~5%,1a with a predominance among females. In the Framingham, Massachusetts, study, the prevalence of non-toxic goiter was 4.2%, with a 4:1 female to male ratio.2

PATHOGENESIS OF NONTOXIC GOITER Nontoxic goiter is the result of an interplay of factors intrinsic and extrinsic to the thyroid. In most patients, subtle limitations on hormonogenesis—imposed, for example, by iodine insufficiency or constitutional biochemical abnormalities in the gland—lead to nontoxic goiter when permissive normal or increased circulating levels of thyroid-stimulating hormone (TSH, thyrotropin) foster gland enlargement. Impaired hormonogenesis, TSH-dependent goiter, and eumetabolism may also reflect the presence of Hashimoto thyroiditis or the ingestion of goitrogens. In the absence of TSH, as in hypopituitarism, nontoxic goiter is rare. Some have argued that the mechanisms of diffuse and nodular nontoxic goiter are different, with the latter reflecting an autonomous state independent of the permissive effect of TSH.3 Although gland autonomy is present in certain patients with nontoxic nodular goiters, it occurs in few such patients, and whether autonomy develops early or late in the course of goitrogenesis is unknown.4 The suggestion has also been made that the cyclic hypersecretion of TSH in response to periodic underproduction of hormone by the thyroid gland is essential to the development of nodular goiter. Formation of fibrous strands or scar tissue (consequent to cyclic follicular hyperplasia, microscopic hemorrhage, and necrosis) has been postulated to impose the anatomical limits within which nodules form.5 Nevertheless, the possibility seems likely that subtle constitutional derangements of hormonogenesis and the permissive action of circulating TSH act together to produce both diffuse and nodular nontoxic goiter. Various abnormalities of hormone production have been described in cases of nontoxic goiter (Table 38-1).6,7,8,9,10,11 and 12,12a These biochemical abnormalities, or altered sensitivity to TSH, might be expressed only in certain follicles (biochemical follicular heterogeneity), which would explain the occurrence of nodular goiter. Excluded from Table 38-1 are genetically determined biochemical lesions that exist in familial goiter with hypothyroidism (see Chap. 47).

TABLE 38-1. Biochemical Findings Described in Nontoxic Goitrous Tissue

Other mechanisms leading to the development of goiter are infiltrative diseases of the thyroid, such as amyloidosis or granulomatous diseases, and the action of nonpituitary thyrotropic substances. Among the latter are polypeptides that circulate in pregnancy (human chorionic gonadotropin [hCG] and, less importantly, a placental TSH-like substance13) and thyroid-stimulating antibodies (see Chap. 197). Some of these thyroid-stimulating immunoglobulins (hTSI) have been identified in human serum14,15,16,17,18,19,20,21 and 22 and act either through the thyroid cell membrane TSH receptor or by enhancing thyroid cell activity independently of the TSH receptor. The hTSIs that interact with the TSH receptor are recognized in vitro in a competitive binding assay involving radiolabeled TSH and thyroid cell membranes (TSH-binding inhibitory immunoglobulin [TBII] assay). Thyroid-stimulating antibodies that are inactive in the TBII assay are recognized by their “growth” effects, measured in vitro by sensitive histocytochemical methods or other methods. The hTSIs that cross-react with an antigenic site at the TSH receptor are probably unimportant in the pathogenesis of nontoxic goiter,15 although they play a role in diffuse toxic goiter. On the other hand, thyroid-stimulating antibodies that act independently of the TSH site are found in the sera of more than 50% of patients with nontoxic goiter. Although these observations support an autoimmune contribution to the pathogenesis of nontoxic goiter, a search for helper T lymphocytes sensitized to thyroid membrane antigens in the peripheral blood of patients with nontoxic nodular goiter yielded negative results.15 Hence, the importance of hTSIs in the formation of nontoxic goiter is not yet clear. Certain cytokines have been implicated in the pathogenesis of nontoxic goiter. For example, the content of insulin-like growth factor-I (IGF-I) is increased more than two-fold in thyroid nodules compared with normal tissue obtained from the same gland.17 Thyroid glands from patients with Graves disease and Hashimoto thyroiditis show no increase in IGF-I. Transforming growth factor-b (TGF-b) has been proposed to be an autocrine growth inhibitor in thyroid follicular cells, at least in the setting of iodide deficiency.18 TGF-b blocks the growth stimulation, measured as thymidine incorporation, imposed on thyroid follicular cells in vitro by IGF-I, epidermal growth factor, and TGF-a, and endogenous TGF-b levels are decreased in iodide-deficient nontoxic goiter.18 Although these studies suggest that cytokines can promote or mediate goiter formation, evidence exists that at least in goiter due to iodine-insufficiency, thyroid growth–promoting factors are not detectable in serum.23 The goiter of pregnancy may be a response both to placental TSHs and to maternal iodide insufficiency. The latter is a complex function of increased maternal renal iodide clearance, iodide parasitism by the fetus, and maternal losses during lactation.

SPECIFIC ETIOLOGIES OF GOITER CONGENITAL GOITER Congenital goiter is either familial—that is, an expression of genetic disorders of intrathyroidal hormonogenesis (see Chap. 47)—or sporadic. In the neonate, sporadic congenital goiter reflects intrauterine iodide insufficiency or fetal exposure to goitrogens. Maternal consumption of naturally occurring goitrogens is rarely the cause of thyroid enlargement in the neonate; this contrasts with endemic goiter, in which community-wide ingestion of a goitrogen occurs. Use of antithyroid medications is a significant cause of sporadic congenital goiter (i.e., thionamide administration to pregnant thyrotoxic women). Thionamides cross the placental barrier and limit fetal hormonogenesis after the 12th week of gestation. The requisites for such a congenital goiter include fetal TSH production, maturation of the fetal thyroid to the point of TSH responsiveness, and sensitivity of the pituitary to low levels of thyroid hormone. Because insubstantial quantities of maternal thyroxine (T4) or triiodothyronine (T3) cross the placenta, the maternal ingestion of T4 or T3 concomitantly with thionamides does not prevent fetal goiter. A specialized metabolically active analog,

3,5-dimethyl-3¢-isopropyl-L -thyronine (DIMIT),24 does cross the placenta, but it is not available for clinical use. Other iatrogenic goitrogens, such as aminoglutethimide and carbutamide, do not appear to cause fetal goiter, and neither agent is likely to be prescribed to pregnant women. Lithium and amiodarone, both potential causes of fetal goiter, are contraindicated during pregnancy. ENDEMIC GOITER Iodine deficiency has been extensively studied as a cause of endemic goiter.25 Before the introduction of iodized salt in 1925, “goiter belts” of low iodine content in water supplies existed in the United States near the Great Lakes, in the Appalachian region, and elsewhere.26 During World War I, a 5% incidence of goiter was seen among U.S. Army inductees,27 presumably reflecting widespread iodine deficiency. By World War II, the nationwide effort to increase dietary iodine had reduced the incidence of goiter in military recruits to 0.06%. Dietary iodine insufficiency no longer is a significant cause of euthyroid goiter in the United States but persists in enclaves in South America, Africa, and Asia,1,28 and as many as 200 million people worldwide are estimated to experience iodine-deficiency goiter. The excessive dietary intake of iodine is an occasional but nonendemic cause of goiter in the United States. In a localized region in Japan, however, high iodine intake has caused endemic goiter. The inhibition of thyroid hormonogenesis by excess iodine is termed the Wolff-Chaikoff effect. Severe fetal iodine deprivation culminates in endemic and sporadic cretinism (see Chap. 47), with distinctive clinical findings. Moderate fetal iodine insufficiency results in euthyroid fetal goiter. Iodide crosses the placenta, and parasitism of maternal iodide by the fetus contributes to the “goiter of pregnancy” in women whose dietary iodine intake is marginal. Both well-characterized and poorly characterized dietary goitrogens have been reported in a variety of epidemiologic studies. Water supplies have been a frequent source.29 Certain vegetables (e.g., cabbage and other members of the Brassica family—turnip, kale, rape) contain thionamide-like substances.1 Cassava is an important goitrogen in Africa, but no evidence exists for dietary goitrogens as a cause of endemic goiter in the United States. Most patients with endemic goiter have no constitutional abnormalities of thyroid hormonogenesis. The goiter formation reflects either extrinsic iodine deficiency or goitrogen ingestion. SPORADIC GOITER As with endemic goiter, sporadic goiter may reflect variations in iodide intake30 and dietary goitrogen content. Although iodine insufficiency rarely causes thyroid enlargement in the United States, iodine excess can lead to goiter in two settings. First, although the thyroid gland escapes from the Wolff-Chaikoff effect in normal subjects after several weeks of excessive iodine intake, the thyroid in some patients does not. Resulting defective hormonogenesis and release cause increased pituitary TSH secretion and goiter formation. A new steady state is achieved in which the enlarged, inefficient gland produces sufficient hormone to maintain euthyroidism. This proposed pathogenetic sequence, based on increased TSH production, seems likely, but this hypothesis has not been validated by prospective application of sensitive serum TSH assays to populations at risk. Patients at risk for goiter, due to failure to escape from iodine inhibition of thyroid hormonogenesis, are those with Hashimoto thyroiditis or with other gland damage syndromes, such as those due to thyroid surgery or radiation of the neck.31 A second setting in which goiter occurs with excess iodine administration is jodbasedow,32 a syndrome of latent hyperthyroidism that is rare in the United States (see Chap. 37). Increasing iodine intake promotes goiter and, nearly concomitantly, thyrotoxicosis. The source of increased iodine intake in goitrous patients is either medicinal—for example, saturated solution of potassium iodide (SSKI) continues to be used by some physicians as a mucolytic agent in patients with chronic lung disease, despite its clinical ineffectiveness—or paramedicinal—for example, from kelp ingestion. Radiologic contrast media and skin antiseptics contain iodine, but, because of their short-term use, are not causes of goiter. Long-term iodine excess followed by the acute interruption of iodine intake can have a distinct effect on thyroid function tests. A variety of medications have antithyroid and goitrogenic effects (Table 38-2).33,34,35,36,37,38 and 39 The mechanisms of the antithyroid actions of these agents can be interference with hormonogenesis at the level of organification of iodide or inhibition of hormone release or both. Lithium, a monovalent cation, has an effect on the thyroid similar to that of the anion iodide: hormone release is inhibited after the drug is concentrated in the gland.36,40 Amiodarone, an antiarrhythmic drug, has been associated with the induction of hyperthyroidism and hypothyroidism.37 Most patients who receive the drug remain euthyroid. The effects of amiodarone appear to be largely due to its iodine content, and this agent appears to promote hyperthyroidism in patients with limited, possibly insufficient, iodine intake. Aminoglutethimide35 and oral sulfonylureas39 interfere with hormonogenesis. Among the sulfonylureas, only carbutamide, which is used in Europe, causes goiter. Aminoglutethimide, an antiadrenal agent used as medical treatment of states of excess adrenocortical hormone production, can cause hypothyroidism and goiter.

TABLE 38-2. Drugs Reported to Cause Goiter in Humans

Thyroiditis syndromes may be associated with goiter (see Chap. 46). The thyroid gland in these conditions is frequently symptomatic locally, and the glandular enlargement may be due to inflammatory infiltration rather than to TSH. The differential diagnosis of nontoxic goiter must always include early hyperthyroidism or, particularly in elderly subjects, monosystemic or apathetic hyperthyroidism. In these states, clinical hypermetabolism may be absent and isolated features such as myopathy, cardiomyopathy, or weight loss may be encountered41 (see Chap. 199). Pregnancy is also a cause of nontoxic goiter, which usually remits postpartum. After multiple pregnancies, substantive thyromegaly may occur. The possible contributions of other environmental factors to the development of nontoxic goiter are more difficult to evaluate. For example, cigarette smoking—the inhalation of thiocyanate—has been implicated in goitrogenesis,42 and serum thyroglobulin levels appear to be higher in smokers than in nonsmokers. The incidence of goiter in patients with end-stage renal disease is variable43 and unlikely to be explained by a single factor. The variability in incidence appears to be regional and may relate to diet or to the use in dialysis baths of tap water that contains ions or other factors that are goitrogenic.

CLINICAL FEATURES OF NONTOXIC GOITER HISTORY AND PHYSICAL FINDINGS Patients with nontoxic goiter have no systemic symptoms, and the goiter is usually found on routine physical examination or is detected by the patient as a neck mass. Local symptoms occasionally include those due to tracheal compression (stridor or respiratory distress) and/or displacement of the esophagus (dysphagia). Rarely, substernal extension of nontoxic goiter may lead to recurrent laryngeal nerve entrapment and hoarseness. Spontaneous hemorrhage into a goiter manifests as local pain and tenderness in one thyroidal lobe. When Hashimoto thyroiditis causes nontoxic thyroidal enlargement, the gland is usually without local symptoms but diffuse tenderness or adenopathy occasionally occurs (see Chap. 46). History-taking from patients with nontoxic goiter includes a focused review of possible environmental or genetic factors that may be causative. The emphasis is on dietary or medicinal sources of excessive iodine intake, antithyroid drugs, or, in regions of the world where iodine deficiency or naturally occurring dietary goitrogens

are common, on establishment of a possible role for these factors. Goiter during pregnancy or the enhancement of established thyromegaly in successive pregnancies should raise the possibility of marginal iodine deficiency. Low-dose head-and-neck irradiation in childhood increases the risk of nodular goiter and thyroid carcinoma.44 Systemic illnesses such as sarcoidosis or amyloidosis raise the rare possibility of infiltrative disease of the thyroid. Isolated thyroid nodules in the euthyroid patient often indicate adenomas or cysts (see Chap. 39). Cancer originating in other organs may (rarely) metastasize to the thyroid.45 At autopsy, infection of the thyroid is relatively common in patients who have succumbed to septicemia, but such involvement is seldom clinically apparent before death.46 Physical examination of the nontoxic goiter should include the following: size, consistency, the absence or presence of nodularity, bruit, tenderness, and concomitant anterior cervical lymphadenopathy. The clinical and etiologic implications of each of these factors are controversial, but their accurate description is useful in the subsequent evaluation of the patient. Extreme firmness of a diffuse goiter is suggestive of infiltrative disease (amyloidosis, Riedel struma) or intense TSH stimulation (Hashimoto thyroiditis); firmness of a nodule has been suggested to be a sign of thyroid cancer. However, in the individual patient these generalizations are frequently incorrect. Tenderness of the thyroid gland can be a useful clinical finding, indicating inflammatory disease or intrathyroidal hemorrhage, usually into a preexisting cyst. Tenderness may also reflect acute thyroiditis contiguous with intercurrent tracheitis and occasionally is encountered in Hashimoto thyroiditis. Granulomatous or subacute thyroiditis is usually associated with exquisite thyroidal tenderness and pain radiating to the mandible. In all age groups, anterior cervical and supraclavicular adenopathy is a hallmark of locally metastatic papillary thyroid carcinoma (see Chap. 40). Because a centrally located node immediately above the isthmus of the thyroid gland drains both lobes, it has been ascribed “oracular” qualities (delphian node), usually reflecting the presence of thyroid cancer. However, this node is sometimes enlarged in patients with Hashimoto thyroiditis. Dullness to percussion over the upper sternum may be noted when a large substernal goiter is present. Instructive examples of nontoxic goiters are shown in Figure 38-1, and radiologic findings associated with substernal goiter are depicted in Figure 38-2.

FIGURE 38-1. Examples of nontoxic goiter. Top left, Sporadic multinodular goiter in an 85-year-old woman. No local symptoms were present. Top right, 65-year-old woman with a 40-year history of nontoxic goiter and thyrotoxicosis of recent onset. Multiple ablative doses of 131I controlled the hyperthyroidism but had no appreciable effect on thyroid gland size. Bottom right, Multinodular goiter with clinical findings of thoracic inlet obstruction and tracheal compression syndrome in a 70-year-old woman. The goiter was treated surgically to relieve obstructive symptoms. Bottom left, Endemic goiter in a 45-year-old Peruvian woman.

FIGURE 38-2. Radiographic and imaging studies in substernal goiter. Patient was a 28-year-old woman who presented with acute bronchitis. Left, Posteroanterior chest roentgenogram revealed right paratracheal mass contiguous with the aortic arch. Middle, Lateral roentgenogram of the chest showed tracheal compression (arrow) by anterior mediastinal mass. Right, Transverse superior mediastinal computed tomographic scan. The trachea (arrow) is enveloped and compressed by goiter, which extended posteriorly to the vertebral body.

NONTHYROIDAL ANTERIOR CERVICAL NECK MASSES The thyroidal origin of neck masses is usually apparent from their location and their movement with the thyroid gland on deglutition. Previous neck surgery, obesity, and hypertrophied bellies of sternocleidomastoid muscles can obscure the thyroidal nature of neck masses. Included in the differential diagnosis of neck lesions are branchial cleft cyst and thyroglossal duct cyst (Fig. 38-3). Branchial cleft cysts are anterolateral, whereas thyroglossal duct cysts are midline and are found at any point between the isthmus and the area above the laryngeal prominence (cartilage). When thyroglossal duct cysts remain contiguous with the base of the tongue (foramen caecum linguae), they move upward with voluntary protrusion of the tongue. Cystic hygroma (diffuse, fluid-filled, multiloculated lymphangioma that is present at birth) arises from the supraclavicular fossa and should not be confused with goiter.

FIGURE 38-3. Abnormal nonthyroidal cervical structures included in the differential diagnosis of goiter. Left, Infected branchial cleft cyst in a 28-year-old woman with a neck mass and fever. The lateral, superior location in the neck is typical for this lesion. Right, Thyroglossal duct cyst in an asymptomatic 42-year-old woman. Lesion was smooth and moved with deglutition and with protrusion of the tongue. The midline or near-midline location and cystic consistency are typical for this lesion. The thyroid gland was not enlarged.

LABORATORY EVALUATION OF THE PATIENT WITH GOITER

The principles of diagnostic evaluation of the patient with non-toxic goiter are: (a) confirmation of eumetabolism; (b) establishment of the cause, insofar as doing so is cost effective; and (c) determination of the impact of the lesion on important non-thyroidal structures in the neck or upper mediastinum. Confirmation of the euthyroid state is essential, regardless of the chronicity of the goiter and its historically eumetabolic nature. The development of hyperthyroidism in the setting of long-standing nontoxic goiter is frequent—thyrotoxicosis may develop in as many as one-third of patients with a large non-toxic goiter of 30 years' duration. Nevertheless, in a large series of elderly subjects with hyperthyroidism,41 patients with long-standing goiter comprised 1.7 µg/kg per day of T4.81 The dose is increased incrementally by 0.025 mg per day every 5 to 6 weeks with TSH monitoring until TSH levels are reduced to the desired range. Usually the degree of TSH suppression need not be to an undetectable level but rather to a level of 0.1 to 0.3 mU/L. This should be the case particularly in postmenopausal women, in whom possible adverse effects of a fully suppressive dosage on bone mineral density are to be avoided. Nodules are assessed for change in size by physical examination (or ultra-sonography if required) every 6 weeks for the first 6 months. The follow-up intervals may be more prolonged when significant decreases in size are observed, extending eventually to annual follow-up.

FIGURE 39-3. Approach to the solitary nodule. *22- to 25-gauge needle, repeated with 18-gauge needle if fluid is obtained. †Evidence of carcinoma (papillary, medullary, poorly differentiated, follicular) or lymphoma. ‡For example, sheets of follicular cells. §For example, small groups of uniform follicular cells with little colloid.|| Selected patients with above-normal surgical risk or other extenuating circumstances may be followed closely with suppressive therapy. ¶Changes in cytologic findings redirect clinician to the appropriate arm of the algorithm. (FNA, fine-needle aspiration; T4, thyroxine.)

FNA biopsy should be repeated immediately when a nodule is found to be enlarging on suppressive therapy, with surgical exploration likely to be inevitable unless cystic fluid or hemorrhage with benign cytology is found. FNA should also be repeated if no significant reduction in nodule size is obtained after 6 to 12 months of suppressive therapy. Of repeat FNA biopsies, 98% confirm the original diagnosis.82 Lesions with benign cytology on initial FNA biopsy may be identified as malignancies in only a very few cases on subsequent repeat aspiration.72,82 However, the

prognosis for neoplasms discovered during a later evaluation is unlikely to be different from the prognosis for neoplasms discovered earlier. The authors believe that reaspiration is preferable to the requirement of nodule size reduction as a confirmation of benignity in such cases and that use of this management approach significantly reduces the frequency of unnecessary surgery. Patients with a history of irradiation present a special situation.83 Historically, these patients have been immediately referred for surgery because of their high cancer rate. Some clinicians now advocate FNA biopsy in the management of these patients, although sufficient evidence for the reliability of benign results is still lacking. The surgical approach to thyroid nodules varies. An ipsilateral lobectomy and isthmusectomy are commonly used for single nodules when the preoperative diagnosis is uncertain. Frequently, histologic evaluation of frozen sections is inconclusive or unreliable, and the final diagnosis requires careful examination of permanent sections. When the ultimate diagnosis is carcinoma, the customary practice is to complete a near-total thyroidectomy within 1 week of the first operation.84 Near-total thyroidectomy is the initial procedure of choice in patients with thyroid nodules and a history of thyroid irradiation (see Chap. 44). When only a lobectomy, or lobectomy and isthmusectomy, is adequate (i.e., with benign histopathology), a question of indication for postoperative T4 therapy often arises. In one group of patients with a history of neck irradiation who underwent partial thyroidectomy for nodules, the T4-treated patients had a nodule recurrence rate of 8.4%, compared with a rate of 35.8% in untreated patients.85 However, T4 therapy may not prevent postoperative recurrence of nontoxic goiter.86,87 The importance of referral to a surgeon highly skilled and experienced in thyroid surgery cannot be overemphasized.88 An alternative nonsurgical approach popularized in Europe is the percutaneous injection of 95% ethanol into thyroid nodules previously proved benign by FNA. This approach has been applied to benign thyroid cysts and hyperfunctioning nodules89,90; it also has been applied to cold nodules.91,92,93,94 and 95 The procedure can be very painful for the patient and has been associated with transient increases in serum thyroglobulin and the thyroid hormones, with self-limited thyrotoxicosis. Fever, local pain, hematomata, and vocal cord paralysis are also possible in inexperienced hands. MANAGEMENT OF THYROID CYSTS Simple thyroid cysts are often successfully drained by thyroid aspiration. In many cases, a single aspiration is curative; in other cases, cysts recur even after multiple aspirations. If the initial aspiration leaves no solid residual, the risk of cancer is very low; an additional therapeutic effect can be obtained in recurrent lesions by instilling sclerosing agents such as tetracycline96 or ethanol.89 No convincing evidence exists that thyroid hormone therapy reduces the likelihood of recurrence. Surgery is usually reserved for cysts >4 cm in diameter or for recurrent cysts that produce local symptoms. In cysts defined only by needle aspiration of at least 1 mL of fluid, it may be prudent to consider surgery for lesions that recur after two or more aspirations, especially if the fluid is hemorrhagic, if the nodules are >3 cm in diameter, or if a residual abnormality remains after maximum drainage.97 MANAGEMENT OF AUTONOMOUS NODULES Hot nodules are commonly identified by thyroid scanning.98 The need for definitive treatment is determined by the degree of functional activity of the nodules rather than by their malignant potential. Nontoxic hot nodules are most often followed clinically. Toxic nodules may be treated with antithyroid drugs temporarily but require definitive therapy with 131I or surgery. The choice between these two modalities remains controversial. Surgery is recommended for those with a history of radiation, for children, and for women of childbearing age. The use of 131I usually reduces the nodule's functional activity, but palpable nodules often persist.99 For patients with toxic nodules, doses of 131I (5–20 mCi) appear to be successful in rendering patients euthyroid, but this may depend on the ratio of dose to nodular area100; higher doses—especially when given to euthyroid patients with autonomous nodules—produce hypothyroidism101 (see Chap. 45). Percutaneous injection of ethanol (see earlier) has also been used with some success.90,102 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.

Ashcraft MW, VanHerle AJ. Management of thyroid nodules. I. History and physical examination, blood tests, x-ray tests, and ultrasonography. Head Neck Surg 1981; 3:216. Ashcraft MW, VanHerle AJ. Management of thyroid nodules. II. Scanning techniques, thyroid suppressive therapy, and fine needle aspiration. Head Neck Surg 1981; 3:297. Mazzaferri EL. Management of a solitary thyroid nodule. N Engl J Med 1993; 328:553. Burch HB. Evaluation and management of the solid thyroid nodule. Endocrinol Metab Clin North Am 1995; 24:663. Sheppard MC, Franklyn JA. Management of the single thyroid nodule. Clin Endocrinol 1992; 37:398. Ridgway EC. Clinical review 30: clinician's evaluation of a solitary thyroid nodule. J Clin Endocrinol Metab 1992; 74:231. Molitch ME, Beck JR, Dreisman M, et al. The cold thyroid nodule: an analysis of diagnostic and therapeutic options. Endocr Rev 1984; 5:185. Hermus AR, Huysmans DA. Treatment of benign nodular thyroid disease. N Engl J Med 1998; 338:1438. Gharib H, Mazzaferri EL. Thyroxine suppressive therapy in patients with nodular thyroid disease. Ann Intern Med 1998; 128:386. Ridgway EC. Medical treatment of benign thyroid nodules: have we defined a benefit? Ann Intern Med 1998; 128:403. Rallison ML, Dobyns BM, Meikle AW, et al. Natural history of thyroid abnormalities: prevalence, incidence, and regression of thyroid diseases in adolescents and young adults. Am J Med 1991; 91:363. Jackson SG, Wartofsky L. The thyroid nodule. In: Monaco F, ed. Thyroid diseases: clinical fundamentals and therapy. Boca Raton, FL: CRC Press, 1993:chap VIIA,253. Singer PA, Cooper DA, Daniels GH, et al. Treatment guidelines for patients with thyroid nodules and well differentiated thyroid cancer. Arch Intern Med 1996; 156:2165. Brander A, Viikinkoski P, Nickels J, Kivisaari L. Thyroid gland: US screening in a random adult population. Radiology 1991; 181:683. Carroll BA. Asymptomatic thyroid nodules: incidental sonographic detection. AJR Am J Roentgenol 1982; 138:499. Horlocker TT, Hay JE, James EM. Prevalence of incidental nodular thyroid disease detected during high resolution parathyroid ultrasonography. In: Medeiros-Neto G, Gaitan E, eds. Frontiers in thyroidology, vol 1. New York: Plenum Press, 1986:1309. Brander A, Viikinkoski P, Nickels J, Kivisaari L. Thyroid gland: ultrasound screening in middle aged women with no previous thyroid disease. Radiology 1989; 173:507. Walfish AG, Hazani E, Strawbridge HTG, et al. Combined ultrasound and needle aspiration cytology in the assessment and management of hypo-functioning thyroid nodules. Ann Intern Med 1977; 87:270. Brown CL. Pathology of the cold nodule. Clin Endocrinol Metab 1981; 10:235. Cooper DS, Schneyer CR. Follicular and Hürthle cell carcinoma of the thyroid. Endocrinol Metab Clin North Am 1990; 19:577. Goldman ND, Coniglio JU, Falk SA. Thyroid cancers I: papillary, follicular, and Hürthle cell. Otolaryngol Clin North Am 1996; 29:593. Hamburger JI. Evolution of toxicity in solitary nontoxic autonomously functioning thyroid nodules. J Clin Endocrinol Metab 1980; 50:1089. Hamburger JI. The autonomously functioning thyroid nodule: Goetsch's disease. Endocr Rev 1987; 8:439. Porcellini A, Ciollo I, Laviola L, et al. Novel mutations of thyrotropin receptor gene in thyroid hyperfunctioning adenomas. J Clin Endocrinol Metab 1994; 76:657. Paschke R, Ludgate M. The thyrotropin receptor in thyroid diseases. N Engl J Med 1997; 337:1675. Duprez L, Hermans J, Van Sande J, et al. Two autonomous nodules of a patient with multinodular goiter harbor different activating mutations of the thyrotropin receptor gene. J Clin Endocrinol Metab 1997; 82:306. Russo D, Arturi F, Suarez HG, et al. Thyrotropin receptor gene alterations in thyroid hyperfunctioning adenomas. J Clin Endocrinol Metab 1996; 1548. Marsh DJ, Learoyd DL, Robinson BG, et al. Medullary thyroid carcinoma: recent advances and management update. Thyroid 1995; 5:407. Heshmati HM, Gharib H, van Heerden JA, Sizemore GW. Advances and controversies in the diagnosis and management of medullary thyroid carcinoma. Am J Med 1997; 103:60. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1998. CA Cancer J Clin 1998; 48:6. McHenry C, Smith M, Lawrence AM, et al. Nodular thyroid disease in children and adolescents: a high incidence of carcinoma. Am Surg 1988; 54:444. Gorlin JB, Sallan SE. Thyroid cancer in childhood. Endocrinol Metab Clin North Am 1990; 19:649. Hung W. Nodular thyroid disease and thyroid carcinoma. Pediatr Ann 1992; 21:50. Robbins J. Lessons from Chernobyl: the event, the aftermath fallout: radioactive, political, social. Thyroid 1997; 7:189. Becker DV, Robbins J, Beebe GW, et al. Childhood thyroid cancer following the Chernobyl accident. A status report. Endocrinol Metab Clin North Am 1996; 25:197. Sarne D, Schneider AB. External radiation and thyroid neoplasia. Endocrinol Metab Clin North Am 1996; 25:181. Smith RE, Adler RA, Clark P, et al. Thyroid function after mantle irradiation in Hodgkin's disease. JAMA 1981; 245:46. Rosen IB, Simpson JA, Sutcliff S, et al. High-dose radiation and the emergence of thyroid nodular disease. Surgery 1984; 96:988. Blum M, Rothschild M. Improved nonoperative diagnosis of the solitary cold thyroid nodule. JAMA 1980; 243:242. Belfiore A, LaRosa GL, LaPorta GA, et al. Cancer risk in patients with cold thyroid nodules: relevance of iodine intake, sex, age, and multinodularity. Am J Med 1992; 93:363. Spencer CA, Wang C-C. Thyroglobulin measurement: techniques, clinical benefits, and pitfalls. Endocrinol Metab Clin North Am 1995; 24:841. Torrens JI, Burch HB. Serum thyroglobulin measurement: utility in clinical practice. Endocrinologist 1996; 6:125. Vierhapper H, Raber W, Bieglmayer C, et al. Routine measurement of plasma calcitonin in nodular thyroid disease. J Clin Endocrinol Metab 1997; 82:1589. Pacini F, Fontanelli M, Fugazzola L, et al. Routine measurement of serum calcitonin in nodular thyroid diseases allows the preoperative diagnosis of unsuspected sporadic medullary thyroid carcinoma. J Clin Endocrinol Metab 1994; 78:826. Saad MF, Fritsche HA Jr, Samaan NA. Diagnostic and prognostic values of carcinoembryonic antigen in medullary carcinoma of the thyroid. J Clin Endocrinol Metab 1984; 58:889. Cavalieri RR. Nuclear imaging in the management of thyroid carcinoma. Thyroid 1996; 6:485. Patton JA, Sandler MP, Partain CL. Prediction of benignancy of the solitary “cold” thyroid nodule by fluorescent scanning. J Nucl Med 1985; 26:461. de los Santos ET, Keyhani-Rofagha S, Cunningham JJ, Mazzaferri EL. Cystic thyroid nodules: the dilemma of malignant lesions. Arch Intern Med 1990; 150:1422. Simeone JF, Daniels GH, Mueller PR, et al. High-resolution real-time sonography of the thyroid. Radiology 1982; 145:431. Tan GH, Gharib H. Thyroid incidentalomas: management approaches to nonpalpable nodules discovered incidentally on thyroid imaging. Ann Intern Med 1997; 126:226. Blum M. Non-isotopic imaging of the neck in patients with thyroid nodules or cancer. In: Thyroid cancer: clinical management. Totowa, NJ: Humana Press, 2000, 9. Bloom AD, Adler LP, Shuck JM. Determination of malignancy of thyroid nodules with positron emission tomography. Surgery 1993; 114:728. Wartofsky L. Management of patients with scan negative, thyroglobulin positive differentiated thyroid carcinoma. In: Wartofsky L, Sherman SI, Gopal J, Schlumberger M, Hay ID, eds. The use of radioactive iodine in patients with papillary and follicular thyroid cancer. J Clin Endocrinol Metab 1998; 83:4195. Gharib H, James EM, Charboneau JW, et al. Suppressive therapy with levothyroxine for solitary nodules. N Engl J Med 1987; 317:70. Cheung PSY, Lee JMH, Boey JH. Thyroxine suppressive therapy of benign solitary thyroid nodules: a prospective randomized study. World J Surg 1989; 13:818. Reverter JL, Lucas A, Salinas I, et al. Suppressive therapy with levothyroxine for solitary thyroid nodules. Clin Endocrinol 1992; 36:25. Berghout A, Wiersinga WM, Drexhage HA, et al. Comparison of placebo with L-thyroxine alone or with carbimazole for treatment of sporadic non-toxic goitre. Lancet 1990; 336:193. Celani MF, Mariani M, Mariani G. On the usefulness of levothyroxine suppressive therapy in the medical treatment of benign solitary, solid, or predominantly solid thyroid nodules. Acta Endocrinol 1990; 123:603. Papini E, Bacci B, Panunzi C, et al. A prospective randomized trial of levothyroxine suppressive therapy for solitary thyroid nodules. Clin Endocrinol 1993; 38:507.

60. LaRosa GL, Lupo L, Giuffrida D, et al. Levothyroxine and potassium iodide are both effective in treating benign solitary solid cold nodules of the thyroid. Ann Intern Med 1995; 122:1. 61. Zelmanovitz F, Genro S, Gross JL. Suppressive therapy with levothyroxine for solitary thyroid nodules: a double-blind controlled clinical study and cumulative metaanalyses. J Clin Endocrinol Metab 1998; 83:3881. 62. Papini E, Petrucci L, Guglielmi R, et al. Long-term changes in nodular goiter: a 5-year prospective randomized trial of levothyroxine suppressive therapy for benign cold thyroid nodules. J Clin Endocrinol Metab 1998; 83:780. 63. Lima N, Knobel M, Cavaliere H, et al. Levothyroxine suppressive therapy is partially effective in treating patients with benign, solid thyroid nodules and multinodular goiters. Thyroid 1997; 7:691. 64. Wartofsky L. Does replacement L-thyroxine therapy cause osteoporosis? Adv Endocrinol Metab 1993; 4:157. 65. Marcocci C, Golia F, Bruno-Bossio G, et al. Carefully monitored levothyroxine suppressive therapy is not associated with bone loss in premenopausal women. J Clin Endocrinol Metab 1994; 78:818. 66. Schneider DL, Barrett-Connor EL, Morton DJ. Thyroid hormone use and bone mineral density in elderly women. JAMA 1994; 271:1245. 67. Franklyn J, Betteridge J, Holder R, Sheppard MC. Effect of estrogen replacement therapy upon bone mineral density in thyroxine treated post-menopausal women with a past history of thyrotoxicosis. Thyroid 1995; 5:359. 68. Oertel YC. Fine needle aspiration and the diagnosis of thyroid cancer. Endocrinol Metab Clin North Am 1996; 25:69. 69. Gharib H, Goellner JR. Fine needle aspiration biopsy of the thyroid: an appraisal. Ann Intern Med 1993; 118:282. 70. Gharib H, Goellner JR, Johnson DA. Fine-needle aspiration cytology of the thyroid. A 12-year experience with 11,000 biopsies. Clin Lab Med 1993; 13:699. 71. Hall TL, Layfield LJ, Philippe A, Rosenthal DL. Sources of diagnostic error in fine needle aspiration of the thyroid. Cancer 1989; 63:718. 72. Hamburger JI, Husain M, Nishiyama R, et al. Increasing the accuracy of fine-needle biopsy for thyroid nodules. Arch Pathol Lab Med 1989; 113:1035. 73. Caraway NP, Sneige N, Samaan NA. Diagnostic pitfalls in thyroid fine-needle aspiration: a review of 394 cases. Diagn Cytopathol 1993; 9:345. 74. Leenhardt L, Hejblum G, Franc B, et al. Indications and limits of ultrasound-guided cytology in the management of nonpalpable thyroid nodules. J Clin Endocrinol Metab 1999; 84:24. 75. Hatada T, Okada K, Ishii H, et al. Evaluation of ultrasound-guided fine-needle aspiration biopsy for thyroid nodules. Am J Surg 1998; 175:133. 76. Wartofsky L, Oertel Y. Fine needle aspiration biopsy of thyroid nodules. In: Van Nostrand D, ed. Nuclear medicine atlas. Philadelphia: JB Lippincott, 1988:193. 77. Tani EM, Skoog L, Löwhagen T. Clinical utility of fine-needle aspiration cytology of the thyroid. Annu Rev Med 1988; 39:255. 78. Bäckdahl M, Wallin G, Löwhagen T, et al. Fine-needle biopsy cytology and DNA analysis. Surg Clin North Am 1987; 67:197. 79. Pasielka JL, Zedenius J, Auer G, et al. Addition of nuclear DNA content to the AMES-risk group classification for papillary thyroid cancer. Surgery 1992; 112:1154. 80. Davila RM, Bedrossian CWM, Silverberg AB. Immunocytochemistry of the thyroid in surgical and cytologic specimens. Arch Pathol Lab Med 1988; 42:51. 81. Hennessey JV, Evaul JE, Tseng YL, et al. L-Thyroxine dosage: a reevaluation of therapy with contemporary preparations. Ann Intern Med 1986; 105:11. 82. Erdogan MF, Kamel N, Aras D, et al. Value of re-aspirations in benign nodular thyroid disease. Thyroid 1998; 8:1087. 83. DeGroot LJ. Clinical review 2: diagnostic approach and management of patients exposed to irradiation to the thyroid. J Clin Endocrinol Metab 1989; 69:925. 84. Brooks JR, Starnes HF, Brooks DC, Pelkey JN. Surgical therapy for thyroid carcinoma: a review of 1249 solitary thyroid nodules. Surgery 1988; 104:940. 85. Fogelfeld L, Wiviott MBT, Shore-Freedman E, et al. Recurrence of thyroid nodules after surgical removal in patients irradiated in childhood for benign conditions. N Engl J Med 1989; 320:835. 86. Anderson PE, Hurley PR, Rosswick P. Conservative treatment and long term prophylactic thyroxine in the prevention of recurrence of multinodular goiter. Surg Gynecol Obstet 1990; 171:309. 87. Hegedus L, Hansen JM, Veiergang D, Karstrup S. Does prophylactic thyroxine treatment after operation for non-toxic goitre influence thyroid size? BMJ 1987; 294:801. 88. Sosa JA, Bowman HM, Tielsch JM, et al. The importance of surgeon experience for clinical and economic outcomes from thyroidectomy. Ann Surg 1998; 228:320. 89. Lippi F, Ferrari C, Manetti L, et al. Treatment of solitary autonomous thyroid nodules by percutaneous ethanol injection: results of an Italian multicenter study. J Clin Endocrinol Metab 1996; 81:3261. 90. Cho YS, Lee HK, Ahn IM, et al. Sonographically guided ethanol sclerotherapy for benign thyroid cysts: results in 22 patients. AJR Am J Roentgenol 2000; 174:213. 91. Papini E, Pacella CM, Verde G. Percutaneous ethanol injection (PEI): what is its role in the treatment of benign thyroid nodules? Thyroid 1995; 5:147. 92. Caraccio N, Goletti O, Lippolis PV, et al. Is percutaneous ethanol injection a useful alternative for the treatment of the cold benign thyroid nodule? Five years experience. Thyroid 1997; 7:699. 93. Bennedbæk FN, Nielsen LK, Hegedus L. Effect of percutaneous ethanol injection therapy versus suppressive doses of L-thyroxine on benign solitary solid cold thyroid nodules: a randomized trial. J Clin Endocrinol Metab 1998; 83:830. 94. Goletti O, Monzani F, Lenziardi M, et al. Cold thyroid nodules: a new application of percutaneous ethanol injection treatment. J Clin Ultrasound 1994; 22:175. 95. Zingrillo M, Collura D, Ghiggi MR, et al. Treatment of large cold benign thyroid nodules not eligible for surgery with percutaneous ethanol injection. J Clin Endocrinol Metab 1998; 83:3905. 96. Treece GL, Georgitis WJ, Hofeldt FD. Resolution of recurrent thyroid cysts with tetracycline instillation. Arch Intern Med 1983; 140:2285. 97. Rosen IB, Provias JP, Walfish PG. Pathologic nature of cystic thyroid nodules selected for surgery by needle aspiration biopsy. Surgery 1986; 100:606. 98. Hamburger JI. The autonomously functioning thyroid nodule: Goetsch's disease. Endocr Rev 1987; 8:439. 99. Goldstein R, Hart IR. Follow-up of solitary autonomous thyroid nodules treated with 131 I. N Engl J Med 1983; 309:1473. 100. Estour B, Millot L, Vergely N, et al. Efficacy of low doses of radioiodine in the treatment of autonomous thyroid nodules: importance of dose/area ratio. Thyroid 1997; 7:357. 101. Ross DS, Ridgway EC, Daniels GH. Successful treatment of solitary toxic thyroid nodules with relatively low-dose iodine 131, with low prevalence of hypothyroidism. Ann Intern Med 1984; 101:488. 102. Tarantino L, Giorgio A, Mariniello N, et al. Percutaneous ethanol injection of large autonomous hyperfunctioning thyroid nodules. Radiology 2000; 214:143.

CHAPTER 40 THYROID CANCER Principles and Practice of Endocrinology and Metabolism

CHAPTER 40 THYROID CANCER ERNEST L. MAZZAFERRI Radiation and Thyroid Carcinoma Papillary Thyroid Carcinoma Prevalence Familial Papillary Carcinoma Pathology Diagnosis Factors Influencing Prognosis Staging Systems Consensus Views Concerning Therapy Surgery Radioiodine Therapy Thyroid Hormone Therapy Other Therapy Follow-Up Follicular Thyroid Carcinoma Prevalence Pathology Diagnosis Factors Influencing Prognosis Therapy Medullary Thyroid Carcinoma Multiple Endocrine Neoplasia Type 2 Syndromes Prevalence Pathology Diagnosis Factors Influencing Prognosis Therapy Follow-Up Family Screening Chapter References

Approximately 18,100 new cases of thyroid carcinoma were diagnosed in the United States in 1999, ranking it 22nd in incidence among the major malignancies.1 Its frequency varies with gender and age and is highest among women between the ages of 30 and 70 years. The lifetime risk of being diagnosed with thyroid cancer is ~0.64% for women and 0.25% for men. 2 In 1998, ~1200 of the 135,000 persons with thyroid carcinoma in the United States died of their disease.1 Between 1973 and 1992, its incidence among all ages rose steadily (almost 28%), whereas its mortality rates dropped more than 23%.2 The declining mortality rates are largely due to early diagnosis and effective therapy applied at an early tumor stage when it is most amenable to surgery and 131I therapy.3

RADIATION AND THYROID CARCINOMA Thyroid carcinoma may be caused by exposure to ionizing radiation. It was the first solid tumor found to have a significantly increased incidence in A-bomb survivors, but only among those younger than 20 years of age at the time of exposure.4 External Radiation. The risk of developing papillary thyroid cancer after therapeutic external radiation, used in the past to treat children with benign head and neck conditions, is well known.5 Exposure before the age of 15 years poses a major risk that becomes progressively greater with increasing amounts of radiation between 0.10 Gy (10 rad) and 10 Gy (1000 rad). This increases the incidence of thyroid carcinoma within 5 years of exposure, and this increased incidence continues for 30 years, at which time it begins to decline.5 Girls are only slightly more likely than boys to develop thyroid carcinoma after irradiation.5 Radioiodine-Induced Thyroid Carcinoma. Until recently, most studies have suggested that 131I is less effective than external gamma radiation in inducing thyroid cancer.4 A slightly elevated risk of thyroid cancer found in a large Swedish population exposed to diagnostic doses of 131I was attributed to cancers found among those examined for a suspected thyroid tumor.4 Another study from the United States reported a slight elevation of thyroid cancer mortality after treatment of hyperthyroidism with 131I, but the absolute risk was small, and the underlying thyroid disease appeared to play a role.6,6a However, most of these studies involved adults. When a large number of children developed thyroid carcinoma after being exposed to radioiodine fallout from the Chernobyl nuclear reactor accident in 1986, it became clear that 131I and other short-lived radioiodines were potent thyroid carcinogens in children, particularly those exposed when younger than age 10 years.7 Nuclear Weapon Fallout. Nuclear weapons that were tested above ground in Nevada between 1951 and 1963 resulted in radiation exposure to individuals across the continental United States. The chance of a significant exposure was highest in the 1950s for children who routinely drank milk from a backyard cow or goat that had ingested grass contaminated with radioactive fallout. The average cumulative thyroid dose from radioiodine fallout was 0.02 Gy (2 rad) for the American population collectively and 0.1 Gy (10 rad) for those younger than age 20 years at the time of exposure in the range known to cause thyroid carcinoma in children.5,8 Approximately 50,000 cases of thyroid cancer are likely to result from these exposures—nearly half of which have probably already appeared—but these are highly uncertain estimates.9 Nevertheless, a large study found an association between thyroid carcinoma and exposure to radioiodine fallout from nuclear tests in Nevada for children exposed at younger than 1 year of age and for those in the 1950 to 1959 birth cohort.10 Screening for thyroid disease that is due to fallout exposure is not recommended, but physicians should discuss the risk and palpate the neck of concerned patients.9

PAPILLARY THYROID CARCINOMA PREVALENCE Thyroid carcinoma is classified into four major types, which in decreasing order of frequency are papillary, follicular, medullary, and anaplastic carcinomas. Papillary and follicular carcinomas—often termed differentiated thyroid carcinomas—arise from follicular cells, synthesize thyroglobulin (Tg), and tend to be slow growing. Medullary carcinoma, which originates from thyroidal C cells that secrete calcitonin, may be sporadic or familial. Anaplastic carcinoma usually arises from well-differentiated thyroid tumors and is almost invariably fatal within a brief period (see Chap. 41). Papillary thyroid carcinoma accounts for ~80% of all thyroid cancers in the United States,11 including those induced by radiation.12 It is diagnosed in ~1 in 17,000 people in the United States annually, although occult microscopic papillary thyroidcarcinoma is found in 10% or more of autopsy and surgical thyroid specimens from men and women throughout adult life.13 In contrast, clinically manifest papillary thyroid carcinoma has a peak incidence in the fourth decade and is three times more frequent in women than in men. In children, the sex ratio is nearly equal, which may reflect radiation-induced disease. FAMILIAL PAPILLARY CARCINOMA Approximately 5% of papillary carcinomas are familial tumors, which are sometimes inherited as an autosomal dominant trait.14,15,15a Others are inherited as a component of familial adenomatous polyposis (Gardner syndrome), occurring at a young age as bilateral, multicentric tumors with an excellent prognosis, particularly those with ret-PTC activation.16,17 Cowden disease is an autosomal dominant syndrome characterized by multiple mucocutaneous hamartomas, keratoses, fibrocystic breast disease or breast cancer, and well-differentiated thyroid cancer.18,19 Carney complex—a syndrome with spotty skin pigmentation, myxomas, schwannomas, and multiple neoplasia that affects multiple glands—also includes thyroid carcinoma.20 PATHOLOGY

Gross Tumor. Papillary carcinoma usually is an unencapsulated and invasive tumor with ill-defined margins that is 2 to 3 cm when diagnosed, but this varies widely (Fig. 40-1A).3 Approximately 10% extend through the thyroid capsule into surrounding neck tissues, and another 10% are fully encapsulated.21 They typically are firm and solid, but some develop chronic hemorrhagic necrosis, yielding a thick brownish fluid on needle biopsy that may be mistaken for a benign cyst.22 Small tumors as large as 1.0 cm that often have a stellate appearance, termed microcarcinomas, usually are found by serendipity but rarely are invasive or metastasize (see Fig. 40-1B).23

FIGURE 40-1. Papillary thyroid carcinoma: gross pathology. A, Large papillary thyroid carcinoma completely replacing right thyroid lobe and extending beyond the thyroid capsule. Such lesions are associated with high recurrence and mortality rates. B, Occult papillary thyroid carcinoma found incidentally at surgery. This lesion almost invariably has a benign clinical course.

Architectural Features. Most papillary carcinomas contain complex branching papillae with a fibrovascular core covered by a single layer of tumor cells, intermingled with follicular structures (Fig. 40-2A), although some have a pure follicular or trabecular appearance.24 The term mixed papillary–follicular carcinoma has no clinical value because the follicular component does not alter 131I uptake or prognosis.25 Moreover, a tumor's nuclear features are more important than its architectural appearance in establishing a diagnosis of papillary carcinoma. Those with cellular features of papillary carcinoma but a pure follicular pattern are termed follicular variant papillary carcinoma (see Fig. 40-2D), which some argue comprise most tumors diagnosed as follicular carcinoma.26

FIGURE 40-2. Papillary thyroid carcinoma: histology. A, Microscopic papillary thyroid carcinoma (arrow) showing a mixed papillary and follicular architecture, and encapsulation. Isolated microscopic papillary carcinomas, found in ~10% of the general population, are incidental findings at surgery or autopsy and almost never are of clinical significance. ×20 B, Papillary thyroid carcinoma with typical papillae containing a fibrovascular core, and cells with large, pale-staining nuclei that appear crowded and overlapping. ×100 C, Fine-needle aspiration biopsy specimen of papillary thyroid carcinoma showing typical cytologic features of the tumor. ×400 D, High-power magnification of follicular variant papillary thyroid carcinoma showing characteristic cellular features, including nuclear inclusions, irregularly placed, pale-staining nuclei, and abundant cytoplasm. ×400

Cytology Features. Papillary carcinoma has cellular features that distinguish it from other thyroid neoplasms regardless of its architecture, permitting an accurate diagnosis by fine-needle aspiration (FNA) (see Fig. 40-2C). The cells are large and contain pink to amphophilic finely granular cytoplasm and large pale nuclei (“Orphan Annie eye” nuclei) with inclusion bodies and nuclear grooves that identify the tumor as papillary carcinoma (see Fig. 40-2D). Psammoma bodies (Fig. 40-3)—the “ghosts” of infarcted papillae that are virtually pathognomonic of papillary carcinoma—are calcified, concentric lamellated spheres found in approximately half the cases within or near the tumor and lymph nodes or in cytology specimens.27 Lymphocytic infiltration—ranging from focal areas of lymphocytes and plasma cells to classic Hashimoto disease (see Chap. 46)—is often seen in papillary carcinoma. Papillary carcinomas are commonly found in multiple sites within the thyroid gland, which are generally thought to be intraglandular metastases. Multiple microscopic tumor foci are found in as many as 80% of cases when the gland is examined in considerable detail, but in routine clinical practice their frequency is as high as 45%, and they usually are bilateral (Fig. 40-4).11

FIGURE 40-3. Psammoma bodies of papillary carcinoma (arrows) are calcified, dark-staining, lamellated spheres that are virtually pathognomonic of papillary thyroid carcinoma. ×400

FIGURE 40-4. Microscopically multifocal papillary thyroid carcinoma (arrows) usually represents intraglandular metastases. ×20

Lymph Node Metastases. Papillary carcinoma commonly metastasizes to lymph nodes in the lateral neck, central neck compartment, and mediastinum. Gross lymph node metastases are present in approximately half the cases at the time of diagnosis, while even more—as many as 85% in studies from Japan28,29— have microscopic nodal metastases.25 Their number rises as primary tumor size increases. Nodal metastases tend to be bilateral when the isthmus or both thyroid lobes are involved with tumor, or they may extend into the mediastinum or extend beyond the lymph node capsule into soft tissues, which are all poor prognostic signs.25,30 Distant Metastases. Fewer than 5% of patients have distant metastases at the time of diagnosis, and another 5% develop them later.11 In a review of 1231 patients with distant metastases, 49% were in the lung, 25% were in bone, 15% were in lung and bone, and 12% were in the central nervous system or in multiple organs.11 Pulmonary metastases may be large discrete nodules or may have a “snowflake” appearance caused by diffuse lymphangitic tumor spread (Fig. 40-5A, Fig. 40-5B and Fig. 40-5C). Pulmonary metastases not seen on radiographs may be detected only with whole body scanning done after therapeutic doses of 131I have been administered.31,32,33 and 34

FIGURE 40-5. Distant metastases of papillary thyroid carcinoma. A, Papillary thyroid carcinoma with diffuse lymphangitic spread of tumor giving a typical “snowflake” appearance on the chest radiograph. Such tumors typically concentrate 131I in younger patients and tend to have a good prognosis. B, Papillary thyroid carcinoma with diffuse nodular infiltrates that concentrated 131I poorly. This patient had pulmonary metastases at the time of initial diagnosis. C, Papillary thyroid carcinoma (left) in a young woman that did not concentrate 131I and grew steadily over a 4-year period (right).

Thyroglossal Duct. Papillary thyroid carcinoma arising within a thyroglossal duct is almost always small and usually has a benign course.35 Papillary Microcarcinoma. Different histologic variants or subtypes of papillary carcinoma have distinct biologic behaviors (Table 40-1). Five histologic subtypes or variants were recognized in the 1988 World Health Organization classification,24 and at least five others have been described since then.36,37 Papillary microcarcinoma is a tumor 1.0 cm or smaller. Approximately 20% are multifocal, and as many as 60% have cervical lymph node metastases,29 which may be their presenting feature.23 Lung metastases occur rarely with multifocal tumors that have bulky cervical metastases; these are the only micro-carcinomas with significant morbidity and mortality.23,37 Otherwise, the recurrence and cancer-specific mortality rates are near zero.13,23

TABLE 40-1. Tumor Histologic Variants That Influence Prognosis in Papillary and Follicular Thyroid Carcinoma

Encapsulated Papillary Carcinoma. Encapsulated papillary carcinoma is an otherwise typical tumor completely surrounded by a fibrous capsule. It accounts for ~10% of papillary carcinomas and is approximately half as likely to metastasize as is typical papillary carcinoma.37 Tumor recurrence is lower than usual, and death that is due to cancer almost never occurs with these tumors.27,37 Follicular Variant Papillary Carcinoma. Follicular variant papillary carcinoma accounts for ~10% of papillary carcinomas.37 Usually not encapsulated, its microfollicular architectural pattern is otherwise indistinguishable from follicular carcinoma, but the typical nuclear features of papillary carcinoma identify its true nature, which can be diagnosed by FNA cytology.27 Its metastases often have psammoma bodies and may show the typical features of papillary carcinoma. Some claim that its prognosis is similar to the usual papillary carcinoma, and others suggest that distant metastases are more likely and that long-term outcome may be unfavorable with this tumor.36,37 and 38 Diffuse Macrofollicular Variant Papillary Carcinoma. Diffuse macrofollicular variant papillary carcinoma is an uncommon tumor that can be confused with goiter or macrofollicular adenoma on frozen section.36 It occurs predominantly in women with goiter, approximately one-third of whom have hyperthyroidism. Most have distant metastases with very high mortality rates.36 Tall Cell Variant. Approximately 10% of papillary thyroid carcinomas show extensive papillae formation with cells twice as tall as they are wide that comprise at least 30% of the tumor.36 Compared with typical papillary carcinoma, tall cell tumors tend to be diagnosed approximately two decades later (in the mid-50s), are larger tumors associated with significantly more invasion into extrathyroidal soft tissues, and have more distant metastases.18,36,37 The tumor, which can be identified on a cytology specimen from fine-needle biopsy, often expresses the p53 suppressor oncogene, perhaps accounting for its frequent loss of 131I uptake and long-term mortality rates that are two- to threefold those of typical papillary carcinomas.18,37 Columnar Cell Variant. Columnar cell variant is a rare variant possibly related to tall cell carcinoma. It occurs mainly in men and is composed of rectangular cells with clear cytoplasm.36 More than 90% develop distant metastases, which usually are unresponsive to 131I therapy or chemotherapy and result in death.36,37 When encapsulated, it has a much better prognosis.21 Diffuse Sclerosis Variant. Approximately 5% of papillary carcinomas are of the diffuse sclerosis variant type, which is even more common in children and adolescents.36,37 Approximately 10% of the tumors in the children of Chernobyl are of this type.39 Usually involving both lobes, diffuse sclerosis variant presents as a goiter with extensive squamous metaplasia, sclerosis, abundant psammoma bodies, and lymphatic invasion involving the whole thyroid gland. Lymph node metastases are almost always present, and ~25% have lung metastases.36,37 Cytology specimens reveal squamous metaplasia, inflammatory cells, and psammoma bodies, but

may be difficult to differentiate from thyroiditis. Although it has a higher incidence of local and pulmonary metastases than typical papillary carcinomas, there is some disagreement about whether its long-term prognosis is worse than usual.36,37 Oxyphilic (Hürthle Cell) Variant. Approximately 2% of papillary carcinomas have nuclei resembling those of Hürthle cell follicular carcinomas.40 Multiple oxyphilic thyroid tumors and a familial occurrence have been noted in some cases.41 This tumor cannot be identified by FNA cytology but is recognized by its papillary architecture on the final histologic sections. Compared with typical papillary carcinomas, oxyphilic carcinomas have fewer neck nodal metastases at diagnosis but have higher rates of recurrence and cause-specific mortality; in this respect, they resemble oxyphilic follicular carcinoma.37,40 Solid or Trabecular Variant. A tumor with a predominantly (>75%) solid architectural pattern that maintains the typical nuclear features of papillary carcinoma, the solid or trabecular variant has a propensity for extrathyroidal spread and lung metastases.36,37 Some, however, find it to be more common in children and report that its prognosis is the same as with typical papillary carcinoma.42 Insular Carcinoma. Approximately 5% of all thyroid carcinomas show solid clusters of cells with small follicles that contain Tg but resemble pancreatic islet cells. This is often categorized as a variant of follicular carcinoma, but it may show papillary differentiation, and it has been suggested that it should be considered a separate entity derived from follicular epithelium.43 Insular carcinomas usually are large and highly invasive tumors that grow through the tumor capsule and into tumor blood vessels. Compared with differentiated thyroid carcinoma, insular carcinoma presents at an older age (54 vs. 36 years); with larger tumors (4.7 vs. 2.5 cm); with fewer neck metastases (36% vs. 50%) but more distant metastases (26% vs. 2%); and has a worse 30-year cancer-specific mortality rate (25% vs. 8%). 18 Insular carcinoma also displays aggressive behavior in children.44 DIAGNOSIS History. In the past, many papillary carcinomas were large and invasive when first diagnosed; however, with the use of FNA, most are small tumors found at an early stage as an asymptomatic thyroid nodule on routine examination or in screening programs of patients with head and neck irradiation.3 The timeliness of diagnosis affects the prognosis (see Prognosis).3 Occasionally, it causes pain, hoarseness, dysphagia, or hemoptysis, or it infiltrates surrounding structures or grows rapidly—findings that are associated with high mortality rates. However, only a few patients have such symptoms. A history of irradiation is important, but only 30% of patients develop palpable thyroid nodules after irradiation, and of these, only one-third are malignant.45 Thyroid nodules that appear after radiation should undergo an evaluation similar to that used for nodules that occur spontaneously. Physical Examination. Papillary carcinoma usually is manifest as a palpably discrete thyroid nodule that moves upward when the patient swallows, but a cervical lymph node metastasis may be its only sign. A midline mass above the thyroid isthmus may be a metastatic lymph (Delphian) node or may be carcinoma within a thyroglossal duct that can be identified from its upward movement when the tongue is protruded. Papillary carcinoma may be a firm, nontender, discrete mass, but many are soft and cystic or diffusely infiltrate one lobe or the entire thyroid. Distant metastases are found less often at the time of diagnosis of papillary carcinoma than follicular carcinoma, but when they are found, the primary tumor is almost invariably large (see Fig. 40-1A and Fig. 40–5B). Fine-Needle Aspiration. Approximately 10% of thyroid nodules show clear evidence of malignancy, such as vocal cord paralysis, signs of invasion, or bulky lymph node metastases, whereas the others appear benign on examination. Neither the history nor the physical examination offers sufficient evidence of a nodule's benign nature that further testing can be deferred (see Chap. 34, Chap. 35 and Chap. 39). The first test in a clinically euthyroid patient is FNA.46 Other tests are too nonspecific to be used first. Except for hyperfunctional (“hot”) nodules that are rarely malignant, most thyroid nodules—whether single in an otherwise normal gland or a dominant nodule in a multinodular goiter—require FNA for diagnosis. Although large-needle cutting biopsies can be done, FNA is considerably safer and is highly effective in obtaining sufficient cytology to identify the distinctive features of papillary carcinoma, including most of its variants46 (see Chap. 39). Nodules that yield diagnostic or suspicious cytology are excised; the others are scanned to exclude hot nodules, whereas benign nodules are simply observed or sometimes treated with thyroxine-suppressive therapy. Thyroxine suppression should not be done without prior FNA and should not be used as a diagnostic test, because thyroid carcinomas may appear to shrink.46 Prior Neck Irradiation. FNA should be the first test performed on a palpable thyroid nodule in a euthyroid patient with a history of head-and-neck irradiation. An asymptomatic patient with a palpably normal thyroid gland who has history of head and neck irradiation should not undergo ultrasonography because nearly 90% of such patients have thyroid nodules found by this test, most of which are benign.9,47 Ultrasonography reveals thyroid lesions 30,000 rad does not increase the success rate.101,129 Response rates are significantly lower when patients have less than a total or near-total thyroidectomy or have a thyroid remnant calculated to be >2 g.101 Therapeutic 131I for Residual or Recurrent Carcinoma. Residual or recurrent carcinoma should be treated surgically whenever possible; however, only ~50% to 75% of differentiated thyroid carcinomas and their metastases and approximately one-third of Hürthle cell carcinomas concentrate 131I.131,132 and 133 One study found that two-thirds of 283 patients with lung or bone metastases had tumors that concentrated 131I.134 This is crucial to survival. For example, 10-year survival rates in one study were 83% or 0%, respectively, depending on whether pulmonary metastases did or did not concentrate 131I.75 The lung metastases that concentrate 131I best are the smallest lesions found in young patients.75 Empiric Fixed Doses. There are basically three approaches to therapy: empiric fixed doses, upper bound limits set by blood dosimetry, and quantitative dosimetry.130 With the first, a fixed amount of 131I is given based on what is being treated. For example, 30 to 50 mCi are given to ablate thyroid remnants, 150 to 175 mCi for residual carcinoma in cervical nodes or neck tissues, and 200 mCi or more for distant metastases. Tumor 131I uptake in amounts adequate for imaging with 4-mCi scanning doses is usually sufficient for 131I therapy, using empiric doses from 30 to 200 mCi.135 Upper Bound Limits Set by Blood Dosimetry. Blood dosimetry is done to establish an upper limit on the amount of 131I that can be given safely, which is generally considered to be 200 rad to the whole blood from a single dose.136 Quantitative Tumor Dosimetry. The third approach is to calculate the dose of 131I that is required to deliver 30,000 rad to ablate the thyroid remnant or 8000 to 12,000 rad to treat nodal or discrete soft tissue metastases. For pulmonary metastases, the amount of 131I is administered that delivers 200 rad to whole blood with no more than 80 mCi of whole blood retention at 48 hours.130 The two most important factors in determining success are the mass of residual tissue and the effective half-time of 131I in that tissue.130 An 80% response was found in tumor deposits that received at least 8000 rad.101 Lesions that receive 1.5 cm) should be evaluated every 6 to 12 months for 10 years. Whole-body 131I scan performed 12 months after surgery and 131I ablation often can document complete absence of tumor. Thereafter, scanning can be done at infrequent intervals unless there is a change in the examination or a rise in Tg.

FIGURE 40-7. Use of recombinant human thyroid-stimulating hormone (TSH) in the management of differentiated thyroid carcinoma. Top shows method of recombinant human TSH (rhTSH) administration, and bottom shows algorithm for its use. (CT, computed tomography; IM, intra-muscularly; MRI, magnetic resonance imaging; Tg, thyroglobulin.)

RECOMBINANT HUMAN THYROID-STIMULATING HORMONE Periodic withdrawal of thyroid hormone therapy is required during follow-up. This causes symptomatic hypothyroidism to raise the serum TSH concentrations sufficiently to stimulate thyroid tissue so that 131I scanning and serum Tg measurements can be obtained. Intramuscular administration of rhTSH stimulates thyroidal 131I uptake and Tg release while the patient continues thyroid hormone suppression therapy, thus avoiding symptomatic hypothyroidism.160,161 Now approved for diagnostic use, rhTSH has been tested in two large international multicenter studies. The first study found that whole-body 131I scan results done after two 0.9-mg doses of rhTSH (while thyroid hormone therapy was continued) were of good quality; they were equivalent to the scans obtained after thyroid hormone withdrawal in 66% of the patients, superior in 5%, and inferior in 29%.161 This study found that rhTSH stimulates 131I uptake for whole-body scanning, but the sensitivity of scanning after rhTSH administration was less than after the withdrawal of thyroid hormone.161 Scanning with rhTSH was associated with significantly fewer symptoms and dysphoric mood states. A second multicenter international study was done to test two dosing schedules of rhTSH on the results of whole-body scans and serum Tg levels compared with those obtained after thyroid hormone withdrawal. The whole-body 131I scanning method was more carefully standardized in the second study than in the first.162 The scans in this study were concordant in 89% of the patients, with superior whole-body scans seen in 4% of the subjects after rhTSH and in 8% after thyroid hormone withdrawal, differences that were not statistically significant. The combination of whole-body scanning and serum Tg measurements detected 93% of the patients with disease or tissue limited to the thyroid bed and detected 100% of the patients with metastatic carcinoma.162 Although not yet approved for preparation of patients for 131I therapy, rhTSH has been used successfully for this purpose.163 Recombinant human TSH, 0.9 mg, is given intramuscularly every day for 2 days, followed by a minimum of 4 mCi of 131I on the third day and a whole-body scan and Tg measurements on the fifth day (see Fig. 40-7). Whole-body 131I images are acquired after 30 minutes of scanning or after obtaining 140,000 counts. A serum Tg of ³2.0 ng/mL obtained 72 hours after the last rhTSH injection indicates that thyroid tissue or thyroid carcinoma is present, which almost always can be identified on the rhTSH-stimulated whole-body scan using the indicated scanning method.162 The drug is well tolerated, with mild headache and nausea being its main adverse effects. Serum Thyroglobulin Measurement. Serum Tg determinations and whole-body 131I imaging together detect recurrent or residual disease in most patients who have undergone total thyroid ablation. The serum Tg concentration reflects the mass of normal thyroid tissue or differentiated thyroid carcinoma, the degree of thyroid physical damage or inflammation, and the level of TSH receptor stimulation.164 Serum anti-Tg antibodies should be measured in the sample obtained for serum Tg assay because these antibodies, which are found in as many as 25% of patients with thyroid carcinoma, invalidate serum Tg measurements in most assays.124,165 Tg measurement is more sensitive when T4 has been stopped or rhTSH is given to elevate the serum TSH.31,162 Under these circumstances, serum Tg has a lower false-negative rate than whole-body 131I scanning.31,162 Detecting serum Tg by a newly introduced Tg mRNA method is a more sensitive marker of residual thyroid tissue or cancer than measuring Tg by immunometric assay, particularly when Tg mRNA is detected during T4 treatment or with circulating anti-Tg antibodies.166 The author uses a sensitive Tg immunometric assay with a detection limit of 0.5 ng/mL. A serum Tg above this level during T4 therapy after total or near-total thyroidectomy and 131I ablation has been achieved is a sign of persistent normal tissue or differentiated thyroid carcinoma, which is an indication for repeat scanning when there is no other evidence of disease (see Fig. 40-7). If serum Tg rises above 10 ng/mL after T 4 is discontinued or rises above 2.5 ng/mL after rhTSH is administered, normal or malignant thyroid tissue is usually present, even if the 2- to 4-mCi 131I diagnostic scan is negative (i.e., 10 ng/mL, the author gives a therapeutic dose of 131I, usually 100 to 150 mCi, and perform a posttreatment scan. In the author's experience, ~20% of such patients have lung metastases. Others use different cutoff values and different doses of 131I, but the Tg level to trigger treatment has gradually come down.34 Scanning Patients with High Serum Thyroglobulin Concentrations and Negative 131I Scans. Several radionuclide scanning techniques may be used to identify the location of tumor in patients with high serum Tg levels, negative diagnostic 131I scans, and negative neck ultrasonography. Thallium-201. Thallium-201 (201Tl) scintigraphy may identify metastases or uptake in the neck when the serum Tg is elevated but the diagnostic whole-body 131I scan is negative.167 In a study comparing 201Tl and 131I scans, the sensitivities and specificities, respectively, were 94% and 96% for 201Tl, and 29% and 100% for 131I.168 However, another study found that 131I is much more sensitive than 201Tl in demonstrating residual thyroid tissue after surgery (100% and 33%, respectively).169 Others also find that 131I scintigraphy is more sensitive and more specific than 201Tl scintigraphy in identifying both residual neck uptake and metastases.170 In a study of 36 paired whole-body 131I and 201Tl scan results, residual uptake in the neck was seen on all the 131I scans but on only 17% of the 201Tl scans.170 In this study, multiple metastatic lesions identified on 14 131I scans were interpreted as negative or nonspecific, or as showing fewer lesions on the corresponding 201Tl scans. Sixteen 201Tl scans gave false-positive results. The authors concluded that 131I scintigraphy is more sensitive and more specific than 201Tl scintigraphy for detecting distant metastases and for identifying residual activity in the neck after thyroidectomy.170 Other studies have found the two scanning techniques approximately equally useful in detecting local recurrences or distant metastases.171 Technetium-99m. Technetium-99m (99m Tc) may localize differentiated thyroid carcinoma. One study found that the sensitivities of 201Tl, 99m Tc-tetrofosmin, and 131I in identifying distant metastases were comparable (85%, 85%, and 78%, respectively).169 However, 131I was much more sensitive than 99mTc-tetrofosmin for demonstrating remnant thyroid tissue after surgery (100% and 33%, respectively).169 Scanning with 99m Tc-methoxyisobutyl isonitrile (99m Tc-MIBI) also detects

metastases of thyroid carcinoma and may be useful in the postoperative follow-up. One large study found increased accumulation of 99m Tc-MIBI in 75% of the patients with lung metastases, in 100% of those with lymph node metastases, and in 94% of the patients with bone metastases.172 The 99m Tc-MIBI scans identified more lesions in the lung than did 201Tl or 131I scans, which, respectively, were positive in 80% and 85% of patients with lung metastases, in 100% and 42% of patients with lymph node metastases, and in 90% and 87% of patients with bone metastases.172 Whole-Body Positron Emission Tomography. Whole-body positron emission tomography (PET) with 18F-fluorodeoxyglucose (FDG) may identify differentiated thyroid carcinoma metastasis that cannot be identified by scintigraphy with 131I or 99m Tc. Although PET has better sensitivity, resolution imaging, and spatial localization, this has to be balanced with its higher cost when compared with thallium scintigraphy.173 False-positive 18F-fluorodeoxyglucose uptake may occur with benign lung disease.174 FDG-PET scans are useful in predicting survival in differentiated thyroid cancer. In one study, multivariate analysis demonstrated that the single strongest predictor of survival was the volume of disease that displayed avidity for FDG-PET. The probability of surviving 3 years with FDG volumes of 125 mL or less was 96% compared with 18% in those with a FDG volume >125 mL. All of the 10 patients with distant metastases and negative PET scans were alive and well at the end of the study, whereas those with positive PET scans were more likely to die of disease.174a False-Positive 131I Scans. Body secretions, transudates, inflammation, nonspecific mediastinal uptake, and neoplasms of nonthyroidal origin may concentrate 131I.175 This also can be seen with physiologic secretion of 131I from the nasopharynx, salivary and sweat glands, stomach, genitourinary tract, and from skin contamination with sputum or tears.176 Pathologic pulmonary transudates and inflammation that is due to cysts, as well as lung lesions caused by fungal and other inflammatory disease, may produce false-positive scans. Diffuse hepatic uptake of 131I is rarely due to occult liver metastases but usually is from hepatic clearance of Tg labeled with 131 I by functioning thyroid remnants or extrahepatic thyroid cancer metastases. The more 131I uptake that appears in the residual thyroid tissue, the more it appears in the liver. In one large study,177 diffuse hepatic 131I uptake was seen in 60% of 399 patients undergoing 131I scans and in nearly 36% of 1115 131I scintigraphy studies. Twelve percent of the diagnostic scans in this study showed uptake in the liver. The frequency of hepatic uptake in the posttherapy scans was related to the dose of 131I as follows: 39% with 30 mCi; 61.5% with 75 to 100 mCi; and 71.3% with 150 to 200 mCi.177 Patients whose 131I scans show hepatic uptake without uptake in the thyroid bed or in extrahepatic metastases, however, often have occult liver metastases.177 I Treatment of Patients with Negative 131I Scans and High Serum Thyroglobulin. When the serum Tg level is elevated and a tumor cannot be found by localizing techniques, pulmonary metastases are sometimes found only after administrating therapeutic doses of 131I.34,178 The Tg cutoff level has been gradually coming down for treating patients with large doses of 131I whose only evidence of disease is an elevated serum Tg. It was ~30 or 40 ng/mL 10 years ago, but now is closer to 10 ng/mL when the patient is off thyroid hormone.34 There is increasing evidence that this treatment is beneficial. Multivariate analyses have shown the prognostic importance of the size of pulmonary metastases at the time of discovery.179 Another multivariate analysis of prognostic factors in 134 patients with pulmonary metastases showed that early (normal chest roentgenogram) scintigraphic diagnosis and 131I therapy for lung metastases were the most important elements in obtaining both a significant improvement in survival rate and a prolonged disease-free time interval.76 131

Two studies32,33 found beneficial effects of 131I therapy for patients with high serum Tg levels, a normal chest roentgenogram, and a negative diagnostic 131I scan. Of the patients reported in these two series, 80% achieved a negative 131I on posttherapy scan, 60% had a serum Tg 4 cm are associated with a much higher mortality rate.181

FIGURE 40-9. A, Follicular thyroid carcinoma with a large isolated nodular pulmonary metastasis. B, Hürthle cell carcinoma with multiple nodular pulmonary and bone metastases that did not concentrate 131I.

FIGURE 40-10. Follicular carcinoma metastatic to pelvis. A, Radiograph showing large osteolytic lesions of pubic bones. B, Follicular carcinoma, 131I uptake in lesions shown in A. Note that 131I uptake is more extensive than might be predicted from the radiograph.

DIAGNOSIS Follicular carcinoma may present as an asymptomatic neck mass, usually without palpable cervical lymph nodes, or a distant metastasis may be the first sign of tumor. Metastases may appear as large discrete pulmonary nodules, osteolytic bone lesions causing pathologic fractures, or as central nervous system tumors with serious neurologic sequelae, which are rarely seen in the absence of a palpable thyroid lesion.11 Bulky metastatic lesions may be functional and can cause thyrotoxicosis, which can be triiodothyronine toxicosis (see Chap. 42). The diagnostic evaluation is similar to that for papillary thyroid carcinoma, except a problem is usually encountered in differentiating benign Hürthle cell tumors and follicular adenomas from their malignant counterparts by FNA or by study of the frozen sections at surgery.95a Typically, such tumors are simply designated as follicular neoplasms, because their benign or malignant character cannot be determined. Large-needle aspiration biopsies and cutting-needle biopsies usually yield results similar to those of FNA but have more serious complications.46 Some physicians suggest that large-needle biopsy, if it can be done efficiently and safely, may be diagnostically useful for follicular nodules ³3 cm that demonstrate suspicious cytology on FNA. However, most recommend surgical excision of thyroid nodules that are suspicious on FNA46,95a (see Chap. 39). FACTORS INFLUENCING PROGNOSIS The 10-year mortality from follicular carcinoma is ~10%182 but ranges to 50%, depending on the degree of vascular invasion and capsular penetration by the tumor.11,182,183 and 184 Hürthle cell carcinomas have the worst prognosis.183 Tumor recurrence in distant sites is seen more often with follicular than with papillary carcinomas and occurs most frequently with highly invasive tumors, when the primary lesions are >4 cm, and in Hürthle cell carcinomas.11,180,181,183 Marked cellular atypia or frank ana-plastic transformation is also associated with a poor prognosis. Women with this tumor may fare better than men do, but age has the most important influence on prognosis. Patients younger than 40 years of age at the time of diagnosis have the best prognosis, whereas older patients tend to have higher recurrence and mortality rates from tumors that often do not concentrate 131I.11 Ten- and 15-year survival rates with distant metastases (~30% and 10%, respectively) are lowest among patients older than 40 years of age.11,181 THERAPY Appropriate treatment can improve survival with this tumor. Relapse, which occurs less often with surgery followed by 131I therapy,11,180 is seen least often in distant sites after total thyroidectomy and 131I therapy.65 Survival rates are best if the tumor initially concentrates 131I and can be completely ablated with the initial surgery and 131 I therapy.11 The argument for total or near-total thyroidectomy for follicular carcinoma is stronger for tumors that are more highly invasive, for Hürthle cell carcinomas, and for primary tumors >4 cm in diameter. 11,181,184 Less extensive thyroid surgery may be adequate for tumors that are minimally invasive, because such tumors ordinarily have an excellent prognosis95a,185; however, because this distinction cannot be made preoperatively, most clinicians prefer total thyroidectomy for follicular carcinoma. When the diagnosis of follicular carcinoma is not initially appreciated (frozen-section histologic analysis is notoriously inaccurate with this tumor), only lobectomy and isthmusectomy are usually done. If the permanent histologic sections show the tumor to be malignant, then complete thyroidectomy should be done within 2 to 3 days.11 In most clinics, patients with moderately to extensively invasive follicular carcinomas are routinely given 131I postoperatively to ablate residual 131I uptake in the thyroid bed. When the tumor is metastatic, the principal treatment is 131I, as described for papillary carcinoma, provided the tumor concentrates the isotope. If it does not, treatment with external radiation or doxorubicin may be effective.95a Thyroid hormone suppression is always given unless the patient is thyrotoxic from metastatic tumor tissue, which produces thyroid hormone.

MEDULLARY THYROID CARCINOMA Medullary thyroid carcinoma (MTC), first recognized in 1959, is a pleomorphic neoplasm with amyloid struma that arises from the calcitonin-secreting C cells of the thyroid. Approximately 20% are familial tumors that are transmitted as an autosomal dominant trait and are often associated with other endocrine neoplasms. The genes responsible for the familial forms of MTC map to the pericentromeric region of chromosome 10.186 MULTIPLE ENDOCRINE NEOPLASIA TYPE 2 SYNDROMES MTC may occur in four settings. Sporadic MTC is a unilateral thyroid tumor with no somatic lesions other than the neoplasm. Multiple endocrine neoplasia type 2A (MEN2A) is familial, bilateral MTC with hyperparathyroidism and bilateral pheochromocytoma. Multiple endocrine neoplasia type 2B (MEN2B) may be sporadic or familial and features bilateral MTC, bilateral pheochromocytoma, an abnormal phenotype with multiple mucosal ganglioneuromas, and musculoskeletal abnormalities

suggestive of the marfanoid habitus. Familial non-MEN MTC (FMTC) comprises bilateral MTC with no other endocrine tumors or somatic abnormalities.187 MEN2A is always inherited as an autosomal dominant trait, whereas MEN2B may be similarly transmitted or can occur sporadically. FMTC is an autosomal dominant trait that is the least common form, and manifests at a later age.187 Medullary Thyroid Carcinoma. The MTC in both MEN2 syndromes is generally bilateral and multicentric, as opposed to sporadic MTC, which is usually unilateral. The C cells are normally located in the upper and middle thirds of the lateral thyroid lobes. In MEN2, the C cells initially undergo hyperplasia, which precedes the development of MTC. Parathyroid Disease. One-third to one-half of the patients with the MEN2A syndrome have hyperparathyroidism, most of whom (85%) have parathyroid hyperplasia.188 In contrast, almost no patients with MEN2B have parathyroid disease.189 Patients with MEN2A seldom present with symptoms of hypercalcemia but often form kidney stones. Parathyroid hyperplasia is discovered during MTC surgery in most MEN2A patients, even those without clinical or biochemical evidence of hyperparathyroidism.189 Pheochromocytoma. Adrenal medullary disease, ranging from diffuse or nodular hyperplasia to large bilateral multilobular pheochromocytomas, occurs in both MEN2 syndromes. Pheochromocytomas or medullary hyperplasia is typically bilateral and occurs in ~40% of patients, with a range of 6% to 100% in different kindreds.188 The symptoms are typically subtler than those encountered with sporadic pheochromocytoma.190 The diagnosis is usually established by demonstrating high urinary or plasma catecholamine levels, but the total urinary catecholamines may be normal, with only an increased epinephrine:norepinephrine ratio, particularly in those with adrenal medullary hyperplasia190 (see Chap. 86). Multiple Endocrine Neoplasia Type 2B. MEN2B is characterized by a constellation of somatic abnormalities consisting of ganglioneuromas of the tarsal plates and the anterior third of the tongue, and a marfanoid habitus. Nodules may occur in the lips, causing them to appear lumpy and patulous. Ganglio-neuromas in the alimentary tract may be associated with constipation, diarrhea, and megacolon. The marfanoid characteristics include long limbs, hyperextensible joints, scoliosis, and anterior chest deformities, but not ectopic lens or cardiovascular abnormalities189 (see Chap. 188). Point mutations of the RET protooncogene have been identified in germline and tumor DNA of unrelated patients from kindreds with MEN2A, MEN2B, and FMTC.186,191,192 A relationship exists between specific RET protooncogene mutations and the disease phenotype in MEN2.192 Although several different and independent point mutations in the genomic sequence of the RET protooncogene have been identified, all involving codons for cysteine residues, the normal function of RET is not yet known, and its role in the development of these inherited neoplasms remains unclear. Nevertheless, the identification of point mutations provides an important direct means of identifying affected MEN and FMTC kindreds. PREVALENCE MTC accounts for ~10% of all thyroid malignancies. Approximately 80% to 90% of MTC cases occur sporadically, and 10% to 20% are inherited. MTC may occur at any age, but sporadic disease is diagnosed later in life than is familial MTC. The median age of patients seen at the Mayo Clinic with sporadic MTC was 51 years, compared with 21 years for those with familial tumors.190 MEN2A and FMTC are both characterized by the development of bilateral MTC, but the age at onset of FMTC is usually later, ranging from 40 to 50 years, as compared with an average of 20 to 30 years for MEN2A. Familial MTC occurs with equal frequency in both sexes, while sporadic MTC has a female-male ratio of 1.5:1. PATHOLOGY Sporadic and familial MTC are histologically similar, although a wider spectrum of appearances is encountered in familial tumors, ranging from isolated hypertrophied C cells to large bilateral multicentric tumors that are usually in the superior portions of the thyroid lobes. Typically, the tumor is composed of fusiform or polygonal cells surrounded by irregular masses of amyloid and abundant collagen (Fig. 40-11). Calcifications are present in approximately one-half of the tumors, and occasionally trabecular bone formation is seen. Calcitonin can usually be demonstrated in the tumor by immunohistochemical studies.

FIGURE 40-11. Medullary thyroid carcinoma showing trabecular architecture with spindle and round cells and abundant stroma that stained positively with Congo red, for amyloid. ×100

Cervical node metastases occur early in the disease and can be seen with primary lesions as small as several millimeters. Tumors >1.5 cm in diameter are more likely to metastasize to distant sites. The pattern of tissue calcitonin staining may differentiate virulent from less aggressive tumors. In one study, patients with primary tumors that showed intense homogeneous calcitonin staining were all clinically well on follow-up examination, whereas patients whose tumors showed patchy localization of calcitonin either developed metastatic disease or died of cancer within 6 months to 5 years of initial surgery.193 DIAGNOSIS Clinical Features. Patients with sporadic disease or previously unrecognized familial MTC usually present with one or more painless thyroid nodules in an otherwise normal gland, but the tumor may cause pain, dysphagia, and hoarseness. The dominant sign may be enlarged cervical lymph nodes or, occasionally, distant metastases, most commonly to the lung, followed in frequency by metastases to the liver, bone (osteolytic or osteoblastic lesions), and brain. In approximately one-half of patients with sporadic MTC, cervical lymph node metastases are present at the time of diagnosis. The thyroid nodule may be cold or normal on radionuclide imaging and is usually solid on echography and malignant on FNA. Radiographs may show dense, irregular calcifications of the primary tumor and cervical nodes, and mediastinal widening that is due to metastases. Rarely, an abnormal phenotype may correctly identify the tumor, or a paraneoplastic syndrome may be the presenting manifestation. Hormonal Features. In addition to calcitonin, MTC may synthesize calcitonin gene-related peptide, L -dopa decarboxylase, serotonin, prostaglandins, adrenocorticotropin, histaminase, carcinoembryonic antigen, nerve growth factor, and substance P.190 Elevated calcitonin, histaminase, L -dopa decar-boxylase, and carcinoembryonic antigen serum levels occur frequently in MTC patients, but not in other thyroid cancers. Approximately 10% of MTC patients have episodes of flushing often induced by alcohol ingestion, calcium infusion, and pentagastrin injection; these episodes may be due to tumor release of prostaglandins and serotonin.190 The only recognized clinical manifestation of high circulating calcitonin levels is a secretory diarrhea that occurs in ~30% of patients and is usually seen only with advanced tumors. Because of the indolent course of MTC, the secretion of adrenocorticotropin by the tumor may cause typical Cushing syndrome (see Chap. 75 and Chap. 219). Calcitonin Determination. Assay of plasma calcitonin is helpful in the early diagnosis of C-cell hyperplasia and MTC and has influenced the management of these disorders, particularly the MEN2 syndromes. Elevated basal plasma calcitonin levels are found in almost all patients with a palpable MTC and correlate directly with tumor mass.190 Basal calcitonin may not be elevated in patients with small tumors and is almost invariably normal in those with C-cell hyperplasia; however, the plasma calcitonin levels increase to abnormally high levels after stimulation with calcium or pentagastrin. Pentagastrin, 0.5 µg/kg, is given intravenously over 5 seconds, and blood samples are taken at 0, 2, and 5 minutes. Depending on the calcitonin antibody used, this stimulus causes calcitonin levels to rise approximately three- to fivefold

with MTC or C-cell hyperplasia.190 Calcium and pentagastrin infused together is more reliable than either used alone because some patients respond to one but not the other. In the combined test, 2 mg/kg of elemental calcium is infused over 60 seconds, followed by 0.5 µg/kg of pentagastrin infused over 5 to 10 seconds, and plasma calcitonin is measured before and 1, 2, 3, 5, and 10 minutes after infusion. Depending on the calcitonin antibody used, the plasma calcitonin level rises approximately five-fold with MTC or C-cell hyperplasia if the basal calcitonin level is minimally elevated. Diagnosis of familial cases is now possible with genetic screening long before the thyroid tumor is clinically manifest or calcitonin levels are elevated (see Factors Influencing Prognosis). Hypercalcitoninemia is not absolutely diagnostic of MTC because it occurs in other conditions.190 When the differential diagnosis of a high plasma calcitonin value is between MTC and another malignancy, a higher calcitonin value and a palpable thyroid tumor usually can identify patients with MTC. In addition, calcitonin secretion by other cancers is poorly stimulated by pentagastrin190 (see Chap. 53). FACTORS INFLUENCING PROGNOSIS MTC is much more aggressive than papillary and follicular carcinoma and has a cancer-specific mortality of ~20% at 10 years.182 Also, a significant number of deaths occur from pheo-chromocytoma.194 The survival rate is substantially worse with sporadic tumors or when metastases are found at the time of diagnosis, with the MEN2B phenotype, and among patients older than 50 years of age at the time of diagnosis. However, the most important prognostic factors are age and tumor stage at the time of diagnosis and the presence of residual disease postoperatively.195,196,196a Prognosis is best with FMTC and MEN2A. Early detection and treatment has a profound impact on the clinical course of MTC. The 10-year survival rates are nearly similar to those in unaffected subjects when nodal metastases are not present, but fall to ~45% with nodal metastases.190 Patients operated on during the first decade of life generally have no evidence of residual disease postoperatively.190 However, persistent or recurrent disease is seen in approximately one-third of patients operated on in the second decade and gradually increases in frequency with age until the seventh decade, when approximately two-thirds of patients have persistent disease after surgery.190 This is largely due to the clinical stage of the disease at the time of surgery.195,196,196a Before 1970, MTC was usually diagnosed in the fifth or sixth decade. With periodic calcitonin screening, affected patients with MEN kindred have been diagnosed at a much earlier age, usually in the second decade or earlier, when they have C-cell hyperplasia or microscopic carcinoma confined to the thyroid.197 Now, with the availability of genetic testing, affected patients can be identified at birth.198 However, the sensitivity of genetic screening for MEN2A offered by diagnostic laboratories that limit RET analysis to exons 10 and 11 is ~83%.199 Genetic testing that includes RET exon 14 results in a more complete and accurate analysis with a sensitivity approaching 95%.199 It is recommended that clinicians confirm the comprehensiveness of a laboratory's genetic screening approach for MEN2A to ensure thoroughness of sample analysis. THERAPY Initial Surgery. Surgery offers the only chance for cure and should be performed as soon as the disease is detected.195,196 Before thyroidectomy, however, pheochromocytoma must be rigorously searched for and excised. The treatment of MTC confined to the neck is total thyroidectomy because the disease is often multicentric, even in patients with a negative family history who are often unsuspected relatives of affected MEN2 kindred.195,196 Because cervical node metastases occur early and adversely influence survival, all patients with palpable MTC or clinically occult disease that is visible on cut section of the thyroid should undergo routine dissection of lymph nodes in the central neck compartment. The lateral lymph nodes should be dissected when they contain tumor, but radical neck dissection is not recommended unless the jugular vein, accessory nerve, or sternocleidomastoid muscle is invaded by tumor.107 Residual or Recurrent Medullary Thyroid Carcinoma. Residual or recurrent MTC, manifested by elevated calcitonin levels, occurs commonly after primary treatment of the tumor. Reoperation in appropriately selected patients is the only treatment that consistently and reliably reduces calcitonin levels and may result in excellent local disease control. Although reoperative neck microdissections can normalize calcitonin levels when metastatic MTC is confined to regional lymph nodes, there is no curative therapy for widely metastatic disease. Improved results are reported with surgical management of recurrent MTC, mainly through better preoperative selection of patients and the institution of routine laparoscopic liver examination preoperatively, which identifies distant metastases in patients with normal computed tomography and magnetic resonance imaging.200 Patients with widely metastatic MTC often live for years, but many develop symptoms secondary to tumor persistence or progression. Judicious palliative, reoperative resection of discrete, symptomatic lesions provides significant long-term relief of symptoms with minimal operative mortality and morbidity.201 Patients with metastatic MTC causing significant symptoms or physical compromise may respond to palliative reoperative resection despite the presence of widespread incurable metastatic disease.201 Inoperable Disease. Patients with inoperable disease are often given palliative treatment, with external radiation for localized disease, or with doxorubicin or other chemotherapy combinations for widespread, life-threatening disease, which is of limited benefit. Radiolabeled metaiodobenzylguanidine and111 In-octreotide are potentially useful in palliative care. Octreotide does not improve the natural course of advanced stages of MTC.202 Initial reports of the aggressive use of radioimmunotherapy with radiolabeled monoclonal antibodies against carcinoembryonic antigen in patients with far advanced disease appear hopeful. The phase I studies, which show the safety of administering high myeloablative doses of 131I-MN-14 F(ab)2 labeled carcinoembryonic antigen, are encouraging but require confirmation.202a FOLLOW-UP The efficacy of surgery for MTC is assessed postoperatively by measurement of plasma calcitonin levels, which may require as long as 6 months to normalize. A normal basal and provoked calcitonin level after surgery indicates cure in most cases. Persistent modest basal calcitonin elevation is often seen after surgery in patients who may remain well for many years, particularly those from a MEN2A kindred who should be observed without further aggressive therapy. When plasma calcitonin levels are extremely high or when diarrhea occurs postoperatively, metastases can be localized by neck palpation, chest radiography, computed tomography, isotope bone (but not liver) scans, liver biopsy, angiography, and venous catheterization with calcitonin measurement. Both 99m Tc-dimercaptosuccinic acid and111 In-octreotide studies have similar sensitivity to localized primary MTC; however, these scans do not detect small lymph node involvement (micrometastases) before initial surgery.203 Unfortunately, both scans have no clinical implication for preoperative MTC staging. FAMILY SCREENING Genetic Screening. Genetic screening should be done in all first-degree relatives of any patient who tests positive for a MEN2 or FMTC mutation. One study found that without genetic testing, even when the mean age at the time of thyroidectomy was ~10 years, a significant number of patients (21%) with MEN2A or MEN2B have persistent or recurrent MTC over a follow-up time of approximately a decade.204 Prophylactic Total Thyroidectomy. In 1993, when mutations of the RET protooncogene were found to account for hereditary MTC, surgeons gained the opportunity to prophylactically operate on patients at an asymptomatic stage or before the disease became clinically manifest. With this approach, microscopic or grossly evident MTC is often present in the excised thyroid glands, but almost none of the patients have metastasis of their MTC to regional lymph nodes at the time of surgery.198 Almost all patients are biochemically cured with prophylactic total thyroidectomy, which can be performed safely in experienced centers.107 Prophylactic total thyroidectomy done by an experienced surgeon is recommended at age 6 years for patients who test genetically positive for MEN2A.107,205 Thyroidectomy should be done at an even earlier age for children who test positive for MEN2B, because of its aggressive behavior.204,206 Central neck lymph node dissection should be included when calcitonin levels are elevated or if patients are older than 10 years.107 Surgical Management of Hyperparathyroidism. Surgical management of hyperparathyroidism in those with MEN2A is controversial. Some advocate total parathyroidectomy and heterotopic autotransplantation for all MEN2A patients with hyperparathyroidism.207 Others advocate preservation of the parathyroids, which obviates the potential morbidity associated with total parathyroidectomy and autotransplantation.208,209 Screening for Pheochromocytoma. Before thyroidectomy is performed, those with MEN2 also should have catecholamine and metanephrine measurements. All patients with MTC require postoperative follow-up of serum calcitonin levels. CHAPTER REFERENCES 1. Landis SH, Murray T, Bolden S, Wingo PA. Cancer statistics, 1999. CA Cancer J 1999; 49:8.

2. 3. 4. 5. 6.

Kosary CL, Ries LAG, Miller BA, et al. SEER cancer statistic review, 1973– 1992: tables and graphs. Bethesda, MD: National Cancer Institute. NIH Pub. No. 96-2789, 1995. Mazzaferri EL, Jhiang SM. Long-term impact of initial surgical and medical therapy on papillary and follicular thyroid cancer. Am J Med 1994; 97:418. Hall P, Holm LE. Radiation-associated thyroid cancer: facts and fiction. Acta Oncol 1998; 37:325. Ron E, Lubin JH, Shore RE, et al. Thyroid cancer after exposure to external radiation: a pooled analysis of seven studies. Radiat Res 1995; 141:259. Ron E, Doddy MM, Becker DV, et al. Cancer mortality following treatment for adult hyperthyroidism. JAMA 1998; 280:347.

6a.Franklyn JA, Maisonneuve P, Sheppard M, et al. Cancer incidence and mortality after radioiodine treatment for hyperthyroidism: a population-based cohort study. Lancet 1999; 353:2111. 7. Jacob P, Goulko G, Heidenreich WF, et al. Thyroid cancer risk to children calculated. Nature 1998; 392:31. 8. National Cancer Institute. Estimated exposures and thyroid doses received by the American people from iodine-131 in fallout following Nevada atmospheric nuclear bomb tests: a report from the National Cancer Institute. Washington, DC: US Department of Health and Human Services, National Institutes of Health, National Cancer Institute, 1997. 9. Institute of Medicine. Committee on exposure of the American people to I-131 from the Nevada atomic-bomb test: implications for public health. Washington, DC: National Academy Press, 1999. 10. Gilbert ES, Tarone R, Bouville A, Ron E. Thyroid cancer rates and 131 I doses from Nevada atmospheric nuclear bomb tests. Am J Med 1998; 90:1654. 11. Mazzaferri EL. Thyroid carcinoma: papillary and follicular. In: Mazzaferri EL, Samaan N, eds. Endocrine tumors. Cambridge, England: Blackwell Scientific Publications Inc., 1993:278. 12. Robbins J, Schneider AB. Radioiodine-induced thyroid cancer: studies in the aftermath of the accident at Chernobyl. Trends 1998; 9:87. 13. Moosa M, Mazzaferri EL. Occult thyroid carcinoma. Cancer J 1997; 10:180. 14. Burgess JR, Duffield A, Wilkinson SJ, et al. Two families with an autosomal dominant inheritance pattern for papillary carcinoma of the thyroid. J Clin Endocrinol Metab 1997; 82:345. 15. Malchoff CD, Sarfarazi M, Tendler B, et al. Familial papillary thyroid carcinoma is genetically distinct from familial adenomatous polyposis coli. Thyroid 1999; 9:247. 15a.Lupoli G, Vitale G, Caraglia M, et al. Familial papillary thyroid microcarcinoma: a new clinical entity. Lancet 1999; 353:37. 16. Bell B, Mazzaferri EL. Familial adenomatous polyposis (Gardner's syndrome) and thyroid carcinoma: a case report and review of the literature. Dig Dis Sci 1993; 38:185. 17. Cetta F, Olschwang S, Petracci M, et al. Genetic alterations in thyroid carcinoma associated with familial adenomatous polyposis: clinical implications and suggestions for early detection. World J Surg 1998; 22:1231. 18. Burman KD, Ringel MD, Wartofsky L. Unusual types of thyroid neoplasms. Endocrinol Metab Clin North Am 1996; 25:49. 19. Dahia PLM, Marsh DJ, Zheng ZM, et al. Somatic deletions and mutations in the Cowden disease gene, PTEN, in sporadic thyroid tumors. Cancer Res 1997; 57:4710. 20. Stratakis CA, Courcoutsakis NA, Abati A, et al. Thyroid gland abnormalities in patients with the syndrome of spotty skin pigmentation, myxomas, endocrine overactivity, and schwannomas (Carney complex). J Clin Endocrinol Metab 1997; 82:2037. 21. Evans HL. Encapsulated columnar-cell neoplasms of the thyroid: a report of four cases suggesting a favorable prognosis. Am J Surg Pathol 1996; 20:1205. 22. de los Santos ET, Keyhani-Rofagha S, Cunningham JJ, Mazzaferri EL. Cystic thyroid nodules: the dilemma of malignant lesions. Arch Intern Med 1990; 150:1422. 23. Baudin E, Travagli JP, Ropers J, et al. Microcarcinoma of the thyroid gland: the Gustave-Roussy Institute experience. Cancer 1998; 83:553. 24. Hedinger C, Williams ED, Sobin LH. The WHO histological classification of thyroid tumors: a commentary on the second edition. Cancer 1989; 63:908. 25. Mazzaferri EL, Young RL, Oertel JE, et al. Papillary thyroid carcinoma: the impact of therapy in 576 patients. Medicine (Baltimore) 1977; 56:171. 26. LiVolsi V, Asa SL. The demise of follicular carcinoma of the thyroid gland. Thyroid 1994; 4:233. 27. LiVolsi VA. Papillary lesions of the thyroid. In: LiVolsi VA, ed. Surgical pathology of the thyroid. Philadelphia: WB Saunders, 1990:136. 28. Noguchi M, Yamada H, Ohta N, et al. Regional lymph node metastases in well-differentiated thyroid carcinoma. Int Surg 1987; 72:100. 29. Sugino K, Ito K Jr, Ozaki O, et al. Papillary microcarcinoma of the thyroid. J Endocrinol Invest 1998; 21:445. 30. Yamashita H, Noguchi S, Murakami N, et al. Extracapsular invasion of lymph node metastasis is an indicator of distant metastasis and poor prognosis in patients with thyroid papillary carcinoma. Cancer 1997; 80:2268. 31. Pacini F, Lari R, Mazzeo S, et al. Diagnostic value of a single serum thyroglobulin determination on and off thyroid suppressive therapy in the follow-up of patients with differentiated thyroid cancer. Clin Endocrinol 1985; 23:405. 32. Pineda JD, Lee T, Ain K, et al. Iodine-131 therapy for thyroid cancer patients with elevated thyroglobulin and negative diagnostic scan. J Clin Endocrinol Metab 1995; 80:1488. 33. Schlumberger M, Arcangioli O, Piekarski JD, et al. Detection and treatment of lung metastases of differentiated thyroid carcinoma in patients with normal chest X-rays. J Nucl Med 1988; 29:1790. 34. Schlumberger M, Mancusi F, Baudin E, Pacini F. 131-I therapy for elevated thyroglobulin levels. Thyroid 1997; 7:273. 35. Yoo KS, Chengazi VU, O'Mara RE. Thyroglossal duct cyst with papillary carcinoma in an 11-year-old girl. J Pediatr Surg 1998; 33:745. 36. LiVolsi VA. Unusual variants of papillary thyroid carcinoma. In: Mazza-ferri EL, Kreisberg RA, Bar RS, eds. Advances in endocrinology and metabolism, 6th ed. St. Louis: Mosby-Year Book, Inc., 1995:39. 37. Ain KB. Papillary thyroid carcinoma etiology, assessment, and therapy. Endocrinol Metab Clin North Am 1995; 24:711. 38. Sobrinho-Simoes M, Soares J, Carneiro F. Diffuse follicular variant of papillary carcinoma of the thyroid: report of eight cases of a distinct aggressive type of tumor. Surg Path 1990; 3:189. 39. Nikiforov Y, Gnepp DR. Pediatric thyroid cancer after the Chernobyl disaster: pathomorphologic study of 84 cases (1991–1992) from the Republic of Belarus. Cancer 1994; 74:748. 40. Herrera MF, Hay ID, Wu PS, et al. Hürthle cell (oxyphilic) papillary thyroid carcinoma: a variant with more aggressive biologic behavior. World J Surg 1994; 16:669. 41. Katoh R, Harach HR, Williams ED. Solitary, multiple, and familial oxyphil tumours of the thyroid gland. J Pathol 1998; 186:292. 42. Rosai J. Papillary carcinoma. Monogr Pathol 1993; 138. 43. LiVolsi VA. Follicular lesions of the thyroid. In: LiVolsi VA, ed. Surgical pathology of the thyroid. Philadelphia: WB Saunders, 1990:173. 44. Hassoun AAK, Hay ID, Goellner JR, Zimmerman D. Insular thyroid carcinoma in adolescents: a potentially lethal endocrine malignancy. Cancer 1997; 79:1044. 45. Schneider AB, Shore-Freedman E, Ryo UY, et al. Radiation-induced tumors of the head and neck following childhood irradiation: prospective studies. Medicine (Baltimore) 1985; 64:1. 46. Mazzaferri EL. Management of a solitary thyroid nodule. N Engl J Med 1993; 328:553. 47. Schneider AB, Bekerman C, Leland J, et al. Thyroid nodules in the follow-up of irradiated individuals: comparison of thyroid ultrasound with scanning and palpation. J Clin Endocrinol Metab 1997; 82:4020. 48. Ezzat S, Sarti DA, Cain DR, Braunstein GD. Thyroid incidentalomas: prevalence by palpation and ultrasonography. Arch Intern Med 1994; 154:1838. 49. Tan GH, Gharib H. Thyroid incidentalomas: management approaches to nonpalpable nodules discovered incidentally on thyroid imaging. Ann Intern Med 1997; 126:226. 50. Fogelfeld L, Wiviott MBT, Shore-Freedman E, et al. Recurrence of thyroid nodules after surgical removal in patients irradiated in childhood for benign conditions. N Engl J Med 1989; 320:835. 51. Dottorini ME, Vignati A, Mazzucchelli L, et al. Differentiated thyroid carcinoma in children and adolescents: a 37-year experience in 85 patients. J Nucl Med 1997; 38:669. 52. Harach HR, Williams ED. Childhood thyroid cancer in England and Wales. Br J Cancer 1995; 72:777. 53. Hung W. Well-differentiated thyroid carcinomas in children and adolescents: a review. Endocrinologist 1994; 4:117. 54. Samuel AM, Rajashekharrao B, Shah DH. Pulmonary metastases in children and adolescents with well-differentiated thyroid cancer. J Nucl Med 1998; 39:1531. 55. Zimmerman D, Hay ID, Gough IR, et al. Papillary thyroid carcinoma in children and adults: long-term follow-up of 1039 patients conservatively treated at one institution during three decades. Surgery 1988; 104:1157. 56. Schlumberger M, De Vathaire F, Travagli JP, et al. Differentiated thyroid carcinoma in childhood: long term follow-up of 72 patients. J Clin Endocrinol Metab 1987; 65:1088. 57. Hay ID. Papillary thyroid carcinoma. Endocrinol Metab Clin North Am 1990; 19:545. 58. Pasieka JL, Thompson NW, McLeod MK, et al. The incidence of bilateral well-differentiated thyroid cancer found at completion thyroidectomy. World J Surg 1992; 16:711. 59. Shaha AR, Loree TR, Shah JP. Prognostic factors and risk group analysis in follicular carcinoma of the thyroid. Surgery 1995; 118:1131. 60. Sanders LE, Cady B. Differentiated thyroid cancer: reexamination of risk groups and outcome of treatment. Arch Surg 1998; 133:419. 61. Sarda AK, Bal S, Kapur MM. Near-total thyroidectomy for carcinoma of the thyroid. Br J Surg 1989; 76:90. 62. DeGroot LJ, Kaplan EL, McCormick M, Straus FH. Natural history, treatment, and course of papillary thyroid carcinoma. J Clin Endocrinol Metab 1990; 71:414. 63. Taylor T, Specker B, Robbins J, et al. Outcome after treatment of high-risk papillary and non-Hürthle-cell follicular thyroid carcinoma. Ann Intern Med 1998; 129:622. 64. Mazzaferri EL. Thyroid remnant 131 I ablation for papillary and follicular thyroid carcioma. Thyroid 1997; 7:265. 65. Massin JP, Savoie JC, Garnier H, et al. Pulmonary metastases in differentiated thyroid carcinoma: study of 58 cases with implications for the primary tumor treatment. Cancer 1984; 53:982. 66. Loh KC, Greenspan FS, Gee L, et al. Pathological tumor-node-metastasis (pTNM) staging for papillary and follicular thyroid carcinomas: a retrospective analysis of 700 patients. J Clin Endocrinol Metab 1997; 82:3553. 67. Cady B. Staging in thyroid carcinoma. Cancer 1998; 83:844. 68. Tsang TW, Brierley JD, Simpson WJ, et al. The effects of surgery, radioiodine, and external radiation therapy on the clinical outcome of patients with differentiated thyroid carcinoma. Cancer 1998; 82:375. 69. Simpson WJ, McKinney SE, Carruthers JS, et al. Papillary and follicular thyroid cancer: prognostic factors in 1578 patients. Am J Med 1987; 83:479. 70. Kurozumi K, Nakao I, Nishida T, et al. Significance of biologic aggressiveness and proliferating activity in papillary thyroid carcinoma. World J Surg 1998; 22:1237. 71. Sugitani I, Yanagisawa A, Shimizu A, et al. Clinicopathologic and immunohistochemical studies of papillary thyroid microcarcinoma presenting with cervical lymphadenopathy. World J Surg 1998; 22:731. 72. Akslen LA, Varhaug JE. Oncoproteins and tumor progression in papillary thyroid carcinoma: presence of epidermal growth factor receptor, c-erbB-2 protein, estrogen receptor related protein, p21-ras protein, and proliferation indicators in relation to tumor recurrences and patient survival. Cancer 1995; 76:1643. 73. Sellers M, Beenken S, Blankenship A, et al. Prognostic significance of cervical lymph node metastases in differentiated thyroid cancer. Am J Surg 1992; 164:578. 74. Sisson JC, Giordano TJ, Jamadar DA, et al. 131 I treatment of micronodular pulmonary metastases from papillary thyroid carcinoma. Cancer 1996; 78:2184. 75. Nêmec J, Zamrazil V, Pohunková D, Röhling S. Radioiodide treatment of pulmonary metastases of differentiated thyroid cancer: results and prognostic factors. Nuklearmedizin 1979; 18:86. 76. Casara D, Rubello D, Saladini G, et al. Different features of pulmonary metastases in differentiated thyroid cancer: natural history and multivariate statistical analysis of prognostic variables. J Nucl Med 1993; 34:1626. 77. Kashima K, Yokoyama S, Noguchi S, et al. Chronic thyroiditis as a favorable prognostic factor in papillary thyroid carcinoma. Thyroid 1998; 8:197. 78. Schäffler A, Palitzsch KD, Seiffarth C, et al. Coexistent thyroiditis is associated with lower tumour stage in thyroid carcinoma. Eur J Clin Invest 1998; 28:838. 79. Viswanathan K, Gierlowski TC, Schneider AB. Childhood thyroid cancer: characteristics and long-term outcome in children irradiated for benign conditions of the head and neck. Am J Dis Child 1994; 148:260. 80. Filetti S, Belfiore A, Amir SM, et al. The role of thyroid-stimulating antibodies of Graves' disease in differentiated thyroid cancer. N Engl J Med 1988; 318:753. 81. Belfiore A, Garofalo MR, Giuffrida D, et al. Increased aggressiveness of thyroid cancer in patients with Graves' disease. J Clin Endocrinol Metab 1990; 70:830. 81a.Mazzaferri EL. Thyroid cancer and Graves' disease: the controversy ten years later. Endocrine Practice 2000; 6:221. 82. Fusco A, Grieco M, Santoro M, et al. A new oncogene in human thyroid papillary carcinomas and their lymph-nodal metastases. Nature 1987; 328:170. 83. Donghi R, Sozzi G, Pierotti MA, et al. The oncogene associated with human papillary thyroid carcinoma (PTC) is assigned to chromosome 10 q11-q12 in the same region as multiple endocrine neoplasia type 2A (MEN2A). Oncogene 1989; 4:521. 84. Smanik PA, Furminger TL, Mazzaferri EL, Jhiang SM. Breakpoint characterization of the ret/PTC oncogene in human papillary thyroid carcinoma. Hum Mol Genet 1995; 4:2313. 85. Nikiforov YE, Nikiforova M, Fagin JA. Radiation-induced post-Chernobyl pediatric thyroid carcinomas. Oncogene 1998; 17:1983. 86. Jhiang SM, Sagartz JE, Tong Q, et al. Targeted expression of the ret/PTC1 oncogene induces papillary thyroid carcinomas. Endocrinology 1996; 137:375. 87. Santoro M, Chiappetta G, Cerrato A, et al. Development of thyroid papillary carcinomas secondary to tissue-specific expression of the RET/PTC1 oncogene in transgenic mice. Oncogene 1996; 12:1821. 88. American Joint Committee on Cancer. Head and neck tumors: thyroid gland. In: Beahrs OH, Henson DE, Hutter RVP, Myers MH, eds. Manual for staging of cancer, 4th ed. Philadelphia: JB Lippincott, 1992:53. 89. Van De Velde CJH, Hamming JF, Goslings BM, et al. Report of the consensus development conference on the management of differentiated thyroid cancer in the Netherlands. Eur J Cancer Clin Oncol 1988; 24:287. 90. Baldet L, Manderscheid JC, Glinoer D, et al. The management of differentiated thyroid cancer in Europe in 1988: results of an international survey. Acta Endocrinol (Copenh) 1989; 120:547. 91. Solomon BL, Wartofsky L, Burman KD. Current trends in the management of well differentiated papillary thyroid carcinoma. J Clin 1996; 81:333.

92. Chen H, Zeiger MA, Clark DP, et al. Papillary carcinoma of the thyroid: can operative management be based solely on fine-needle aspiration? J Am Coll Surg 1997; 184:605. 93. Cady B. Our AMES is true: how an old concept still hits the mark: or, risk group assignment points the arrow to rational therapy selection in differentiated thyroid cancer. Am J Surg 1997; 174:462. 94. Schlumberger MJ. Medical progress: papillary and follicular thyroid carcinoma. N Engl J Med 1998; 338:297. 95. Mazzaferri EL. Treating differentiated thyroid carcinoma: where do we draw the line? Mayo Clin Proc 1991; 66:105. 95a.Mazzaferri EL. NCCN thyroid carcinoma practice guidelines. Oncology 1999; 13:391. 96. Newman KD, Black T, Heller G, et al. Differentiated thyroid cancer: determinants of disease progression in patients 400 mg per day or 40 mg per day, respectively). If there is any question that PTU or MMI may be causing an adverse effect, the individual drug should be discontinued and the patient either observed without medication for a short time or the alternative agent substituted if needed. If there was a serious adverse reaction to one agent, the alternative agent should probably not be used. There is an estimated 20% to 50% likelihood that a patient with a reaction to one agent (e.g., PTU) will have an adverse reaction to the other (i.e., MMI). There has been interest in the use of recombinant human granulocyte colony-stimulating factor in the treatment of antithyroid agent-induced agranulocytosis.64,64a In a review of generalized reactions to PTU or MMI, fewer than 100 cases had been reported since 1945.65 The most common, generalized reactions included vasculitis, lupus erythematosus, and polyarthritis. Reactions were more common with PTU than with MMI, but occasionally, even with discontinuation of the medication, fatality can occur. Cross-reactions to PTU and MMI may occur.

TABLE 42-6. Potential Side Effects of Antithyroid Medications

Most patients tolerate PTU or MMI, and therapy is continued as biochemical and clinical euthyroidism returns. At this time a decision is made whether to continue the antithyroid medications (Fig. 42-8). Some physicians maintain patients on PTU or MMI for ~1 year in an effort to “induce” a remission (Table 42-7 and Table 42-8). In the United States, it is likely that only 20% to 40% of patients will have a remission; a small goiter size, response to low doses of PTU or MMI, mild thyrotoxicosis at presentation, and perhaps low-titer TSH receptor antibodies and lack of HLA-Dw3 and B8 antigens all increase the likelihood of remission (Table 42-9). Different

geographic areas with variable iodine intake may be associated with different remission rates. In one multicenter study, the efficiency of 40 mg versus 10 mg of MMI was compared in the treatment of Graves disease.60 In this study, 196 of 309 (63%) patients achieved a remission, with the differences in the two groups not being significantly different. Adverse reactions occurred in 16% of the group taking 10 mg and in 26% of the group taking 40 mg, with ~0.8% having a serious hematologic abnormality.

FIGURE 42-8. Upper panel: Patient with Graves disease treated with antithyroid drug. Left: Appearance before therapy. Right: Four months after commencement of therapy. Note the markedly decreased stare and the weight gain. Lower panel: Patient with Graves disease treated with radioiodine. Left: Before therapy. Right: Six months later.

TABLE 42-7. Indications for the Different Therapies of Graves Disease

TABLE 42-8. Complications of Therapies for Graves Disease

TABLE 42-9. Long-Term Antithyroid Medication as a Treatment for Graves Disease

b-BLOCKERS AND OTHER AGENTS When a thyrotoxic patient is symptomatic, a b-adrenergic blocker is often prescribed at the same time PTU or MMI therapy is initiated (see Table 42-5).66 b-Blockers improve many of the symptoms of thyrotoxicosis, improve myocardial efficiency, reduce myocardial oxygen demand, and may decrease nitrogen loss. It may be advisable to recommend that a hyperthyroid patient receiving a b-blocker record his or her pulse several times daily. The time-honored b-blocker is propranolol, and the oral dose range is 20 to 40 mg every 8 hours for mild to moderate thyrotoxicosis and as high as 60 to 80 mg four times a day for more severe cases. The therapeutic objective is subjective improvement and the restoration of the pulse rate to ~80 beats per minute. Newer b-blocking agents have the advantages of cardioselectivity and longer duration of action. A thyrotoxic patient responds very rapidly to the administration of b-blockers. If no clinical or subjective response is noted in 2 to 3 days, the dose should be increased. b-Blockers have no direct effect on radioactive iodine uptake or the synthesis or secretion of T4 or T3. These agents should be given with caution, if at all, to patients with a history of asthma. High-output cardiac failure may be improved by b-blockers, but their use in such patients should be judicious. It is usually not prudent to use long-term b-blockers as the sole therapy in the treatment of Graves disease. Because b-blockers do not alter T4 or T3 synthesis or secretion, the underlying thyrotoxicosis may worsen while a patient is taking b-blockers alone. If a patient is allergic to both PTU and MMI, if the thyrotoxicosis has not responded appropriately, or if the thyrotoxicosis requires more rapid restoration of the euthyroid state, alternative agents such as ipodate, other iodides, or lithium carbonate should be considered (see Table 42-5).64,67,68 and 69 Patients taking one of these latter antithyroid agents usually need closer medical supervision and possibly hospitalization. IODIDES Iodides may be very effective antithyroid agents but should be used in thyrotoxicosis only when specific indications exist.70 Iodides should usually not be used as customary or sole therapy because there can be an “escape” from their inhibitory effects, and the resultant exacerbation in thyrotoxicosis may be extremely difficult to

control. For this reason, iodides are typically used in combination with thioureas, and therapy is not instituted until a patient has been started on PTU or MMI. Iodides can decrease serum T4 and T3 levels by 30% to 50% in 10 days; iodide and lithium may have an additive effect on decreasing serum T4.68,69 A thorough medical history, with specific inquiry as to allergies to iodides, should be taken before initiating therapy with ipodate or other iodides. The doses of iodides given with these agents are enormous when compared with the minimal daily requirement, and the consequent suppression of radioactive iodine uptake compromises the ability to use radioiodine for either diagnostic studies or therapy. Data suggest that 131I can probably be used effectively again as early as 4 to 6 weeks after ipodate is discontinued,67 but this area requires further investigation. Iodide (or ipodate) does not interfere technically with the measurement of serum iodothyronines or TSH. Iodides, ipodate, and lithium are especially useful in lowering the serum T4 and T3 levels, and thus decrease the chance of thyroid storm when a thyrotoxic patient is being prepared for surgery and there is insufficient time to administer PTU or MMI alone for several weeks. Of the three agents noted, ipodate is a good choice in nonpregnant individuals; lithium therapy requires extremely close observation with frequent serum monitoring of electrolytes, liver function tests, and lithium levels. Ipodate is an iodide-containing radiocontrast agent that is very effective in the treatment of thyrotoxicosis.67 The ipodate molecule inhibits extrathyroidal T4 to T3 conversion, and a dose of 0.5 to 3 g decreases the serum T3 level by 50% within several days. The iodide released from ipodate may help decrease thyroidal secretion when used for short periods (i.e., 55 mm per hour); normal or near-normal leukocyte count; transient elevations of antithyroid antibodies in low titers; and serum T4 and T3 values that may be high, normal, or low, depending on the phase of the disease. In the thyrotoxic phase, the serum thyroid-stimulating hormone (TSH) value is suppressed but may be detectable in third-generation assays in nearly 80% of those evaluated.10 TSH is normal during the euthyroid phase and elevated during the hypothyroid phase. A thyroid scan during an episode of subacute thyroiditis shows a cold region in the involved section of the thyroid. Additionally, during the thyrotoxic and euthyroid phases, the radioactive iodine (RAI) uptake is often 20 g, but usually it is 1 to 3 g.4 Lobulation, a red-brown color, and a homogeneous appearance are characteristic. Microscopically, two major patterns are found. In diffuse chief cell hyperplasia, solid masses of cells are present with minimal or absent stromal fat. Usually, these sheets comprise chief cells with rare oxyphil elements. Nodular (adenomatous or pseudoadenomatous) hyperplasia consists of circumscribed nodules of chief or oxyphil cells. Each nodule is devoid of fat, and little fat is found in the intervening stroma.1,4 Usually, no rim of normal parathyroid can be found; however, in nodular glands, relatively parvicellular zones may encompass totally cellular nodules, mimicking an adenoma.4 Bizarre nuclei are rare in primary hyperplasia; mitoses are found occasionally.19 Electron microscopy reveals active-appearing chief cells, with large Golgi complexes, secretory vacuoles, and prominent endoplasmic reticulum virtually identical with chief cell adenomas.11 Usually, intracellular fat is decreased or absent at the ultrastructural level and by fat stain.13,15,20 CLEAR CELL (WATER CLEAR CELL) HYPERPLASIA Clear cell (water clear cell) hyperplasia,21 a rare cause of primary hyperparathyroidism, occurs as a sporadic, nonfamilial disorder. Grossly, unlike other hyperplasias, the superior glands are larger than the lower. Total weight of such parathyroids always exceeds 1 g, and weights of 5 to 10 g may occur.1,4 The glands are irregular and exhibit cysts. No nodules are seen. The color is distinctly mahogany. Histologically, the glands are composed of diffuse sheets of water clear cells without the admixture of other cell types. No rim of normal tissue is present. The clear cell is large, 10 to 15 mm in diameter, with sharp cell borders that occasionally fuse. The nuclei are located basally. Nuclear pleomorphism is absent, although multinucleation may be observed. Acini, papillae, or follicles may form and may be interspersed with solid sheets of clear cells.1,4 Ultrastructurally, the cytoplasm is not empty but contains numerous membrane-bound vacuoles, secretion granules, Golgi apparatus, and endoplasmic reticulum.11

PARATHYROID CARCINOMA Parathyroid carcinoma is responsible for 90% accuracy in distinguishing abnormal from normal parathyroid glands with this technique performed intraoperatively.29

OTHER TYPES OF PARATHYROID PATHOLOGY SECONDARY HYPERPARATHYROIDISM Grossly, in secondary hyperparathyroidism, all four glands are enlarged. Without clinical data, however, primary hyperplasia cannot be distinguished from secondary parathyroid hyperplasia microscopically.4 The anatomic variations of the parathyroid gland locations makes the surgical treatment of secondary hyperparathyroidism a challenge.30 FAMILIAL HYPOCALCIURIC HYPERCALCEMIA In familial hypocalciuric hypercalcemia (see Chap. 58) the parathyroid glands are described as normal or as showing minimal hyperplasia.31 HUMORAL HYPERCALCEMIA OF MALIGNANCY In patients with hypercalcemia caused by a nonparathyroid malignancy that is secreting a humoral substance, the parathyroids would be expected to demonstrate a normal histologic appearance32 or atrophy, because they would be suppressed by the systemic hypercalcemia. Although uncommon, hyperplastic parathyroid tissue in such patients has been described by some authors.5 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Grimelius L, Akerström G, Johansson H, Bergstrüm R. Anatomy and histopathology of human parathyroid glands. Pathol Annu 1981; 16(Part II):1. Akerström G, Malmaeus J, Bergström S. Surgical anatomy of human parathyroid glands. Surgery 1984; 95:14. Wang CA. The anatomic basis of parathyroid surgery. Ann Surg 1976; 183:271. Castleman B, Roth SI. Tumors of the parathyroid glands, series II, fasc 14. Washington: Armed Forces Institute of Pathology, 1978. Akerström G, Grimelius L, Johansson H, et al. Estimation of parathyroid parenchymal cell mass by density gradients. Am J Pathol 1980; 99:685. Dufour DR, Wilkerson SY. Factors related to parathyroid weight in normal persons. Arch Pathol Lab Med 1983; 107:167. Dufour DR, Wilkerson SY. The normal parathyroid revisited: percent of stromal fat. Hum Pathol 1982; 13:717. Dekker A, Dunsford HA, Geyer SJ. The normal parathyroid gland at autopsy: the significance of stromal fat in adult patients. J Pathol 1979; 128:127. Saffos RO, Rhatigan RM, Urgulu S. The normal parathyroid and the borderline with early hyperplasia: a light microscopic study. Histopathology 1984; 8:407. Cinti S, Balercia G, Zingaretti MC, et al. The normal human parathyroid gland. Submicrosc Cytol 1983; 15:661. Johannessen JV. Parathyroid glands. In: Johannessen JV, ed. Electron microscopy in human medicine, vol 10. New York: McGraw-Hill, 1981:111. LiVolsi VA. Pathology of the parathyroid glands. In: Barnes L, ed. Surgical pathology of the head and neck. New York: Marcel Dekker, 1985:1487. Grimelius L, Johansson H. Pathology of parathyroid tumors. Semin Surg Oncol 1997; 13:142. Dekker A, Watson CG, Barnes EL. The pathologic assessment of primary hyperparathyroidism and its major impact on therapy: a prospective evaluation of 50 cases with oil-red-O stain. Ann Surg 1979; 190:671. Ghandur-Mnaymneh L, Kimura N. The parathyroid adenoma. Am J Pathol 1984; 115:70. Bowiby LS, DeBault LE, Abraham SR. Flow cytometric DNA analysis of parathyroid glands. Am J Pathol 1987; 128:338. Bondeson L, Sandelin K, Grimelius L. Histopathological variables and DNA cytometry in parathyroid carcinoma. Am J Surg Pathol 1993; 17:820. Chaitin BA, Goldman RL. Mitotic activity in benign parathyroid disease. (Letter). Am J Clin Pathol 1981; 76:363. Snover DC, Foucar K. Mitotic activity in benign parathyroid disease. Am J Clin Pathol 1981; 75:345. Dufour DR, Durkowski C. Sudan IV staining: its limitations in evaluating parathyroid functional status. Arch Pathol Lab Med 1982; 106:224. Hedback G, Oden A. Parathyroid water clear cell hyperplasia, an O-allele associated condition. Hum Genet 1994; 94:195. Schantz A, Castleman B. Parathyroid carcinoma: a study of 70 cases. Cancer 1973; 31:600. Favia G, Lumachi F, Polistina F, D'Amico DF. Parathyroid carcinoma: sixteen new cases and suggestions for correct management. World J Surg 1998; 22:1225. Cryns VL, Thor A, Hu HJ, et al. Loss of the retinoblastoma tumor suppressor gene in parathyroid carcinoma. N Engl J Med 1994; 330:757.

25. 26. 27. 28. 29. 30. 31. 32.

Reddick RL, Costa JC, Marx SJ. Parathyroid hyperplasia and parathyromatosis. Lancet 1977; 1:549. Fitko R, Roth SI, Hines JR, et al. Parathyromatosis in hyperparathyroidism. Hum Pathol 1990; 21:234. Sokol MS, Kavolius J, Schaaf M, et al. Recurrent hyperparathyroidism from benign neoplastic seeding: a review with recommendations for management. Surgery 1993; 113:456. Bondeson AG, Bondeson L, Ljungberg O, Tibblin S. Fat staining in parathyroid disease: diagnostic value and impact on surgical strategy. Clinicopathologic analysis of 191 cases. Hum Pathol 1985; 16:1255. Wang CA, Rieder SV. A density test for the intraoperative differentiation of parathyroid hyperplasia from neoplasia. Ann Surg 1978; 187:63. Butterworth PC, Nicholson MI. Surgical anatomy of the parathyroid glands in secondary hyperparathyroidism. J R Coll Surg Edinb 1998; 43:271. Thorgiersson U, Costa J, Marx SJ. The parathyroid glands in familial hypocalciuric hypercalcemia. Hum Pathol 1981; 12:229. Dufour DR, Marx SJ, Spiegel AM. Parathyroid gland morphology in nonparathyroid hormone-mediated hypercalcemia. Am J Surg Pathol 1985; 9:43.

CHAPTER 49 PHYSIOLOGY OF CALCIUM METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 49 PHYSIOLOGY OF CALCIUM METABOLISM EDWARD M. BROWN Role of Calcium Forms of Calcium in Blood Overall Calcium Balance Hormonal Control of Calcium Homeostasis Regulation of Parathyroid Hormone Secretion by Calcium and Other Factors Intracellular Mechanisms Regulating Parathyroid Hormone Release Effects of Parathyroid Hormone on Calcium-Regulating Tissues Role of the Skeleton in Calcium Homeostasis Regulation of Circulating Calcium Concentration Clinical Assessment of Calcium Homeostasis Direct Assays of Calcium-Regulating Hormones in Blood Assessment of the Function of Target Organs for Calcium-Regulating Hormones Chapter References

ROLE OF CALCIUM Calcium ions (Ca2+) play numerous critical roles in both intra-cellular and extracellular physiology (Table 49-1). Intracellular Ca2+ is an important regulator of a variety of cellular functions, including processes as diverse as muscle contraction, hormonal secretion, glycogen metabolism, and cell division.1,2 and 3 Many of these functions are accomplished through the interaction of Ca2+ with intracellular binding proteins, such as calmodulin, which then activate enzymes and other intracellular effectors.2,3 The cytosolic free calcium concentration in resting cells is ~100 nmol/L. It is regulated by channels, pumps, and other transport mechanisms that control the movements of Ca2+ into and out of cells and between various intracellular compartments.1,2 and 3 Consonant with its role as a key intracellular second messenger, the cytosolic free Ca2+ concentration (Cai) can rise by as much as 100-fold (i.e., to 1–10 µmol/L) during cellular activation. Such increases in Cai are the result of uptake of extracellular Ca2+ through Ca2+-permeable channels in the plasma membrane, release of Ca2+ from its intra-cellular stores, such as the endoplasmic reticulum, or both factors. Despite the importance of intracellular Ca2+ in cellular metabolism, this compartment comprises only 1% of total body calcium.4

TABLE 49-1. Some Properties of Calcium in Humans

In contrast to intracellular Ca2+ concentration, the extracellular free Ca2+ concentration (Ca2+o) is ~1 mmol/L. It is closely regulated by a complex homeostatic system involving the parathyroid hormone (PTH)–secreting parathyroid glands and calcitonin-secreting thyroidal C cells as well as specialized Ca2+-transporting cells in the intestine, skeleton, and kidney.4,5,6 and 7 This system regulates the flow of Ca2+ into and out of the body as well as between various bodily compartments, particularly between the skeleton and extracellular fluid. The rigid control of Ca2+o ensures a steadysupply of Ca2+ for vital intracellular functions. Ca2+o has otherimportant roles, such as maintaining intercellular adhesion, promoting the integrity of the plasma membrane, and ensuring the clotting of blood, which further emphasizes the importance of maintaining near constancy of Ca2+o. The total amount of soluble extracellular calcium, however, like intracellular Ca2+, constitutes only a minute fraction of total bodily Ca2+ (~0.1%; see Table 49-1). Most of the Ca2+ within the body (>99%) resides as calcium phosphate salts within the skeleton, where it serves two important functions. First, it protects vital internal organs and acts as a rigid framework that facilitates locomotion and other bodily movements. Second, it provides a nearly inexhaustible reservoir of calcium and phosphate ions for times of need when intestinal absorption and renal conservation are insufficient to maintain adequate levels of these ions within the extracellular fluid. Thus, although Ca2+ within all bodily compartments plays essential roles, the fraction that is most closely regulated by the homeostatic system and, in turn, affects all other compartments is Ca2+o. An understanding of overall extracellular calcium homeostasis requires knowledge of the various forms of Ca2+ in the circulation and extracellular fluid as well as the movements of Ca2+ between the organism and the environment and among various bodily compartments (i.e., overall Ca2+ balance). Finally, Ca2+ homeostasis depends critically on a recognition system for Ca2+o and the effector systems by which changes in the circulating levels of calciotropic hormones (i.e., PTH; 1,25-dihydroxyvitamin D3[1,25(OH)2D3]; and calcitonin) restore Ca2+ homeostasis. The form of Vitamin D produced by the body is vitamin D3, which is then metabolized to 25-hydroxyvitamin D3 and 1,25-dihydroxyvitamin D3 as described later in this chapter. Thus, vitamin D3 and its respective metabolites are referred to throughout this chapter. The reader should recognize, however, that some dietary forms of vitamin D and medications contain vitamin D2. Vitamin D2 is metabolized to 25-hydroxyvitamin D2 and 1,25-dihydroxyvitamin D2, which in humans are equivalent in their potencies to the same metabolites of vitamin D3. As is described in more detail later, mineral ions themselves (i.e., extracellular calcium and phosphate ions) can also function in a calciotropic hormone-like manner, directly regulating the functions of many, if not all, of the tissues involved in maintaining Ca2+ homeostasis.7 FORMS OF CALCIUM IN BLOOD Although the extracellular ionized Ca2+ concentration within the interstitial fluid that bathes the various tissues of the body is perhaps most relevant to the Ca2+ homeostatic system, it is not readily measurable. Generally, the total or ionized serum Ca2+concentration is determined. Of the serum total Ca2+ concentration, some 47% is ionized or free Ca2+, an equivalent amount (~46%) is protein-bound, and the remainder (5–10%) is complexed to small anions, including phosphate, citrate, bicarbonate, and others.8 Ultrafilterable Ca2+ comprises both free and complexed Ca2+, but only the former is metabolically active (i.e., available for binding to extracellular Ca2+ binding sites or for uptake into cells). Albumin is the principal protein that binds Ca2+ in the circulation, and ~75% of bound serum calcium resides on albumin. The remainder is bound to various globulins. Several factors alter the amount of Ca2+ bound to proteins. Hypoalbuminemia and hyperalbuminemia reduce and increase, respectively, the amount of Ca2+ in the blood that is bound to albumin (and, therefore, the total Ca2+ concentration) without altering the level of ionized Ca2+ (see Chap. 60). A simple way to correct the total Ca2+ concentration for changes in serum albumin is to add or subtract 0.8 mg/dL of total Ca2+ for each 1 g/dL decrease or increase, respectively, in serum albumin concentration. Rarely, patients with multiple myeloma present with an increase in total Ca2+ concentration despite a normal serum ionized Ca2+ concentration because of a Ca2+-binding paraprotein in the circulation.9 The binding of Ca2+ to serum proteins is also pH dependent; acidosis reduces binding and alkalosis increases it, thereby producing reciprocal changes in serum ionized Ca2+ concentration. The carpal spasm encountered in some patients during hyperventilation is, in part, due to a concomitant reduction in serum ionized Ca2+ concentration caused by alkalosis. For each increase in pH of 0.1 unit, the serum ionized Ca2+ concentration decreases by ~0.1 mEq/L, and vice versa.

OVERALL CALCIUM BALANCE During skeletal and somatic growth in childhood, more Ca2+enters the body through the gastrointestinal (GI) tract than leaves it through the kidneys, GI tract, and perspiration (generally, loss of Ca2+ in sweat is insignificant). That is, the organism is in positive Ca2+ balance. Conversely, in the elderly (see Chap. 64), total body Ca2+ decreases, principally because of loss from the skeleton (i.e., Ca2+ balance is negative). During several decades of adult life, however, the Ca2+ homeostatic system precisely balances intake and output of Ca2+ to maintain a constant total body level (Fig. 49-1) while at the same time maintaining near constancy of Ca2+o. Of the 1000 mg of elemental Ca2+ ingested daily by this hypothetical individual, ~30% or 300 mg is actually absorbed by the intestine. Some 100 mg of Ca2+ is lost into the gut lumen by intestinal secretion daily, so that net absorption is 200 mg. Approximately 500 mg of Ca2+ enters and leaves the skeleton daily through the processes of bone formation and resorption, respectively. To maintain Ca2+ balance, 200 mg of Ca2+ are lost in the urine. The latter comprises only 2% of the daily filtered load of Ca2+ (10 g), illustrating the remarkable efficiency of the kidney in reabsorbing Ca2+ and its potential for modifying Ca2+ balance via changes in net Ca2+ reabsorption.

FIGURE 49-1. Overall Ca2+ balance in a normal individual. Of the 1 g of elemental Ca2+ ingested, net absorption is 200 mg (300 mg true absorption, 100 mg endogenous fecal secretion). Balance is achieved by renal excretion of 200 mg of Ca2+ because equivalent amounts of Ca2+ are laid down and removed from the skeleton on a daily basis in this individual. (ECF, extracellular fluid.) (Adapted from Brown EM, LeBoff MS. Pathophysiology of hyperparathyroidism. In: Rothmund M, Wells SA Jr, eds. Progress in surgery, vol 18. Parathyroid surgery. Basel: Karger, 1986:13.)

Because of the increasing recognition of osteoporosis as a major public health problem for individuals in later life, a great deal of interest has been shown in Ca2+ nutrition as a function of age, hormonal status, and other factors (see Chap. 64). The recommended daily allowance (RDA) in the United States had been 800 mg. However, although this level of intake may be sufficient to maintain some young adults in zero or positive balance, it may be insufficient for growing children and can lead to a negative balance in older people. In particular, a mean calcium intake of ~1000 mg prevents negative balance in perimeno-pausal, estrogen-replete women, whereas nearly 1500 mg is necessary to achieve zero balance in estrogen-deficient women of this age.10 Indeed, the RDAs for various groups of individuals have been modified to the following values: 800 mg until age 10; 1200 mg in adolescence; and 1000 mg thereafter, increasing to 1200 mg during pregnancy and lactation and to 1500 mg if at increased risk of osteoporosis or >65 years of age.11 Intake of calcium in these quantities should not be considered as a supplementation but rather as a nutritional requirement for skeletal health. Interestingly, the RDA for humans is nearly five-fold lower than that for other species of animals ranging in size from 1 to 800 kg.10 The form of calcium ingested has some impact on the efficiency with which it is absorbed. A high phosphate/calcium ratio may have deleterious skeletal consequences, because increases in serum phosphate can lower Ca2+o concentration and promote secondary hyperparathyroidism. Nevertheless, the relatively high content of phosphorus in dairy products does not appear to reduce substantially the absorption of Ca2+from this source. Calcium carbonate is a commonly used dietary Ca2+ supplement. A pH below ~5 is required for solubilization of calcium carbonate in the stomach.12 Therefore, in the presence of achlorhydria—a common condition in the aging population—absorption of the carbonate salt of calcium can be reduced. In this situation, more soluble forms of Ca2+, such as calcium citrate, are preferable.13 Certain dietary sugars, such as lactose, enhance the absorption of ingested Ca2+ by a mechanism that is independent of vitamin D. This potential beneficial effect of lactose is often offset by lactose intolerance in older patients. The treatment of milk with lactase largely overcomes this problem in the lactose-intolerant patient. Conversely, phytate, which is present in some cereal grains, can bind Ca2+ in the intestinal lumen and impair absorption.

HORMONAL CONTROL OF CALCIUM HOMEOSTASIS A complex homeostatic system has evolved that is designed to maintain near constancy of Ca2+o through calciotropic hormone, and direct, Ca2+o-induced alterations in the GI, renal, and/or skeletal handling of Ca2+ (Fig. 49-2).4,5,6 and 7 Usually, this is accomplished in such a way that extracellular (including skeletal) and cellular Ca2+ homeostasis and balance are maintained simultaneously. Conditions of gross Ca2+ excess, however, may induce positive Ca2+ balance (e.g., deposition of Ca2+ in bone and soft tissues) to maintain a normal or near-normal level of Ca2+o.Conversely, severe Ca2+ deficiency leads to mobilization of skeletal Ca2+ to preserve normocalcemia. Overall Ca2+o homeostasismay be understood in terms of the following general principles: (a) The first priority of the homeostatic system is to maintain a normal level of Ca2+o; (b) with moderate stresses on the system, intestinal and renal adaptations are usually sufficient to sus- tain Ca2+o homeostasis without changes in net bone mass; and (c) with severe hypocalcemic stresses, skeletal stores of Ca2+ are mobilized to maintain Ca2+o homeostasis, potentially compromising the structural integrity of the skeleton.

FIGURE 49-2. The hormonal control of calcium and phosphate homeostasis by parathyroid hormone (PTH) and vitamin D (solid lines and arrows). Ca2+ regulates PTH secretion in an inverse fashion. In turn, PTH affects renal handling of calcium and phosphate as well as renal synthesis of 1,25(OH)2D3. PTH and 1,25(OH)2D3 synergistically mobilize calcium and phosphate from bone, whereas the latter increases intestinal absorption of both ions. Also shown are direct actions of calcium and phosphate ions on tissues involved in maintaining mineral ion homeostasis (lines and arrows), illustrating the role of these ions as extracellular, first messengers. For example, elevated Ca2+o directly suppresses renal synthesis of 1,25(OH)2D3, tubular reabsorption of Ca2+ in the thick ascending limb, and the resorptive activities of osteoclasts. Thus, calcium ions modulate their own homeostasis not only through direct actions on the secretion of calciotropic hormones but also by exerting direct effects on target tissues that tend to lower Ca2+o. In effect, calcium and phosphate ions act as Ca2+o-regulating factors or “hormones” because they transmit information about the state of mineral ion metabolism from one part of the homeostatic system to other parts. For additional details, see text. (Reproduced from Brown EM, Pollak M, Hebert SC. Cloning and characterization of extracellular Ca2+-sensing receptors from parathyroid and kidney. Molecular physiology and pathophysiology of Ca2+-sensing. The Endocrinologist 1994; 4:419.)

The homeostatic system comprises two essential components. The first is several cell types that sense changes in Ca2+o and respond with appropriate alterations in

their output of Ca2+o-regulating hormones (PTH, calcitonin, and 1,25[OH]2D3). The parathyroid glands are key sensors of variations in Ca2+o, responding withalterations in PTH secretion that are inversely related to ambient levels of Ca2+o.7 Parathyroid cells recognize changes in Ca2+o through a cell-surface Ca2+o-sensing receptor14—described in more detail later—that is a member of the superfamily of G protein– coupled receptors that includes the receptors for PTH and calcitonin. The C cells of the thyroid gland also respond to changes in Ca2+o with alterations in calcitonin secretion—with high Ca2+o stimulating calcitonin secretion—through the same Ca2+o-sensing receptor. The physiologic relevance of changes in circulating levels of calcitonin in adult humans, however, remains uncertain (see Chap. 53). Finally, the production of 1,25(OH)2D3 is likewisedirectly regulated by Ca2+o, with hypocalcemia stimulating its synthesis in the renal proximal tubule and hypercalcemia inhibiting it. 15 The second key component of the homeostatic system is the effector cells that control the renal, intestinal, and skeletal handling of Ca2+. The translocation of calcium and phosphate ions into or out of the extracellular fluid by these tissues is regulated by PTH, calcitonin, and 1,25(OH)2D3 as well as by mineral ions themselves. The overall operation of the homeostatic system is as follows (see Fig. 49-2). In response to slight decrements in the Ca 2+o, forexample, a prompt increase occurs in the secretory rate for PTH. This hormone has several important effects on the kidney, which include inducing phosphaturia, enhancing distal tubular reabsorption of Ca2+, and increasing generation of 1,25(OH)2D3 from 25-hydroxyvitamin D3, or 25(OH)D3.16 The increased circulating levels of this potent metabolite of vitamin D3 directly stimulate absorption of calcium and phosphate by the intestine through independent transport systems.4,5 and 6,17 PTH and 1,25(OH)2D3 also synergistically enhance the net release of calcium and phosphate from bone. The increased movement of Ca2+ into the extracellular fluid from intestine and bone, coupled with PTH-induced renal retention of this ion, normalize circulating levels of Ca2+o, thereby inhibiting PTH release and closing the negative-feedback loop. In addition, 1,25(OH)2D3, formed in response to the action of PTH on the kidney, directly inhibits the synthesis and secretion of PTH,7 contributing to the negative feedback control of parathyroid function by the homeostatic system. Excess phosphate mobilized from bone and intestine is excreted in the urine via the phosphaturic action of PTH. Both calcium and phosphate ions can exert direct effects on various cells and tissues involved in mineral ion metabolism (see Fig. 49-2). For instance, calcium ions not only reduce PTH secretion but also directly inhibit proximal tubular synthesis of15 stimulate the function of osteoblasts,18 and inhibit 1,25(OH)2D3, the function of osteoclasts.19 Phosphate ions reduce 1a-hydroxylation of vitamin D, stimulate bone formation, inhibit bone resorption,4,5,6 and 7 and stimulate various aspects of parathyroid function,20 as described in more detail in the next section. These actions play important roles in mineral ion homeostasis by enabling both the hormone-secreting and effector elements of the homeostatic system to sense changes in the ambient concentrations of mineral ions and to respond in a physiologically appropriate manner. Indeed, by virtue of their direct actions on cells involved in mineral ion metabolism, calcium and phosphate ions can be viewed as serving a hormone-like role as extracellular first messengers.7 Although the cloning of the Ca2+o-sensing receptorthat mediates direct actions of Ca2+o on parathyroid and renalfunction has clarified substantially how cells sense Ca2+o,21 the mechanism underlying phosphate sensing remains obscure. REGULATION OF PARATHYROID HORMONE SECRETION BY CALCIUM AND OTHER FACTORS A steep, inverse sigmoidal relationship exists between PTH secretion and Ca2+o (Fig. 49-3)—hence, the exquisite sensitivity ofthe parathyroid gland to Ca2+o that is critical to maintaining stable normocalcemia.7 This sigmoidal function can be described in terms of maximal and minimal secretory rates, midpoint or set-point, and slope at the midpoint (see Fig. 49-3). The maximal secretory rate provides a measure of the acute secretory reserve of the gland in response to a maximal hypocalcemic stress. Maximal secretory capacity increases with parathyroid cellular hyper-plasia, as in primary or secondary hyperparathyroidism.4,5,6 and 7,21 The secretion of PTH from the parathyroid cell apparently is incapable of being totally suppressed, even at very high levels of Ca2+o(see Fig. 49-3). Importantly, this implies that a sufficient mass of even normal parathyroid tissue could cause hypersecretion of PTH with resultant hypercalcemia. This has, in fact, been documented in rats by transplanting sufficient numbers of normal rat parathyroid glands to cause hypercalcemia in the recipient animals.22 This may, in part, contribute to the hypersecretion of PTH in hypercalcemic hyperparathyroidism The setpoint of normal human parathyroid cells in vitro is ~1.0 mmol/L of ionized Ca2+and is close to the ambient level of Ca2+o in humans (1.1–1.3mmol/L).7 This parameter is important in determining the level at which the Ca2+o is set by the homeostatic system (the setpointfor Ca2+o is slightly higher than that for the parathyroid cell per se, also being a function of the tissues that are targets for the actions of PTH). In primary hyperparathyroidism, the setpoint of pathologic parathyroid cells is increased to 1.2 to 1.4 mmol/L Ca2+, or sometimes higher, and the serum Ca 2+ concentration is correspondingly reset to an elevated level7,21 (see Chap. 58). The steep slope of the relationship between PTH release and Ca2+oensures a large change in secretory rate for a small change in Ca2+o and is important in maintaining the serum Ca2+ concentra- tion within a narrow range. In addition to its sensitivity to Ca2+o,the parathyroid cell alters its secretory rate within seconds to a perturbation in Ca2+o and may also be responsive to other variables, such as the rate and direction of the change.23,24

FIGURE 49-3. A, Relationship between parathyroid hormone (PTH) secretion and extracellular Ca2+ concentration in normal human parathyroid cells. Dispersed parathyroid cells were prepared and incubated with the indicated ionized Ca2+ concentrations. The PTH released was determined using a radioimmunoassay that recognizes the intact hormone and the COOH-terminal fragments of the molecule. B, The four parameters that can be used to describe the relationship between PTH secretion and the extracellular ionized Ca2+ concentration: maximal and minimal secretory rates at low and high Ca2+ concentrations, respectively; slope of the curve at its midpoint; and setpoint (the level of Ca2+ producing half of the maximal inhibition of PTH release). (A from Brown EM. Regulation of the synthesis, metabolism and actions of parathyroid hormones. In: Brenner BM, Stein H, eds. Contemporary issues in nephrology, vol 2. Divalent ion homeostasis. New York: Churchill-Livingstone, 1983.)

The technique of expression cloning in Xenopus laevis oocytes enabled isolation of a phosphoinositide-coupled, Ca2+o-sensing receptor from bovine parathyroid gland through which parathyroid, kidney, and other cells sense Ca2+o.14 This receptor has a large amino-terminal extracellular domain that is involved in binding Ca2+o, followed by seven membrane-spanning segments characteristic of the superfamily of G protein–coupled receptors, and a cytoplasmic carboxy terminus (Fig. 49-4). In addition to recognizing Ca2+o, the receptor responds to other polyvalent cations as well, including Mg2+, trivalent cations, and neomycin.14 The possibility exists, therefore, that the Ca 2+o-sensing receptor can function as a Mg2+o-sensing receptor, and that some of the directeffects of Ca2+o as well as the toxic actions of aminoglycosides on the kidney are mediated through it.21 The receptor plays a key role in Ca2+o sensing by the parathyroid and kidney because the human homolog of the receptor harbors inactivating or activating mutations in several human genetic diseases that manifest abnormal Ca2+o sensing.21 In familial hypocalciuric hypercalcemia (FHH),25 individuals with heterozygous inactivating mutations of the Ca2+o-sensing receptor have mild to moderate hypercalcemia with inappropriately normal (e.g., nonsuppressed) circulating levels of PTH and normal or even frankly low levels of urinary calcium excretion. In contrast, persons with homozygous FHH, which presents clinically as neonatal severe hyperparathyroidism (NSHPT), have severe hypercalcemia with marked hyperparathyroidism.25 (NSHPT can also be caused by compound heterozygous inactivating mutations of the Ca2+o-sensing receptor [i.e., patients harbor a different mutation in each allele of the receptor26] or by the presence of heterozygous inactivating mutations that exert a dominant negative effect on the wild-type receptor.27) Finally, some families with an autosomal dominant or, occasionally, a sporadic form of hypocalcemia accompanied by relative hypercalciuria (e.g., inappropriately high for the serum calcium concentration) harbor activating mutations in this receptor, a finding that provides further evidence of its importance in Ca2+o sensing in the parathyroid and kidney.28

FIGURE 49-4. Proposed structural model of the predicted bovine parathyroid Ca2+-sensing receptor protein. The large amino-terminal domain is located extracellularly and contains nine consensus N-glycosylation sites that are shown as branched chains. Amino acids of the deduced protein are shown at intervals of 50 in the amino-terminal domain. The amino-acid segments that comprise each of the seven predicted membrane-spanning helices are numbered (M1–M7). Potential protein kinase C (PKC) phosphorylation sites are also shown. Each symbol (circle or triangle) represents an individual amino acid; those that are conserved with the metabotropic glutamate receptors are shown as solid symbols, of which acidic residues are indicated by solid triangles and all others by solid circles. Also indicated are multiple cysteine residues conserved with the metabotropic glutamate receptors (which may be involved in stabilizing the large extracellular domain) as well as clusters of acidic residues in the extracellular domain (open and solid triangles) that could be involved in binding of Ca2+ and other polyvalent cations. (SP, signal peptide; HS, hydrophobic segment.) (From Brown EM, Gamba G, Riccardi D, et al. Cloning and characterization of an extracellular Ca2+-sensing receptor from bovine parathyroid. Nature 1993; 366:575.)

In addition to being present in parathyroid, the Ca2+o-sensing receptor is also present along most of the renal tubule29; it is present at the highest levels in the cortical thick ascending limb of the nephron (a segment that is likely responsible for the abnormal renal handling of Ca2+ in FHH).25 Receptor transcripts are also found in thyroid, brain, and several other tissues, including intestine and bone, as described later.30 The presence of the receptor in kidney probably accounts for several of the long-recognized but poorly understood direct effects of Ca2+o on renal function (such as the reduced urinary concentrating ability encountered in some hypercalcemic patients).31 In the thyroid, the Ca2+o-sensing receptor resides almost exclusively in the C cells and is thought to mediate the stimulatory effect of elevated Ca2+o on calcitonin secretion. Ca2+o-sensing receptors in the brain might mediate as yet unknown actions of Ca2+o on neuronal function.30 Interestingly, among the G protein–coupled receptors, the Ca2+o-sensing receptor exhibits only limited homology with three subclasses of receptors sharing its overall topology, particularly the presence of a very large amino-terminal extracellular domain. These are the metabotropic glutamate receptors (mGluRs), targets for glutamate, the major excitatory neurotransmitter in the brain; the GABAB receptor, which is activated by g-aminobutyric acid (GABA), the major inhibitory neurotransmitter in the brain; and a family of putative pheromone receptors.30 All three classes of receptor presumably arose from a common ancestral gene. Further understanding of their structural similarities may clarify previously unknown relationships between Ca2+o homeostasis or other functions of Ca2+ in the nervous system and in systemic Ca2+o homeostasis. As with Ca2+, high concentrations of Mg2+o inhibit PTH release via the Ca2+o-sensing receptor, although, on a molar basis, Mg2+o is two to three times less potent than Ca2+o.32 Levels of Mg2+o in the blood sufficiently high to modulate PTH secretion are generally achieved only during pharmacologic interventions, such as the intravenous infusion of Mg2+ for the treatment of preeclampsia. Because of the reabsorption of more NaCl, water, and Ca2+than of Mg2+ in the proximal tubule, however, the level of Mg2+ in the thick ascending limb can rise to approximately 1.6-fold higher than that initially filtered at the glomerulus.33 Therefore, locally higher Mg2+ concentrations could interact with Ca2+o-sensing receptors in the thick ascending limb, a major site of hormonally regulated (i.e., PTH-regulated) Mg2+ and Ca2+ reabsorption. Abnormally low Mg2+o is also associated with a reduced rate of PTH release due to an unknown mechanism, which can cause hypocalcemia.34 This biphasic secretory response to Mg2+o differs from that to Ca2+o because near-maximal rates of PTH release can be maintained at very low levels of Ca2+o. Various other factors, including metabolites of vitamin D, a variety of catecholamines and other biogenic amines, prosta-glandins and peptide hormones, phosphate, and monovalent cations such as potassium and lithium also modify PTH release.7,35 Of these, the most physiologically relevant are the active vitamin D metabolite 1,25(OH)2D3 and changes in the serum phosphate concentration. Studies have documented that 1,25(OH)2D3 plays an important role in the longer-term (over days or longer) regulation of parathyroid function, tonically inhibiting PTH secretion,36,37 PTH gene expression,38,39 and probably parathyroid cellular proliferation.40 Therefore, the actions of PTH on its target tissues produce negative feedback regulation of parathyroid cellular function not only by increasing the serum Ca2+ concentration but also by stimulating the formation of 1,25(OH)2D3, which exerts direct negative feedback actions on the parathyroid. Elevations and reductions in serum phosphate concentrations increase and decrease, respectively, PTH gene expression, PTH secretion (perhaps indirectly through changes in the expression of its gene), and parathyroid cellular proliferation.20 The use of lithium in the treatment of manic-depressive illness may induce hypercalcemia and hypersecretion of PTH.41 These findings can mimic the biochemical abnormalities of primary hyperparathyroidism or FHH (e.g., lithium treatment can also be accompanied by hypocalciuria and a tendency toward hypermagnesemia). Indeed, of patients being treated long term with lithium, some 10% to 15% develop overt hyperparathyroidism, although whether this results from a direct action of lithium to induce the disorder or from unmasking of preexistent mild primary hyperparathyroidism is not entirely clear. In vitro, lithium produces a rightward shift in the setpoint of bovine parathyroid cells for Ca2+o42 (e.g., a change similar to that seen in parathyroid adenomas), possibly due to interference with the process of Ca2+o sensing. INTRACELLULAR MECHANISMS REGULATING PARATHYROID HORMONE RELEASE After its biosynthesis as preproPTH, its subsequent metabolism to proPTH, and then its metabolism to PTH, the hormone is stored in secretory granules before its release into the circulation43 (see Chap. 51). The regulation of hormonal secretion involves the transduction of information about changes in Ca2+o by the Ca2+o-sensing receptor into alterations in the levels of one or more intracellular mediators that ultimately modify the secretory process.7,14,43 The definitive identification of the relevant mediators, however, remains elusive. Although cyclic adenosine monophosphate (cAMP) mediates the actions of potent secretagogues, such as b-adrenergic catecholamines, on PTH release, it cannot account quantitatively for the effects of Ca2+o on hormonal secretion.7 Increases in Ca2+o produce an initial transient rise in Cai,44 most likely because of Ca2+-receptor– mediated hydrolysis of polyphosphoinositides (PIs). The resultant production of inositol 1,4,5-trisphosphate14,45 stimulates the release of Ca2+ from intracellular stores. A more sustained increase in Cai then ensues,44 owing to uptake of extracellular Ca2+ through plasma membrane channels for which the properties and mode of regulation remain to be fully defined. Activation of phospholipase C and the resultant hydrolysis of polyphosphoinositides in response to elevations in Ca2+o likelyplays a key role in the concomitant inhibition of PTH secretion and of parathyroid cellular proliferation. However, the identification of the most relevant downstream mediator(s) of these inhibitory effects on parathyroid function remain elusive. Patients with homozygous FHH have dramatic hypersecretion of PTH due to severe primary hyperparathyroidism with a substantial shift to the right in setpoint as well as marked glandular enlargement and parathyroid cellular hyperplasia.46 This experiment in nature shows, therefore, that the Ca2+o-sensing receptor contributes to the tonic inhibition of parathyroid cellular proliferation under normal circumstances. EFFECTS OF PARATHYROID HORMONE ON CALCIUM-REGULATING TISSUES CONTROL OF GASTROINTESTINAL CALCIUM ABSORPTION The net amount of Ca2+ absorbed from the GI tract is the difference between the total mass of Ca 2+ moving from lumen to plasma (absorption) and plasma to lumen.47 The latter, termed endogenous fecal calcium, results from the secretion by the intestinal mucosa of ~100 mg per day of Ca2+; it varies little in different states of Ca 2+ balance. Ca2+ absorption results from both passive diffusion across the intestinal mucosa and active transport. The passive diffusion of Ca2+ is a concentration-dependent, nonsaturable process that accounts for absorption of 10% to 15% of the Ca2+ in the normal diet (i.e., 100–150 mg per day of ingested Ca2+). The active component of Ca2+ absorption is a saturable, carrier-mediated mechanism46a,46b regulated by 1,25(OH)2D3. The highest density of sites of active Ca2+ absorption is in the duodenum.47,48 Clearly, however, vitamin D-responsive Ca2+ absorption occurs in more distal portions of the bowel as well, including the small intestine (ileum more than jejunum) and portions of the colon. Because these segments of the GI tract are anatomically much longer than the duodenum, they may

contribute significantly to overall Ca2+ absorption. After 1,25(OH)2D3 is administered to vitamin D–deficient animals, an increase occurs in the GI absorption of Ca2+ over several hours, which generally is paralleled by the induction in intestinal mucosal cells of several vitamin D–dependent proteins, including a Ca2+-binding protein (calbindin D9K), alkaline phosphatase, and Ca2+-Mg2+-adenosine triphosphatase.47,48 These proteins may be involved in the mechanism by which vitamin D enhances Ca2+ absorption. Probably, Ca2+ ions in the intestinal lumen move down their electrochemical gradient across the brush border of the epithelium via an uptake channel46a,46b and are pumped out of the basolateral aspect of the cell on their way to the extracellular fluid. The metabolite 1,25(OH)2D3 appears to stimulate both the influx and the egress of Ca2+ from the intestinal epithelial cells.49,50 Most phosphate absorption from the intestine occurs in the small bowel through a vitamin D–responsive mechanism distinct from the Ca2+mechanism.4,5 and 6,47 Even in vitamin D deficiency, however, approximately half of dietary phosphorus is absorbed. The less stringent regulation of phosphate absorption in the gut is consonant with the ubiquity of this ion in the diet and the looser control of the serum phosphate concentration. An important aspect of the Ca2+o homeostatic system is itscapacity to adapt the efficiency of Ca2+ absorption to dietary intake. When patients are placed on a low Ca2+ diet, serum levels of 1,25(OH)2D3 increase by 50% within 24 to 48 hours, whereas when they are exposed to a high Ca2+ diet this metabo-lite decreases by 50% over this period.51 In experimental animals, the increase in 1,25(OH)2D3 levels on a low Ca2+ diet is largely abolished by prior parathyroidectomy,52 a finding which suggests that dietary Ca2+-induced changes in 1,25(OH)2D3 concentration arise from changes in serum Ca2+ concentration that alter vitamin D metabolism through alterations in the rate of PTH secretion. Nevertheless, lowering and raising the level of Ca2+o also directly stimulate or inhibit, respectively, the 1-hydroxylation of 25(OH)D3.15 The latter effects of Ca2+o could potentially be mediated by the Ca2+o-sensing receptor presentin the renal proximal tubular cells (see later).29 The Ca2+o-sensing receptor is also expressed along the entire GI tract, but whether it directly regulates the absorption of mineral ions is not known.30 Because of the Ca2+o- and PTH-evoked, 1,25(OH)2D3 mediated adaptation in the efficiency of the absorption of Ca2+by the intestine, the absorption of this ion varies less than its content in the diet. Absorption of supplemental Ca2+ may be predominantly through the vitamin D–independent route.53 Phosphate intake also modulates the production of 1,25(OH)2D3, with hypophosphatemia stimulating and hyperphosphatemia inhibiting its renal synthesis.4,5 and 6 CONTROL OF RENAL CALCIUM EXCRETION PTH- and direct Ca2+o-induced changes in renal Ca2+ handling play an important role in overall fine-tuning of Ca2+balance54,55,55b (see Chap. 206). Conversely, vitamin D and its metabolites have only minor direct effects. Of the approximately 10 g of Ca2+ filtered daily by the kidney, 65% is reabsorbed in the proximal tubule.55 Ca2+ reabsorption in this site is closely linked to bulk solute and water transport. PTH has little effect on Ca2+ transport in this nephron segment. In fact, in some studies, PTH inhibits the proximal tubular reabsorp tion of Ca2+, perhaps because the hormone reduces sodium reabsorption.55 Of the more distal portions of the renal tubule, the descending and ascending thin limbs of Henle loop transport little Ca2+.55 Conversely, the thick ascending limb (TAL) of the loop and the distal convoluted tubule (DCT) reabsorb ~20% and 10%, respectively, of the filtered load of Ca2+. In experimental animals, PTH rapidly increases the reabsorption of Ca2+ in both segments of the nephron by increasing transport of the ion from lumen to plasma.5,6,55 It exerts this effect, like its other actions in the kidney and in other tissues, by interacting with a G protein– coupled cell-surface receptor linked to activation of both adenylate cyclase and phospholipase C.56 Several lines of evidence support a primary role for cAMP in mediating PTH-induced changes in renal Ca2+ handling. In microdissected portions of the mammalian renal tubule, PTH-sensitive adenylate cyclase is present in the proximal tubule, cortical portion of the TAL, and portions of the distal tubule (e.g., DCT).57 The location of the enzyme in the proximal tubule is thought to be linked to the well-known PTH-induced phosphaturia. The PTH-activated adenylate cyclase activity in the latter two locations correlates well with the sites of action of the hormone in promoting Ca2+reabsorption. Moreover, the exposure of renal tubules to analogs of cAMP mimics the effects of PTH on Ca2+ transport, further supporting the mediatory role of cAMP. In the cortical thick ascending limb (CTAL), PTH is thought to increase the overall activity of the Na/K/2Cl cotransporter that drives transcellular reabsorption of NaCl in this nephron segment (Fig. 49-5).31,33,55 This increase in transcellular transport is associated with an increase in the lumen-positive, transepithelial potential difference that is responsible for driving ~50% of the reabsorption of NaCl and most of the reabsorption of Ca2+ and Mg2+ in the CTAL. In contrast, raising Ca2+o by activating the Ca2+osensing receptor present in the same epithelial cells of the CTAL reduces overall cotransporter activity, probably by inhibiting both the cotransporter itself as well as the apical potassium channel that recycles potassium ions back into the tubular lumen.31 The transepithelial potential gradient is consequently reduced (see Fig. 49-5), leading to diminished paracellular reabsorption of both Ca2+ and Mg2+. In effect, Ca2+o, acting in a fashion analogous to that of “loop” diuretics (e.g., furosemide), regulates its own reabsorption (and that of Mg2+) by a direct renal action that antagonizes the action of PTH.31

FIGURE 49-5. Diagram showing how the Ca2+o-sensing receptor (CaR) may regulate intracellular second messengers and, in turn, transport of NaCl, K+, Mg2+, and Ca2+ in the cortical thick ascending limb. Hormones stimulating cyclic adenosine monophosphate (cAMP) accumulation, such as parathyroid hormone, increase the reabsorption of Ca2+and Mg2+ through the paracellular pathway by elevating the lumen-positive transepithelial potential, Vte, via stimulation of the activity of the Na/K/2Cl cotransporter and an apical K+ channel. The CaR, also present on the basolateral membrane, increases arachidonic acid (AA) formation by activating phospholipase A2 (PLA2) (2). AA is metabolized via the P450 pathway to produce an active metabolite, probably 20-hydroxyeicosatetraenoic acid, which inhibits the apical K+ channel (4) and, perhaps, the Na/K/2Cl cotransporter (3). Both actions reduce overall cotransporter activity, thereby diminishing Vte and, in turn, paracellular transport of Ca2+ and Mg2+. The CaR probably also inhibits adenylate cyclase (1), thereby decreasing hormoneand cAMP-stimulated divalent cation transport. (Reproduced with permission from Brown EM, Hebert SC. Calcium-receptor regulated parathyroid and renal function. Bone 1997; 20:303.)

Although the detailed cellular mechanism by which PTH regulates Ca2+ transport in the DCT remains incompletely understood, it likely involves a PTH-stimulated increase in the apical uptake of Ca2+ through a channel.46a,55,55a The ensuing transcellular transport of Ca2+ is facilitated by a vitamin D–dependent Ca2+-binding protein, calbindin D28K, which is expressed in this nephron segment and is distinct from that participating in vitamin D–mediated GI absorption of Ca2+.58 Another small amount of Ca2+ (~5% of the filtered load) is reabsorbed in the collecting ducts, but transport at this site is not PTH regulated. Although the Ca2+o -sensing receptor is present in the DCT,29 its role, if any, in regulating tubular reabsorption of Ca2+ is not known. The net action of PTH on renal Ca2+ handling is to reduce the amount of Ca2+ excreted at any given level of serum Ca2+.54 This has been demonstrated by examining renal Ca2+ excretion as a function of serum Ca2+ concentration in subjects with underactive, normal, or overactive parathyroid function (Fig. 49-6).54 In patients with primary hyperparathyroidism, although the total amount of Ca2+ excreted over 24 hours in the urine may be elevated, less Ca2+ is excreted than in a normal person whose serum Ca2+ concentration is equally elevated. Conversely, hypoparathyroid patients have a renal Ca2+ leak, excreting more calcium than normal at a given serum Ca2+ concentration. Therefore, during therapy with vitamin D, the total serum Ca2+concentration of patients with hypoparathyroidism should be maintained in the range of 8 to 9 mg/dL to avoid hypercalciuria (see Chap. 60). The data in Figure 49-6 also illustrate the steep positive relationship between serum and urine Ca2+, which is likely mediated by the Ca2+o-sensing receptor.31 This relationship is reset as a function of the prevailing state of parathyroid function, shifting rightward and leftward with chronic increases and decreases, respectively, in the circulating levels of PTH.54

FIGURE 49-6. Urinary excretion of Ca2+ expressed as a function of serum Ca2+ in normal individuals (area enclosed by the dotted lines, which shows mean ± 2 standard deviations) as well as in hypoparathyroid subjects (open and solid triangles; solid triangles represent basal values) and hyperparathyroid subjects (solid circles). The shaded area is the normal physiologic situation. (GF, glomerular filtrate.) (From Nordin BEC, Peacock M. Role of the kidney in regulation of plasma calcium. Lancet 1969; 2:1280.)

Along with its effects on renal Ca2+ handling in the distal nephron (i.e., CTAL and DCT), PTH also inhibits phosphate reabsorption in both proximal and distal sites and enhances the synthesis of 1,25(OH)2D3 in the proximal tubule.5,6 The first of these effects, like the actions of the hormone on Ca2+ reabsorption, appears to be mediated by cAMP. PTH also activates PI turnover,56 and evidence has implicated this pathway in the PTH mediated stimulation of the synthesis of 1,25(OH)2D359. In addition to regulating renal handling of Ca2+ and Mg2+, the Ca2+o-sensing receptor probably mediates the known action of hypercalcemia to reduce urinary concentrating ability.31 It is thought to do so by two actions: First, by inhibiting NaCl reabsorption in the TAL, it reduces the medullary hypertonicity needed for passive, vasopressin-stimulated reabsorption of water in the collecting ducts. Second, in the inner medullary collecting duct, raising Ca2+o directly reduces vasopressinstimulated water flow, probably by an action mediated by the Ca2+o-sensing receptor on the apical membrane of these cells60. The resultant excretion of Ca2+ in a more dilute urine may reduce the risk of renal stone formation when a Ca2+ load requires disposal during times of antidiuresis.31 In this way, the Ca2+o-sensing receptor may serve to coordinate the homeostatic mechanisms governing Ca2+o and water metabolism. ROLE OF THE SKELETON IN CALCIUM HOMEOSTASIS During the constant remodeling of the skeleton, osteoclastic bone breakdown is closely coupled to osteoblastic bone formation.4,5 and 6,61 The formation of osteoblasts is coupled to the generation of osteoclasts from their precursors. Osteoclastic resorption of bone, in turn, is tightly linked to the ensuing replacement of the resorbed bone by osteoblasts. This constant turnover and renewal of bone (see Chap. 50) plays an important role in maintaining the structural integrity of this tissue. The precision of the coupling between resorption and formation is dramatically illustrated in Paget disease of bone, in which up to 10-fold increases in the rate of skeletal turnover are often unassociated with any alterations in the serum calcium concentration or overall calcium balance (see Chap. 65). Key mechanisms that link the differentiation and function of osteoblasts and osteoclasts are described briefly below and are discussed in more detail in Chap. 50. PTH and other agents stimulating bone resorption (e.g., interleukin-11, prostaglandin E2, and 1,25[OH]2D3) activate osteoclast maturation and function only indirectly by enhancing the expression of osteoclast differentiating factor (ODF) (also sometimes called TRANCE, RANKL,or osteoprotegerin-ligand).62 ODF is expressed on the cell surface of osteoblasts and stromal cells. It activates osteoclast development and increases the activity of mature osteoclasts by interacting with the ODF receptor (a protein called RANK or a closely related protein) on preosteoclasts, which then differentiate into mature osteoclasts if macrophage colony-stimulating factor (M-CSF) is also present.62 Osteoclastic bone resorption, in turn, is coupled to subsequent osteoblastic formation of bone, at least in part through the release of skeletal growth factors, such as transforming growth factor-b and insulin-like growth factor-I, which stimulate osteoblastic differentiation and activity.63 The skeleton also serves as a nearly inexhaustible reservoir for calcium and phosphate ions.5,6 and 7 Because the skeletal content of Ca2+ is 1000-fold higher than that of the extracellular fluid, this function can be subserved by the net movement of relatively small amounts of Ca2+ into or out of bone. After administration of PTH to animals, the structure of osteoclasts, osteoblasts, and osteocytes (those bone cells trapped within the calcified matrix) is altered within minutes.5 Those morphologic changes are accompanied by increased activity of osteoclasts and inhibition of osteoblastic function, leading to an increase in net skeletal calcium release within 2 to 3 hours. The PTH-induced increase in the size of the periosteocytic lacunae also has been considered presumptive evidence for a role for this cell type in skeletal calcium release. Continued exposure to PTH causes an increase in the number and activity of osteoclasts that is ultimately accompanied by an increase in osteoblastic activity through the coupling of bone resorption to formation, as noted previously. The mechanisms by which PTH modulates bone cell function remain to be fully elucidated. The hormone raises skeletal levels of cAMP64 but may also act through other second messenger systems, including the activation of phospholipase C.56 In addition to modulating bone turnover indirectly, by altering the rate of PTH secretion and/or 1,25(OH)2D3 production, changes in Ca2+o also directly regulate bone cell function in vitro in ways that probably contribute to the control of bone turnover in vivo. Raising Ca2+o stimulates several aspects of the functions of osteoblasts and preosteoblasts, such as enhancing their proliferation and chemotaxis.18 These actions could contribute to the coupling of osteoclastic bone resorption to the subsequent replacement of the missing bone by osteoblasts. Conversely, elevated levels of Ca2+o directly inhibit osteoclastic function,19 possibly providing a way for osteoclasts to autoregulate their activity as a function of the amount of calcium that they have resorbed. Pharmacologic evidence has implicated distinct mechanisms for Ca2+o sensing in osteoblasts and osteoclasts,18,19 which differ in turn from the Ca2+o-sensing receptor that regulates various aspects of parathyroid and renal function.7,14 The latter receptor, however, can be expressed by both osteoblastic65 and osteoclastic cells,66 and further studies are needed to establish with certainty the molecular mechanisms through which bone cells sense Ca2+o. Changes in the level of extracellular phosphate also modulate bone turnover as noted previously. Elevations in phosphate stimulate bone formation and inhibit bone resorption, whereas hypophosphatemia produces the converse effects.7 As with the known direct actions of phosphate on parathyroid function, the mechanisms underlying the sensing of extracellular phosphate ions by bone cells remain obscure. REGULATION OF CIRCULATING CALCIUM CONCENTRATION The information in Figure 49-2 does not delineate the temporal sequence of how the homeostatic system reacts to perturbations in Ca2+o. Actually, there is a hierarchy of responses by both the parathyroid gland and the effector systems that regulate calcium transport in the skeleton, kidney, and intestine.7 The initial alteration in the secretory rate of PTH (release of preformed stores of hormone) occurs within seconds of the change in Ca2+o. Within 15 to 30 minutes, an increase occurs in the net synthesis of PTH, without any change in the levels of messenger RNA (mRNA), because of reduced intracellular degradation of PTH. If the hypocalcemic stimulus persists, the levels of PTH mRNA increase over the ensuing 12 to 24 hours or more.38,39 Eventually, chronic hypocalcemia is associated with enhanced parathyroid cellular proliferation over days to weeks,67 which further increases the secretory rate of the hormone. Reduced levels of 1,25(OH)2D3 are also a likely stimulus for parathyroid cellular proliferation and contribute to the severe hyperparathyroidism that can be encountered in chronic renal insufficiency. The most rapid changes in the handling of calcium by the effector organs of the homeostatic system occur in the skeleton and kidney. Alterations in distal tubular calcium reabsorption take place within minutes in vitro,55 whereas the skeletal release of calcium occurs within 2 to 3 hours.68 If these two functional changes are insufficient to restore normocalcemia, the continued hypersecretion of PTH stimulates increased synthesis of 1,25(OH)2D3 within 1 to 2 days,51 enhances the activity of existing osteoclasts, and promotes the appearance of new osteoclasts within days to weeks. Only rarely (e.g., in severe vitamin D deficiency) is the full complement of homeostatic responses insufficient to restore normocalcemia. Exposure of the homeostatic system to a calcium load produces responses that are largely the opposite of those seen with hypocalcemia. The suppression of parathyroid function induces a renal leak of Ca2+, reduces the release of skeletal Ca2+, and ultimately suppresses GI absorption of Ca2+ by inhibiting the synthesis of 1,25(OH)2D3. The remarkable sensitivity of the system is illustrated by its response to the ingestion of the Ca2+ present in a glass of milk. A minute rise in serum Ca2+ causes approximately a 30% reduction in PTH secretion, which leads to prompt excretion of much of the extra Ca2+ in the urine.5 In response to severe loads of Ca2+, the skeleton can buffer substantial quantities of Ca2+, and the kidney can excrete up to 1 g of Ca2+ over 24 hours; however, hypercalcemia may ensue, particularly if

renal impairment develops. One of the most elegant features of the homeostatic system for Ca2+ is that it simultaneously contributes to the regulation of the serum phosphate concentration. With Ca2+ deficiency sufficient to produce secondary hyperparathyroidism, the excess phosphate mobilized into the extracellular fluid from intestine and bone is excreted in the urine (mild hypophosphatemia may ensue). Conversely, with oral Ca2+ loading, reduced GI and skeletal availability of phosphate are managed by decreased renal phosphate clearance because of reduced PTH secretion. Primary abnormalities in phosphate metabolism also elicit changes in the Ca2+o homeostatic system that tend to correct both serum phosphate and calcium concentrations. Hypophosphatemia, for example, enhances the synthesis of 1,25(OH)2D3, which then increases intestinal absorption and skeletal release of phosphate and calcium52 (Fig. 49-7). The increased movement of Ca2+ into the extracellular fluid suppresses PTH release, thereby enhancing the excretion of the excess Ca2+ as well as retaining the phosphate mobilized from intestine and bone. The net result of these adaptations is some normalization of serum phosphate without any change in the serum Ca2+ concentration. Through poorly defined mechanisms, the kidney also retains phosphate more avidly in states of phosphate depletion, independent of alterations in PTH secretion. Clearly, as with the actions of phosphate on parathyroid and bone cells, some mechanism must exist through which these alterations in extracellular phosphate availability and/or levels are recognized by the kidney and transduced into alterations in 1,25(OH)2D3 synthesis and tubular phosphate reabsorption. Moreover, additional hormones regulating phosphate homeostasis likely will be discovered that are involved in the pathogenesis of inherited and acquired disorders of the regulation of the serum phosphate concentration.69

FIGURE 49-7. Response of the homeostatic system to hypophosphatemia. Hypophosphatemia per se stimulates synthesis of 1,25(OH)2D3. Excess Ca2+ mobilized with phosphate from bone and intestine suppresses parathyroid hormone (PTH) release, facilitating calciuresis. Decreased renal phosphate clearance resulting from reduced circulating levels of PTH retains phosphate available from intestine and bone. Low phosphate also may directly inhibit PTH release, increase bone resorption, and enhance renal phosphate reabsorption. (ECF, extracellular fluid.)

CLINICAL ASSESSMENT OF CALCIUM HOMEOSTASIS The clinical evaluation of normal and abnormal calcium homeostasis requires an accurate assessment of serum calcium and phosphate concentrations as well as the functions of the parathyroid glands and the effector systems regulating calcium homeostasis. Total serum calcium and phosphate measurements are usually reliable. Both dip-type and flow-through electrodes of improved quality are widely available for measuring serum ionized calcium and may be useful when changes in serum protein levels make the accurate estimation of ionized from total calcium levels difficult. Measurement of the serum ionized calcium concentration may also be helpful in cases of mild primary hyperparathyroidism, in which an elevation of serum ionized calcium is sometimes more readily detectable than one of serum total calcium concentration. DIRECT ASSAYS OF CALCIUM-REGULATING HORMONES IN BLOOD PARATHYROID HORMONE Most of the immunoreactive PTH in the circulation represents inactive fragments of the hormone comprising the mid-molecule and carboxy-terminal portions of the molecule70 (see Chap. 51). The amount of biologically active PTH (1-84) in the blood represents C) in exon II, resulting in the substitution of arginine (CGT) for cysteine (TGT) in the signal peptide.28 This places a charged amino acid in the hydrophobic core of the signal peptide, leading to inefficient processing of the mutant preproPTH to PTH. In a family with autosomal-recessive isolated hypoparathyroidism, a donor splice mutation was identified in the PTH gene of affected individuals at the exon II-intron II boundary that resulted in the loss of exon II, which encodes the initiation codon and signal peptide.29 A search for evidence of ectopic PTH synthesis (i.e., synthesis outside of parathyroid tissue) indicates that it occurs only rarely in malignancies associated with hypercalcemia.30 However, with the advent of highly specific PTH immunoassays and mRNA analysis, a few cases of true ectopic PTH production have been documented. In one case of an ovarian carcinoma, the 5' PTH gene regulatory region that normally silences gene transcription in nonparathyroid cells was replaced by a foreign sequence that allowed inappropriate transcription to take place.31

HORMONE SECRETION GENERAL FEATURES OF HORMONAL SECRETION Relatively little PTH is stored in secretory granules within the parathyroid glands. In the absence of a stimulus for release, intraglandular metabolism occurs, causing complete degradation of the hormone to its constituent amino acids or partial degradation to fragments through a calcium-regulated enzymatic mechanism.32 In cases of hypercalcemia, the predominant hormonal entities released from parathyroid glands appear to be fragments composed of midregion or COOH-terminal sequences or both, containing little or no bioactivity.33,34 In cases of hypocalcemia, degradation of PTH within the parathyroid cell is minimized, and the major hormonal entity released appears similar to or identical with the bioactive PTH 1–84 molecule.35 Consequently, in the presence of hypocalcemia, increased amounts of bioactive PTH are secreted, even in the absence of additional synthesis of hormone. With a brief hypocalcemic stimulus, a biphasic secretory response often occurs. Hormone, presumably newly synthesized and derived from “immature” Golgi vesicles, is initially released in a large burst over a few minutes.2 This is followed by a lower response sustained for a longer period, presumably representing hormone stored in secretory granules. However, hormone stores are insufficient to maintain secretion for more than a few hours in the presence of a sustained severe hypocalcemic stimulus; although other mechanisms—transcriptional and posttranscriptional—increase PTH synthesis to some extent, additional PTH secretion ultimately depends on an increase in the number of parathyroid cells. Such an increase appears to be modulated by the reductions in circulating 1,25-dihydroxyvitamin D that often accompany hypocalcemia. A second protein, identical to chromogranin A from the adrenal medulla, is cosecreted with PTH in most conditions leading to the release of PTH.36,37 This 50-kilodalton (kDa) protein is synthesized within the parathyroid gland and stored with PTH within secretory granules. This molecule, which can be glycosylated and phosphorylated, is a member of the chromogranin-secretogranin (granin) family of proteins that occurs in virtually all neuroendocrine cells.38,39 and 40 The family of proteins plays several roles in the process of regulated secretion, including targeting peptide hormones and neurotransmitters to secretory granules of the regulated secretory pathway. Chromogranin A also functions as a precursor of biologically active peptides that modulate neuroendocrine cell secretion in an autocrine or paracrine fashion (see Chap. 175). MODULATORS OF PARATHYROID GLAND SECRETION The calcium ion is the main regulator of parathyroid gland activity, although several other agents influence the release of PTH from parathyroid glands. These include various ions, agents altering the activity of the parathyroid cell adenylate cyclase system (e.g., b-adrenergic catecholamines, histamine), peptides derived from chromogranin A, and vitamin D metabolites. A circadian rhythm has been reported for PTH secretion, with increased blood levels occurring at night and small amplitude pulses of PTH secretion occurring at much shorter intervals.41,42 These studies may suggest neural or central nervous system influences on PTH secretion, or they could reflect circadian alterations in the levels of extracellular fluid calcium.

IONS Cations. The most potent of the cations modulating PTH release and the secretagogue that is most important in altering PTH release under physiologic and pathophysiologic circumstances is the calcium ion (Fig. 51-3). Although there is an inverse relationship between ambient calcium levels and PTH release, this is a curvilinear rather than proportional relationship.43 From in vivo studies of cattle, maximal rates of PTH secretion of ~16 ng/kg per minute appear to be rapidly achieved at calcium levels below 7.5–8.0 mg/dL (1.88–2.00 mmol/L). When calcium levels are reduced from 10.0 to 9.5 mg/dL (2.50–2.38 mmol/L), a small and gradual increase in PTH secretion occurs that does exhibit a proportional relationship to the ambient calcium level. Half-maximal secretion rates normally occur at calcium levels of ~8.5 mg/dL (2.12 mmol/L), which is the set-point for PTH secretion. Basal secretion rates result after ambient calcium levels have risen above 11 mg/dL (2.75 mmol/L) and appear to persist despite further increases in calcium concentration, even up to 16–18 mg/dL (4.00–4.50 mmol/L). Similar results have been observed in studies with human parathyroid tissue in vitro. The nonsuppressible (more correctly, non–calcium suppressible) component of constitutive PTH secretion appears to comprise mainly bioinactive midregion and COOH-terminal fragments. However, direct determination of the activity of hormonal material released when calcium levels are markedly elevated also demonstrates some bioactivity.

FIGURE 51-3. The sites of calcium regulation for parathyroid hormone (PTH) biosynthesis, intraglandular metabolism, and secretion. Sites known to be influenced by calcium (*) include reversible reduction of preproPTH mRNA levels by elevated extracellular fluid calcium levels acting on preproPTH gene transcription and mRNA stability, increase in the production and release of COOH-terminal fragments by elevated calcium levels, and increase in secretion of the mature form of PTH by reduction in extracellular fluid calcium concentration. Whether calcium acts to alter the splicing of the pre-mRNA is unclear (?).

This inverse relationship between PTH and extracellular calcium contrasts with the influence of the calcium ion as a secre-tagogue in most other secretory systems in which elevations in this ion enhance release of the secretory product. This distinction between the parathyroid cell and other secretory cells is maintained intracellularly, where elevations rather than decreases in cytosol calcium correlate with decreased PTH release.44 Alterations in extracellular fluid calcium levels are transmitted through a parathyroid plasma membrane calcium sensing receptor that couples through a G-protein complex to phospholipase C. Increases in extracellular calcium lead to increases in inositol 1,4,5-trisphosphate (IP3) and mobilization of intracellular calcium stores. The manner in which this inhibits hormone secretion is not understood. Although it would be anticipated that increases in diacylglycerol would accompany IP3 increases caused by hydrolysis of phosphoinositides, activate protein kinase C, and result in reduced PTH secretion, this does not appear to occur in the parathyroid cell.45 Paradoxically, agents that do stimulate protein kinase C, such as phorbol esters, stimulate rather than inhibit hormone secretion. This may be a result of phosphorylation of amino acids in the COOH-terminus of the CaSR, leading to its desensitization by blocking interaction of the receptor with its G protein. The precise steps in the pathway from changes in extracellular calcium levels to hormone release remain to be elucidated. The parathyroid cell Ca2+-sensing receptor cDNA was identified by expression cloning in Xenopus laevis oocytes.46 The mRNA of the human receptor encodes a polypeptide of 1078 amino acids that is predicted to contain a very large extracellular domain of ~600 amino acids and a seven transmembrane-spanning region characteristic of G-protein–coupled cell-surface receptors. Compared with the other known G-protein–coupled receptor family members, the Ca2+-sensing receptor shows some homology with the metabotropic glutamate, g-aminobutyric acid-B, and vomeronasal odorant receptors, sharing conserved cysteine residues and a hydrophobic sequence in the NH2-terminal region. The Ca2+-sensing receptor has a low affinity for Ca2+, and consistent with this, the receptor sequence does not contain any of the Ca2+ binding motifs found in high-affinity calcium-binding proteins. Highly acidic regions in the extracellular NH2-terminal domain and the second extracellular loop may bind calcium as they do in other known low-affinity Ca2+ binding proteins. Besides the parathyroid, the Ca2+-sensing receptor is also expressed in other cells having Ca2+-sensing functions, such as those of the kidney tubule, the calcitonin-secreting thyroid C-cells; and in diverse other organs and tissues such as brain, keratinocytes, gastrointestinal tract, and retina. Neomycin binds the receptor, possibly accounting for the toxic renal effects of aminoglycoside antibiotics. The human Ca2+-sensing receptor gene has been partially characterized and shown to consist of seven exons spanning more than 20 kb of genomic DNA.47 The long extracellular domain is encoded by the first two to six exons, and the remainder of the molecule is encoded by exon seven. Exons 1A and 1B encode two alternative 5' untranslated regions of the CaSR mRNA. Inherited abnormalities of the CaSR gene, located on chromosome 3p13.3-21, can lead either to hypercalcemia or to hypocalcemia depending on whether they are inactivating or activating, respectively. Heterozygous loss-of-function mutations give rise to familial (benign) hypocalciuric hypercalcemia (FHH), in which the lifelong hypercalcemia is asymptomatic.47,48 and 49 The homozygous condition manifests itself as neonatal severe hyperparathyroidism (NSHPT), which is a rare disorder characterized by extreme hypercalcemia and the bony changes of hyperparathyroidism that occur in infancy.50,51 In several cases of NSHPT the parents are normocalcemic, and the cases seem to be sporadic. Autosomal-dominant hypocalcemia (ADH) is caused by gain-of-function mutations in the CaSR gene.52 Autosomal-dominant hypocalcemia may be asymptomatic or present with neonatal or childhood seizures. Because of the overactive CaSR in the nephron, these patients are at greater risk of developing renal complications during vitamin D therapy than are patients with idiopathic hypoparathyroidism. A common polymorphism in the intracellular tail of the CaSR, Ala to Ser at position 986, has a modest effect on the serum calcium concentration in healthy individuals.53 CaSR polymorphisms might also affect urinary calcium excretion and bone mass; therefore, the CaSR is a candidate gene for involvement in disorders such as idiopathic hypercalciuria and osteoporosis. The CaSR is a target for phenylalkylamine compounds— so-called calcimimetics— that are allosteric stimulators of the affinity for cations by CaSR.54 These orally active compounds have been evaluated for treatment of primary and secondary hyperparathyroidism and parathyroid carcinoma.55 The CaSR expressed in the developing parathyroid glands and in the placenta plays an important role in regulating fetal calcium concentrations. Normally, the fetal blood calcium level is elevated above the maternal level. This depends on the action of PTHrP (released from the fetal parathyroids and placenta) on placental calcium transport. Disruption of the CaSR, as shown by studies in CaSR-deficient mice, causes fetal hyperparathyroidism and hypercalcemia as a result of fetal bone resorption.56 The transfer of calcium across the placenta is reduced, and renal calcium excretion is increased. In vitro reductions in extracellular fluid calcium do not appear capable of enhancing specific PTH biosynthesis, which seems to occur at a maximal or near-maximal rate per cell; however, in vivo the reverse is the case, with biosynthesis of PTH normally being set at a low level. In vivo, PTH mRNA levels are markedly stimulated by decreased circulating calcium concentrations. It is thought that the relative insensitivity of parathyroid cell cultures to extracellular calcium is the result of reduced expression of the CaSR.57,58 In vitro and in vivo studies have reported that elevated calcium levels reversibly and specifically reduce PTH mRNA.59 This occurs, in part, by a direct effect on transcription of the pre-proPTH gene60 and, in part, by a posttranscriptional mechanism, whereby hypocalcemia stabilizes and hypercalcemia destabilizes the PTH mRNA.61 Although the results of in vitro studies are conflicting, prolonged hypocalcemia in vivo may stimulate DNA replication, cell division, and the production of increased numbers of parathyroid cells or parathyroid hyperplasia. This would increase the synthesis of proteins, including PTH, within the hypercellular parathyroid gland and ultimately would increase PTH release. In primary parathyroid gland hyperfunction resulting in hyperparathyroidism, alterations in the calcium-sensing mechanism may manifest as a setpoint error, producing a shift to the right of the curve relating PTH secretion to extracellular calcium levels.62 Consequently, elevated concentrations of extracellular fluid calcium may be required to reduce PTH secretion, resulting in an adenomatous or hyperplastic parathyroid gland that is incompletely suppressed by calcium. Such a mechanism may underlie the observation that an increase in the mass of parathyroid tissue, such as that produced by transplantation, can be associated with hypercalcemia. The parathyroid glands of patients with primary and severe uremia secondary to hyperparathyroidism have reduced CaSR expression as assessed by immunostaining.63 Loss of a functional CaSR, as in humans with NSHPT or in mice in which the CaSR gene has been ablated, leads to severe parathyroid hyperplasia.64 If basal secretion per cell produces a significant amount of bioactive PTH, the cumulative increase in this basal or non–calcium-suppressible secretion arising from an increase in parathyroid cells also could be responsible for the hypercalcemia. The precise mechanistic relationship of extracellular calcium to parathyroid cell growth remains to be determined. In studies in vitro, magnesium appears to parallel the effects of calcium on PTH release, although with reduced efficacy.65,66 This is consistent with the known affinity of the parathyroid calcium-sensing receptor for magnesium that is lower than that for calcium itself. Mild hypomagnesemia or hypermagnesemia stimulates or suppresses,

respectively, PTH secretion.67 A special situation exists in clinical disorders associated with severe hypomagnesemia in which PTH secretion is impaired67 (see Chap. 68). With the discovery of the syndrome of aluminum toxicity in uremia, characterized by osteomalacia, low circulating PTH levels, and a tendency toward hypercalcemia during treatment with vitamin D, the effects of aluminum on PTH secretion were assessed in vitro. Such studies, performed with rather high ambient aluminum concentrations, have shown suppression of PTH release by this cation,68 whereas, PTH mRNA expression is unaffected. A direct effect of aluminum on inhibition of PTH secretion in vivo therefore could contribute to the low circulating PTH levels seen in the presence of aluminum intoxication. It is unlikely that aluminum plays any physiologic role in the normal regulation of PTH secretion (see Chap. 131 and Chap. 61). Anions. The phosphate ion is of most interest clinically as a potential modulator of PTH release. Hyperphosphatemia induced by intravenous or oral administration of phosphate is associated with increased circulating levels of PTH. The effects have been thought to be indirect and a result of the hypocalcemia that accompanies the rise in serum phosphate.69 However, it has been suggested that the anion can have a direct action on PTH secretion although the mechanism is unknown.70,71 A parathyroid “phosphate sensor” analogous to the CaSR has yet to be identified. The physiologic role of other anions, such as chloride, that have been associated with alterations of PTH levels in vitro remains unclear. MODULATORS OF ADENYLATE CYCLASE ACTIVITY IN THE PARATHYROID CELL A group of adenylate cyclase–stimulating agonists, such as catecholamines, prostaglandins of the E series, calcitonin, gut hormones of the secretin family, and histamine, influences PTH release in vitro.72,73 and 74 b-Adrenergic agonists (e.g., epinephrine, isoproterenol) also enhance PTH release in animals, and some decreases in circulating PTH levels in humans have been reported with b-blockers. Despite the ability of the b-adrenergic catecholamines and histamine to induce PTH secretion, it is uncertain what physiologic or pathophysiologic role these biogenic amines may have in humans.74,75 Agents that may lower cyclic adenosine monophosphate (cAMP) levels within the parathyroid gland, such as a2-adrenergic agonists, prostaglandin F2, and somatostatin, have been associated with decreased PTH secretion, mainly in vitro. CHROMOGRANIN A–DERIVED PEPTIDES Peptides derived from chromogranin A (CgA), such as b-granin (i.e., CgA 1–113), synthetic b CgA 1–40, pancreastatin (i.e., porcine CgA 240–288), and parastatin (i.e., porcine CgA 347–419), have been reported to inhibit low-calcium–stimulated PTH release from parathyroid cells in culture.76,77 The physiologic significance of these observations remains unclear. Within the parathyroid gland, processing of CgA to peptides occurs to a lesser extent than in some other endocrine tissues, such as the B cells of the pancreas, and the parathyroid cells could respond in a paracrine or endocrine fashion to CgA-derived peptides generated elsewhere. STEROLS AND STEROIDS Various studies have demonstrated a role for vitamin D metabolites in the modulation of PTH release. A feedback loop between PTH-induced increase of vitamin D metabolite levels and vitamin D metabolite–induced decrease of PTH levels has been postulated. A high-affinity 1,25-dihydroxyvitamin D receptor (a member of the nuclear/steroid hormone receptor superfamily of transcriptional regulators) exists in parathyroid cells, and with the localization of injected 1,25-[3H]-dihydroxy-vitamin D within the parathyroid cell, interest in this metabolite as a potential regulator of PTH synthesis or secretion has arisen. Efforts that focused on the role of 1,25-dihydroxyvitamin D in the biosynthesis of PTH provided evidence, in vitro and in vivo, that the sterol reversibly reduces levels of mRNA responsible for PTH synthesis.78,79 This is effected by the action of the sterol on preproPTH gene transcription. A vitamin D response element has been identified in the human PTH gene promoter, although the binding of the vitamin D receptor (VDR) to this element apparently does not require the retinoid X receptor, the normal heterodimerization partner of the VDR.80 The precise mechanism of the down-regulation of PTH gene transcription by 1,25-dihydroxyvitamin D remains to be elucidated. Although 1,25-dihydroxyvitamin D is of uncertain importance in influencing immediate hormone release, it plays a role in modulating hormone synthesis within the gland, altering the quantities of hormone available for immediate release by secretagogues. Moreover, an early in vivo effect of reduction in 1,25-dihydroxyvitamin D appears to be an increase in maximal releasable PTH as a result of stimulation of glandular proliferation.79 In vitro studies show that this effect of the sterol on cell proliferation is mediated by its action on early immediate response gene expression such as the C-MYC protooncogene.81 The role of other metabolites of vitamin D in modulating PTH biosynthesis or secretion remains unclear. However, “nonhyper-calcemic” analogs of 1,25-dihydroxyvitamin D have been developed that do diminish PTH secretion in vitro and in vivo and that may serve in the future as therapeutics for hyperparathyroidism. Cortisol, in some studies, has been reported to elevate extracellular levels of PTH in vivo and in vitro.82 It is unclear, however, whether this effect is a direct effect on synthesis, on alteration of hormonal release, or on an alteration of the levels of enzymes metabolizing PTH within the parathyroid cell.82 Although these findings could help explain the mild secondary hyperparathyroidism occasionally reported with glucocorticoid therapy, their overall significance remains unknown. PERIPHERAL METABOLISM Besides metabolism within the parathyroid cell, PTH 1–84 undergoes extensive peripheral metabolism after release into the general circulation (Fig. 51-4). Although PTH 1–84 can be cleared from the circulation by the kidney and hormonal fragments may be generated in the kidney, such fragments probably do not reenter the circulation. Studies using sequence-specific radioimmunoassays coupled with gel filtration of serum or plasma have shown that the major circulating forms of PTH are the midregion and COOH-terminal fragments.83,84,85,86,87 and 88 Such fragments may be released from the parathyroid gland in the presence of hypercalcemia.

FIGURE 51-4. Model of parathyroid hormone (PTH) metabolism in the presence of normal renal function (NRF) and chronic renal failure (CRF). In the presence of NRF, PTH 1–84 released from the parathyroid gland is cleared by the kidney or metabolized in the liver, where amino (N; NH2 ) and carboxyl (C; COOH) fragments are generated. Carboxyl, but generally not NH2 , fragments, are released into the circulation; these COOH fragments enter a pool, which also includes a contribution from the parathyroid gland, and are cleared by the kidney. In chronic renal failure, PTH secretion can be increased by a tendency toward hypocalcemia, and fewer COOH fragments are released from the parathyroid glands. Failure of the nonfunctioning kidneys to clear PTH results in hepatic metabolism of this moiety, with increased production of NH2 - and COOH fragments. Under these conditions, COOH-terminal fragments reach high circulating concentrations because of increased production and reduced renal clearance.

Another major site of origin of PTH fragments appears to be the liver. Studies using 125I-labeled PTH 1–84 injected intravenously into animals have demonstrated that initial cleavage occurs between residues 33 and 34 and at sites COOH-terminal to this. These and other studies have indicated that the hepatic Kupffer cell is the most likely site of peripheral hormonal cleavage.89 The midregion and COOH-terminal fragments thus generated can enter the general circulation and subsequently be metabolized in the kidney. In view of structure-function studies that have localized the bioactivity of PTH 1–84 to the NH2-terminal third of the molecule, such midregion and COOH-terminal fragments should be biologically inert, at least with respect to calcium homeostatic actions mediated through the PTH/PTHrP receptor. Because of their long half-life in the circulation relative to intact PTH 1–84, such inert moieties generally comprise most circulating PTH-related entities.90 Conversely, the corresponding NH2-terminal fragments generated during hepatic cleavage of PTH 1–84 would be expected to be bioactive, but such fragments seem to be completely degraded in the liver and do

not reenter the circulation. The major circulating bioactive moiety is similar to or identical with the intact molecule PTH 1–84. Clearance of midregion and COOH-terminal fragments of PTH apparently depends almost solely on renal mechanisms. In cases of renal insufficiency, concentrations of midregion and COOH fragments may increase considerably (see Chap. 58 and Chap. 61). By using a sensitive renal cytochemical bioassay for PTH in association with gel filtration fractionation of plasma, it was possible to confirm that the major circulating bioactive moiety in hyperparathyroidism is similar to or identical with PTH 1–84.91 These observations have been confirmed with the development of sensitive immunoradiometric assays that simultaneously recognize NH2 and COOH epitopes on the PTH molecule and therefore detect only intact PTH 1–84. In secondary hyperparathyroidism associated with severe chronic renal failure, additional low-molecular-weight bioactive forms may occasionally be detected.4 The finding of a half-life of disappearance of intact bioactive PTH in end-stage renal disease that cannot be differentiated from normal kinetics indicates that, with the gradual evolution of renal insufficiency, extrarenal sites of degradation assume increasing importance in the clearance of the hormone.4 This conclusion agrees with the finding that the clearance of NH2-terminal immunoreactivity is not decreased in patients with renal failure.90

ACTIONS OF PARATHYROID HORMONE PARATHYROID HORMONE FUNCTIONS The major function of PTH appears to be the maintenance of a normal level of extracellular fluid calcium. The hormone exerts important effects on bone and kidney and indirectly influences the gastrointestinal tract. In response to a fall in the extracellular fluid ionized calcium concentration, PTH is released from the parathyroid cell and acts directly on the kidney to enhance renal calcium reabsorption and promote the conversion of 25-hydroxy-vitamin D to 1,25-dihydroxyvitamin D.92 This latter metabolite increases gastrointestinal absorption of calcium and, with PTH, induces skeletal resorption, causing the restoration of extracellular fluid calcium and the neutralization of the signal initiating PTH release. The opposite series of homeostatic events occurs in response to a rise in extracellular fluid calcium levels. Although this scheme outlines the overall events that occur after a fall in calcium levels, aspects of the response may vary with the extent of the fall in calcium concentration and the consequent rise in PTH. There is some evidence that the actions of PTH on target tissues may depend on its prevailing concentration.93 Consequently, certain actions of PTH, such as renal calcium retention and even skeletal anabolic actions, may predominate at relatively low circulating concentrations of PTH, but skeletal lysis may become evident only at higher levels of circulating PTH. In addition to regulating calcium homeostasis, PTH elicits various other responses. Whether these other functions evolved independently or to complement the role of the hormone in maintaining calcium homeostasis is unclear. Among these other responses are perturbations of other ions, the most marked of which are those involving phosphate. As a consequence of PTH-enhanced 1,25-dihydroxyvitamin D production, the gastrointestinal absorption of phosphate is facilitated to some extent, and with PTH-induced skeletal lysis, phosphate and calcium are released. These effects increase the extracellular fluid phosphate levels, but the predominant effect of PTH on phosphate homeostasis is to inhibit renal phosphate reabsorption and produce phosphaturia. Consequently, a net decrease in extracellular fluid phosphate concentration occurs, and this may be viewed as adjunctive to the role of PTH in raising calcium levels. In common with other peptide hormones, PTH interacts through a receptor on the outer surface of the plasma membrane of target cells. In these target tissues, the result of this interaction of PTH with the membrane receptor has classically been appreciated to be the stimulation of the enzyme adenylate cyclase on the inner surface of the plasma membrane, although the same receptor can couple to phosphatidylinositol turnover as well. The product of this adenylate cyclase activity, cellular cAMP, and the products of phospholipase activity, IP3, diacyl-glycerol, and intracellular Ca2+, initiate a cascade of events leading to the final cellular response to the hormone. The PTH receptor may activate both these intracellular signaling path-ways in some target tissues and act preferentially by means of one or another pathway in other cells, but these issues are still under study. STRUCTURE OF PARATHYROID HORMONE CORRELATED WITH FUNCTION The amino acid sequences of the major glandular form of mammalian PTH, PTH 1–84, have now been determined in humans, cattle, pigs, dogs, and rats (Fig. 51-5).11,20 The corresponding amino acid sequence of the chicken peptide has also been determined and contains 88 rather than 84 amino acids. Studies correlating hormonal structure and function have emphasized the importance of the NH2 -terminal region to bioactivity. Considerable deletion of the middle and COOH-terminal region of the intact peptide can be tolerated without apparent loss of biologic activity. A peptide composed of the NH2-terminal 34 residues, PTH 1–34, appears to contain all of the conventional bioactivity of the intact hormone when tested in various bioassay systems.94 Synthetic NH2-terminal fragments of PTHrP, which share a high degree of homology with PTH within the NH2-terminal 13 residues, have been shown to mimic PTH actions in many bioassay systems.95,96 and 97

FIGURE 51-5. Amino acid sequence of mammalian parathyroid hormone 1–84. The backbone is that of the human sequence, and substitutions in each of the bovine, porcine, and rat hormones are shown at the specific sites.

Considerable effort has been expended to define the regions of the NH2 -terminal domain of the PTH molecule (and more recently of the PTHrP molecule) that are required to interact with the PTHR and stimulate downstream signaling. These studies examined full agonist, partial agonist, and antagonist activities of a variety of synthetic analogs of PTH (1-34), generally using in vitro adenylate cyclase–stimulating activity as an index of bioresponse. Early studies demonstrated that either deletion or extension of the NH2 terminus of PTH markedly inhibited the capacity of this hormone to increase cAMP in target tissues.94,98 Subsequently, the NH2 terminus of PTHrP was shown to manifest similar sensitivity.96 This work suggested that the NH2 -terminal domain of these hormones is a critical site for activation of the adenylate cyclase signaling system. Other work focused on the role of the COOH-terminal domain of PTH (1-34) in receptor binding, attempting to dissect a COOH-terminal “binding” domain in this entity from the NH2-terminal “activation” domain. Stimulated by the initial observation that poorly active NH2-terminal deleted analogs of PTH could still function as antagonists,98 this work ultimately suggested that the 14-34 portion of PTH contains the predominant binding domain.99 A number of studies have now been reported in which the structural basis of PTHrP function has also been analyzed, generally in comparison with PTH. For example, although PTH and PTHrP lack significant amino acid sequence homology in the 14-34 domains, these domains bind to osteoblastic cells with similar avidity, suggesting that these regions retain similar molecular topologies.100,101 Despite many additional studies, however, it remains premature to assign exact functional equivalence to most of the different domains and residues in these two hormones. Both PTH and PTHrP assume only a modestly ordered structural configuration in aqueous solution. Several nuclear magnetic resonance spectroscopy studies of PTH (1-34) and of PTHrP (1-34) in nonaqueous solutions (which are intended to mimic the membrane-bound receptor milieu) have disclosed some elements of secondary structure and have assigned two a-helical domains to these molecules, one less stable, extending for 7–11 amino acids in the NH2-terminal region, and another more stable, of 9–15 residues in the COOH domain.102,103,104,105,106 and 107 Some, but not all, studies have suggested interaction between the helices via a hydrophobic interface. In addition, the linker between the two helices is believed to be a turn or flexible hinge region. Cyclization of PTH and PTHrP analogs by a specific lactam apparently promotes a-helical conformation, and the stabilization of the a-helix in the putative COOH-terminal receptor-binding domain of PTH and PTHrP108,109 has been correlated with enhanced bioactivity of these hormones. To date, however, determination of the three-dimensional x-ray structure of PTH or PTHrP has not been feasible because of the difficulty in obtaining high-quality crystals. Although most in vitro studies correlating PTH structure and function have examined adenylate cyclase stimulating activity, several studies have been performed in

which other in vitro activities have been assessed. Such studies have indicated that the NH2 residues of PTH 1–34 may not be essential to induce effects on phosphoinositide metabolism and on the augmentation of intracellular calcium levels; rather that the PTH 23–34 sequence mediates this response. In vitro studies have described the mitogenic effects of the 25–47 region of PTH in osseous cells (not dependent on cAMP accumulation) and alkaline phosphatase–stimulating activity of the 53–84 sequence of PTH in osseous cells. The clinical significance of these in vitro observations remains to be determined.

PARATHYROID HORMONE RECEPTORS The PTH/PTHrP receptor is a member of the G-protein–linked receptor superfamily. These glycoproteins have hydrophobic sequences that are thought to span the plasma membrane seven times. These receptors couple to intracellular effectors through guanine nucleotide binding regulatory proteins or G proteins, which are heterotrimeric, consisting of alpha, beta, and gamma subunits. In the inactive state, the alpha subunit binds GDP; however, when the receptor is occupied by ligand, the G protein is activated by exchange of GTP for GDP and dissociates into a and b-g subunits. The alpha subunit, with GTP bound, can then stimulate (Gsa) or inhibit (Gia), the activity of effector molecules such as adenylate cyclase or phospholipase C. The a subunit also has an intrinsic GTPase activity that hydrolyzes GTP to GDP, terminating the effector activation. The PTH/PTHrP receptor belongs to a subgroup of the G-protein coupled receptor superfamily (group II GPCR) that, by virtue of their structural similarities, also includes the receptors binding secretin, calcitonin, vasoactive intestinal peptide, glucagon, glucagon-like peptide I, and growth hormone–releasing hormone. These receptors show no significant sequence identity with other known G-protein–linked receptors. All of these receptors couple to the adenylate cyclase effector by means of Gsa. The activated PTH/PTHrP receptor couples to at least two effectors: adenylate cyclase and phospholipase C. Other members of this subgroup—including the receptors for calcitonin, glucagon, and glucagon-like peptide I—also couple to multiple signaling pathways. PTH/PTHrP receptor cDNAs have been cloned and characterized from several species and tissues, including opossum kidney, rat bone, and human kidney and bone.110,111,112 and 113 The amino acid sequences of the receptors are highly homologous, demonstrating a marked conservation across species. The opossum and rat PTH/PTHrP receptors bind NH2-terminal analogs of PTH and PTHrP with similar affinity, while the human receptor appears to bind PTH somewhat preferentially. This is consistent with the study results of relative bioactivities of PTH and PTHrP in vivo in humans, that showed PTHrP 1–34 to be less potent than PTH 1–34.114,114a The gene for the PTH/PTHrP receptor contains multiple introns (Fig. 51-6). For example, the mouse gene has at least 15 exons—14 of which encode the receptor protein—that span more than 32 kb of genomic DNA.115 There are eight exons containing predicted membrane-spanning domains. These exons are heterogeneous in length, and three of the exon-intron boundaries fall within putative transmembrane sequences, suggesting that these exons did not arise from duplication events. The exon-intron organization of the PTH/PTHrP gene is similar to that of the growth hormone–releasing hormone gene, especially in the transmembrane regions, suggesting that the two genes evolved from a common precursor. PTH/PTHrP receptor gene transcription in mice is controlled by two promoters; P1, which is selectively active in kidney, and P2, which functions in a variety of tissues. Although P1 and P2 are conserved in the human PTH/PTHrP receptor gene, P1 activity in the kidney is weak, and a third promoter, P3, accounts for the majority of renal PTH/PTHrP receptor transcripts in humans.116 During development, only P2 is active at midgestation in many human tissues, including calvaria and long bone. Thus, factors regulating the well-conserved P2 promoter control PTH/PTHrP receptor gene expression during skeletal development. Later in development, receptor gene expression is up-regulated with the induction of both P1 and P3 promoter activities.116a

FIGURE 51-6. Schematic representation of the parathyroid hormone (PTH) and parathyroid hormone–related protein receptor. The gene contains at least 15 exons. Exon I encodes the 5' untranslated region of the mRNA. The portions of the receptor encoded by the remaining 14 exons are schematically depicted by the alternate blocks of unfilled and filled circles. Exons are depicted as follows: SS; putative signal sequence; E1–E4; extracellular sequences; T1–T7; transmembrane sequences; C; cytoplasmic sequence. The mouse PTH/PTHrP receptor of 591 amino acids is shown.

The PTH/PTHrP receptor gene promoters lack consensus TATA elements, and the downstream promoter P2 and P3 promoters are GC-rich, which would be consistent with the widespread expression of the receptor mRNA. Although the receptor mRNA is highly expressed in kidney and bone, which are the primary target tissues of PTH, it is also expressed in nonclassic target tissues such as liver, brain, smooth muscle, spleen, testis, and skin.112,117 In most of these tissues, the receptor probably mediates the local action of PTHrP. Although the predominant transcript is 2.5 kb, some tissues also express smaller or larger transcripts that probably are the result of alternative splicing of the primary transcript. The functional significance of these different forms of the receptor is unknown. A second related receptor, PTHR2, which is the product of a distinct gene and binds PTH but not PTHrP, has been identified.118 Its expression is limited to brain, pancreas, testis, and placenta; its function is unknown. Ligand-binding specificity of the PTH/PTHrP and the PTH2 receptors resides predominantly, but not exclusively, in its NH2-terminal extracellular domain, whereas activation and generation of the cAMP signal is generated in the membrane-associated domain and involves specific amino acids in transmembrane domain (TM) 3 and extracellular loop 2. This suggests a two-step interaction of PTH and receptor, whereby the ligand first complexes with the NH2 -terminal domain of the receptor, after which the complex interacts with the membrane-associated part of the receptor to generate the signal. Extensive analysis of the PTHR has also been performed in an attempt to delineate the structural basis of its function. Such studies using mutant and chimeric PTHR have indicated that the NH2 -terminal extracellular domain of the PTHR in particular is important for ligand binding and appears to interact with the COOH region of PTH or PTHrP.119 Nevertheless, amino acid residues elsewhere in the PTHR, such as in the third extracellular loop, may also contribute to this binding.120 Residues in the membrane-spanning helices of the PTHR may be required for binding residues in the NH2 -terminal domain of the ligand. Amino acids in the extracellular loops and the transmembrane regions are involved in signaling by PTH analogs, and residues in several intracellular loops and the COOH terminal intracellular tail are implicated in linkage to both Gsa and Gqa. Residues in the second intracellular loop appear to be especially critical for interaction with Gqa.121 Other studies have identified sites in the COOH tail of the PTHR that may be important for receptor internalization and phosphorylation.122 Consequently, a great deal of important information has been gleaned and continues to be generated on specific sites in the PTHR that may be necessary for hormone binding, for hormone signaling, and for PTHR regulation. Nevertheless, a comprehensive understanding of the detailed interaction of PTH and of PTHrP with the PTHR will have to await the determination of the three-dimensional structure of the bimolecular complex. High circulating levels of PTH in hyperparathyroid states have been associated with hormonal desensitization in target tissues, apparently caused by diminished receptor capacity and a postreceptor reduction in functional levels of Gs.123,124 PTH receptors, similar to many peptide hormone receptors, appear to be subject to down-regulation. The renal resistance to PTH often seen in hyperparathyroid states may be the partial result of this kind of regulatory mechanism.125 More widespread reductions in Gs are associated with hormone resistance in the disorder pseudohypoparathyroidism type 1a. The human PTH/PTHrP receptor gene has been localized to chromosome 3p21.1-22112 (and the PTHR2 gene to chromosome 2q33). A search for PTH/PTHrP receptor defects in patients with apparent resistance to endogenous PTH, such as in patients with pseudohypoparathyroidism type 1b, has produced negative results.126,127 These patients have end-organ resistance to PTH without typical features of Albright hereditary osteodystrophy and therefore are thought likely to manifest a defect in PTH/PTHrP receptor expression or function. Because no mutations were identified in the receptor gene in such patients, it was indicated that mutations in genes for other proteins involved in the PTH/PTHrP signaling pathway are most likely responsible for the defects. Linkage to chromosome 20q13.3, which includes the GNAS1 locus encoding Gsa has been established in kindreds with PHP-1b.128 In addition, the genetic defect is imprinted paternally and is therefore inherited in the same fashion as the PTH resistance in kindreds with PHP-1a and/or pseudopseudo-hypoparathyroidism and in a mouse model heterozygous for ablation of the Gnas gene.129 Although the precise nature of mutations causing PHP-1b remains to be elucidated, an understanding of how mutations at a single chromosomal locus cause overlapping and/or distinct phenotypes is likely to be related to (a) the appreciation of the complex nature of the GNAS1 gene that because of its bidirectional imprinting encodes maternally, paternally, and biallelically derived proteins,130 and (b) the subtle cell-specific imprinting of the Gsa transcript. Newly

discovered exons upstream of the Gsa exons encode two different proteins, XLas and NESP55, which are expressed in neuroendocrine cells and are probably important for secretory vesicle formation and function. It therefore is possible that a structural or regulatory mutation within the complex GNAS1 locus could account for PHP-1b, although this remains to be demonstrated. Direct evidence that the PTH/PTHrP receptor mediates the calcium homeostatic actions of PTH and the growth plate actions of PTHrP in humans has come from the study of rare genetic disorders. Jansen metaphyseal chondrodysplasia (JMC) is inherited in an autosomal-dominant fashion although most reported cases are sporadic.131 The disorder comprises short-limbed dwarfism secondary to severe growth plate abnormalities, asymptomatic hypercalcemia, and hypophosphatemia. There is increased bone resorption similar to that in primary hyperparathyroidism and urinary cAMP levels are elevated, but circulating PTH and PTHrP levels are low or undetectable. Although the PTH/PTHrP receptor is widely expressed in fetal and adult tissues, it is most abundant in three major organs, the kidney, bone, and metaphyseal or cartilaginous growth plate. It was hypothesized that the changes in mineral ion homeostasis and the growth plate in JMC are caused by constitutively active PTH/PTHrP receptors. Indeed, heterozygous PTH/PTHrP receptor gene mutations have now been identified in JMC patients. Two recurrent mutations have been described: one changes His to Arg at position 223 at the cytoplasmic end of the second transmembrane domain and the other is a Thr-to-Pro conversion at position 410 toward the cytoplasmic end of the sixth transmembrane domain.132 Current theory holds that in the absence of agonist, G protein-coupled receptors are constrained in an inactive conformation by interactions between select amino acids in the transmembrane domains. Both His-223 and Thr-410 are highly conserved in members of the group II GPCRs, attesting to their functional importance for these receptors. Mutation of position 223 to any positively charged amino acid results in constitutive activity. Possibly, in the wild-type receptor, agonist binding leads to protonation of His-223 and results in activation. In the case of Thr-410 substitution with any other amino acid leads to constitutive activity indicating the key role Thr-410 plays in maintaining the receptor in an inactive state. Inactivating or loss-of-function mutations in the PTH/PTHrP receptor have been implicated in the molecular pathogenesis of Blomstrand lethal chondrodysplasia (which was first described as a disease entity in 1985).133 This rare disease is characterized by advanced endochondral bone maturation, short-limbed dwarfism, and fetal death, thus mimicking the phenotype of PTH/PTHrP receptor-less mice. The majority of patients with Blomstrand lethal chondrodysplasia were born to phenotypically normal, consanguineous parents, suggesting an autosomal-recessive mode of inheritance. In one such fetus, a homozygous missense mutation converting Pro-132 to Leu in the extracellular domain of the PTH/PTHrP receptor was identified.134 Although the mutated receptor is expressed on the plasma membrane, it demonstrates reduced ligand binding and adenylate cyclase signaling. In another patient born to noncon-sanguineous parents, a heterozygous deletion of 11 amino acids in the fifth transmembrane domain of the receptor was found.135 This was the result of a G® A substitution in exon M5, which created a novel splice acceptor site in the maternal allele. The paternal allele is not expressed, although the reasons for this are not known. As in the other example, whereas the mutant receptor is well expressed in cells, it fails to bind ligand or stimulate cAMP or inositol phosphate production. Evidence has accumulated for the possible existence of other receptors in the PTH/PTHrP system in addition to the two receptors already characterized. A receptor responsive to NH2 -terminal PTH and PTHrP, which signals through changes in intracellular free calcium rather than cAMP, has been found to exist in keratinocytes and other cells but has not yet been characterized. A receptor that signals in a similar manner but is only responsive to PTHrP-(1-34) mediates the release of arginine-vasopressin from the supraoptic nucleus. Unique biologic roles are predicted for the mid-region of PTHrP (transplacental calcium transport), the basic region of PTHrP (nuclear import and inhibition of apoptosis), the carboxyl-terminal region of PTHrP (inhibition of bone resorption), and the mid/carboxyl-terminal part of PTH (binding to osteoblast-like cells and modulation of their activity), each of which may act via discrete receptors. Finally, discovery of the PTHR2 suggested the presence of an additional ligand in the PTH family. Thus, the PTHR2, responsive only to PTH and not PTHrP, is expressed mainly in brain, placenta, testis, and pancreas. A novel PTH-like ligand for the PTHR2 has been isolated from the hypothalamus.135a PARATHYROID HORMONE–INDUCED SIGNAL TRANSDUCTION ADENYLATE CYCLASE Adenylate cyclase stimulation and the subsequent generation of cAMP are believed to be important events in the actions of PTH in the kidney and in the skeleton. Cyclic adenosine monophosphate mimics phosphaturic and calcium-retaining effects of PTH in the kidney in vivo and in vitro and mediates PTH-stimulated renal 1a-hydroxylase activity.136 Cyclic adenosine monophosphate has also been implicated in the hypercalcemic action of PTH and may simulate PTH-induced bone resorption in vitro. Furthermore, the bone anabolic effect of PTH appears dependent on the capacity to stimulate cAMP. Cyclic AMP produced in target cells for PTH is believed to stimulate cAMP-dependent protein kinase isoenzymes; types I and II have different biochemical characteristics and may serve different functions. The isoenzymes are tetramers consisting of two regulatory and two catalytic subunits that have similar catalytic but different regulatory components, termed RI and RII, respectively. With binding of cAMP to the regulatory component, the holoenzyme dissociates, releasing the catalytic component that facilitates the transfer of a terminal phosphate group from a nucleotide donor, usually ATP, to an amino acid residue (i.e., serine, threonine, or tyrosine) of the substrate protein. PTH-induced stimulation of the two types of protein kinase has been demonstrated in normal osteoblasts and in a malignant osteoblast line.137 One of the substrates of PKA is the widely expressed cAMP response element binding protein (CREB).40 The phosphorylated CREB bound to the cAMP response element (CRE) in the promoters of responsive genes couples to the basal transcriptional machinery via cointegrators such as CREB-binding protein (CBP). PTH activates several osteoblastic genes, including c-fos,138 by this mechanism. OTHER SECOND MESSENGERS OF PARATHYROID HORMONE ACTION In addition to cAMP, other second messengers, such as the calcium ion, have been implicated as being potentially important in PTH action. In some species, transient hypocalcemia, presumably caused by calcium entry into bone cells, is the earliest event in the action of PTH on the skeleton in vivo. In vitro studies have shown that PTH promotes the uptake of calcium into isolated bone cells, that elevated calcium mimics or potentiates the effects of PTH on the enzymatic activities of isolated bone cells, and that calcium antagonists inhibit and calcium ionophores stimulate bone resorption. PTH stimulates phosphatidylinositol turnover in certain cell types and in renal membranes.139,140 In response to PTH-induced augmentation of phospholipase C action, presumably via Gq stimulation after occupancy of the PTH/PTHrP receptor, increased production of IP3 and diacylglycerol occurs. Increases in cytoplasmic calcium, presumably induced by IP3, have also been demonstrated, as has increased protein kinase C activity. The cellular response to PTH may therefore involve multiple mechanisms of cell signaling, and modulation of one message by another (“cross talk”) may affect the final response to the hormone. The relative contributions of adenylate cyclase versus phospholipase C stimulation to the pleiotropic effects of PTH in bone and kidney still remains to be clarified. EFFECTS OF PARATHYROID HORMONE IN TARGET TISSUES BONE Consistent with its prime function of raising the extracellular fluid calcium concentration, the most appreciated effect of PTH is a catabolic one in bone.141,142 The end result is the breakdown of mineral constituents and bone matrix, as manifested in vivo by the release of calcium and phosphate, by increases in plasma and urinary hydroxyproline, and by other indices of bone resorption. This process appears to be mediated by osteo-clastic osteolysis, but the mechanism by which PTH causes osteoclastic stimulation is indirect. Unlike calcitonin, PTH, when administered in vivo, does not bind directly to multinucleated osteoclasts. In vivo studies using autoradiography have revealed PTH binding to mature osteoblasts and to skeletal mononuclear cells in the intertrabecular region of the metaphysis (apparently to preosteoblastic stromal cells).143,144 Consequently, PTH-mediated increases in the number and function of osteoclasts (which are of hematogenous origin) appear to occur indirectly, through effects on cells of the osteoblast series (which are of mesenchymal origin).145 The results of these studies have been confirmed in experiments with the cloned PTH/PTHrP receptor; these have confirmed expression of the receptor in osteoblastic cells but not osteoclasts.146 PTH-induced stimulation of multinucleated osteoclasts occurs through the action of PTH-stimulated osteoblastic activity.147 This osteoblastic activity is characterized by release of intermediary factors such as cytokines. One of the most important of these intermediary factors is osteoclast differentiation factor (ODF), also known as osteoprotegerin ligand (OPGL).148 This protein is expressed in bone-lining cells or in osteoblast precursors that support osteoclast recruitment. ODF/OPGL, which is a member of the tumor necrosis factor (TNF) family, is identical to the molecule TRANCE/RANKL, which is expressed in T cells and previously was identified as a dendritic cell survival factor in vitro. Several bone resorbing factors, such as PTH, PTHrP, PGE2, some of the interleukins, and 1,25(OH)2D3, up-regulate ODF/OPGL gene expression in osteoblasts and bone stromal cells (Fig. 51-7). Interaction of ODF/OPGL with its receptor on osteoclast progenitors and osteoclasts then stimulates their recruitment and activation and delays their degradation. The ablation of the ODF/OPGL gene in mice has confirmed it to be a key regulator of osteoclastogenesis as well as of lymphocyte development and lymphnode organogenesis.149 Osteoprotegerin (OPG)/osteoclastogenesis inhibitory factor (OCIF)150 is a soluble receptor for ODF/OPGL that inhibits recruitment,

activation, and survival of osteoclasts and therefore inhibits osteoclastic bone resorption. OPG/OCIF acts as a natural decoy receptor to disrupt the interaction between ODF/OPGL, which is released by osteoblast-related cells, and the ODF/OPGL receptor on osteoclast progenitors.

FIGURE 51-7. Model of the activation of bone resorption and formation by parathyroid hormone (PTH) or parathyroid hormone–related protein (PTHrP). PTH of PTHrP bind via their NH2 terminus to the PTH/PTHrP receptor (PTHR) of osteoblastic stromal cells or osteoblasts. Osteoblastic stimulation may lead to bone formation. PTH/PTHrP binding to the PTHR, most likely on the osteoblastic stromal cell, can also lead to the binding of ODF/OPGL/RANKL to receptors (RANK) on cells of the osteoclast series. This cytokine can then enhance commitment, proliferation, differentiation, and fusion of osteoclast precursors to form active bone-resorbing osteoclasts. Alternatively, ODF/OPGL/RANKL may be bound by OPG, preventing osteoclastic bone resorption.

Early and late phases of calcium mobilization have been described after in vivo administration of PTH.151 The early hypercalcemic response occurs from 10 minutes to 3 hours after hormone exposure. This response may be a consequence of increased metabolic activity of preexisting osteoclasts or of other cell types that enhance the transfer to bone of calcium in bone that is already in solution. A more sustained hypercalcemic response to PTH administration, occurring over approximately 24 hours, appears to depend on new protein synthesis and to involve a quantitative increase in osteoclasts, a change in the structure of the osteoclasts (i.e., increased ruffled borders, which is a zone of the cell believed to be involved in skeletal resorption); an increase in the secretion of lysosomal enzymes, including collagenase and acid hydrolases such as acid phosphatase; and acidification of the extracellular milieu of the osteoclast. In contrast to sustained administration of PTH, which causes bone resorption and hypercalcemia, low and intermittent doses of PTH-(1-34)—and PTHrP-(1-34) and related analogs—promote bone formation.152 Furthermore, in clinical studies, daily injections of PTH-(1-34) have been reported to increase hip and spine bone mineral density (BMD). The consequences of the effects of PTH on osteoblast activity are therefore complex. Thus, in addition to using osteoblastic cells to relay signals to the osteoclast lineage to resorb bone, PTH also appears to stimulate osteoblastic cells to enhance new bone formation. This might occur as a result of the ability of PTH to stimulate the production of growth factors, such as IGF-I, by osteoblasts. However, the bone-forming activity of PTH may depend more on promoting differentiation of precursor cells into secretory osteoblasts than on an action on mature osteoblasts.153 The precise mechanisms of the anabolic effect of PTH still remain to be clarified, however. The anabolic and catabolic effects of PTH on osteoblasts may therefore represent a combination of direct and indirect effects; effects of different domains of the PTH molecule that signal differently; discrete functions of morphologically similar but functionally distinct osteoblastic cells; or differences in hormonal effects based on different times of exposure or different hormone concentrations. KIDNEY One of the first-described effects of the administration of PTH intravenously was phosphaturia (see Chap. 206). Renal tubular reabsorption of phosphate is an active process, with a limited transport capacity resulting in a maximum rate of tubular reabsorption (Tm PO4).154 Because the absolute values of Tm PO4 vary considerably among individuals and most of this variation can be explained by differences in glomerular filtration rate (GFR), the Tm PO4/GFR ratio has been suggested as a more accurate index of phosphate reabsorption. This index, which represents the sum of the heterogeneous maximum reabsorption rates of all individual nephrons, is reduced by increased concentrations of PTH and increased in its absence. The molecular basis of the inhibition of Na+-dependent phosphate transport by PTH was clarified by the cloning of the Na+ -phosphate co-transporter, NPT-2.155 Thus, inhibition of phosphate reabsorption by PTH in the renal proximal tubule appears to involve PTH-induced endocytosis of NPT-2 and its subsequent intracellular degradation. Another relatively immediate effect of PTH that contributes to its role in calcium homeostasis is enhanced fractional reabsorption of calcium from the glomerular fluid. However, excessive circulating concentrations of PTH ultimately are associated with a rise in urinary calcium because of increases in extracellular calcium levels and therefore in the filtered load of calcium. The capacity of PTH to produce a mild renal tubular acidosis appears linked to its ability to inhibit the activation of the type 3 Na+-H+ exchanger (NHE3).156 Augmentation of urinary cAMP excretion in response to PTH administration is one of the earliest renal responses and is consistent with its postulated role as a second messenger for many or most renal responses. Nevertheless, evidence for an important role for inositol phosphates, protein kinase C, and intracellular calcium in the renal actions of PTH is also available; the relative in vivo contributions of the two pathways to the multiple effects of PTH in the kidney still requires clarification. The renal actions of PTH to increase urinary phosphate and cAMP excretion form the basis for the modified Ellsworth-Howard test that is used clinically to establish renal PTH responsiveness (see Chap. 60). PTH alters renal function because of its interactions with multiple regions of the nephron. Evidence for PTH binding to the primary foot processes of podocytes of renal corpuscles and for the stimulation of adenylate cyclase activity in rat glomeruli can be correlated with reduction of the GFR in rats by PTH as a result of decreasing the ultrafiltration coefficient.144 Evidence for the expression of the cloned PTH/PTHrP receptor in renal glomeruli confirms the previous findings. PTH appears to reach the luminal surface of polar tubular cells by glomerular filtration and the basolateral surface through the peritubular capillary plexus. High-capacity binding sites have been described on the periluminal portion of the earliest part of the proximal tubule, at which, related to the microvillar surface, PTH degradative events appear to occur. Saturable, specific, low-capacity binding sites for PTH have been localized after injection in rats to the basolateral surface of the proximal convoluted tubule and pars recta, the thick ascending limb of the loop of Henle, and the distal convoluted tubule and pars arcuata of the collecting duct.144 This pattern of PTH binding and activity correlates well with the expression of the cloned PTH/PTHrP receptor and with the localization of PTH-stimulated adenylate cyclase activity in rat nephrons, which was demonstrated in studies using microdissection of tubules.156 These observations emphasize the diversity of PTH actions on the renal tubule, most of which can be mimicked by the infusion of cAMP onto the luminal aspect of tubular cells.157,158 The use of various techniques, including micropuncture and microperfusion, has localized PTH-induced inhibition of phosphate reabsorption to the proximal convoluted tubule and to the pars recta. 158 Inhibition of phosphate reabsorption in the proximal convoluted tubule appears to be accompanied by inhibition of sodium and fluid reabsorption. However, sodium is also reabsorbed more distally. Inhibition of phosphate reabsorption also may occur, although perhaps to a lesser extent, in the distal tubule. The proximal tubule appears to be the major site of action of PTH in stimulating the 1a-hydroxylase and increasing the production of 1,25-dihydroxyvitamin D. The PTH-induced inhibition of bicarbonate transport occurs in the proximal tubule and the pars recta, and the effect of PTH on this nephron segment may explain the rise in urinary bicarbonate produced by PTH infusion.158 The important site of PTH action to increase calcium and probably magnesium transport appears to be in the thick ascending limb of the loop of Henle and in the distal convoluted tubule and earliest portion of the cortical collecting tubule. Despite the demonstration of the dual role of PTH in increasing extracellular fluid calcium concentrations through renal and skeletal routes, the relative importance of each is unclear.159 However, contributions of the kidney and skeleton to calcium homeostasis may depend on the concentration of circulating PTH and the duration of elevated hormonal levels.

OTHER TARGET TISSUES Although the administration of PTH in vivo can enhance intestinal calcium absorption, this effect appears to be indirect and mediated by the increased production of 1,25-dihydroxyvitamin D.92 Among other reported effects of PTH are direct effects on vascular tone, stimulation or inhibition of mitosis of various cells in vitro, increased concentrations of calcium in mammary and in salivary glands, enhanced hepatic gluconeogenesis, and enhanced lipolysis in isolated fat cells. The demonstration of the widespread expression of the PTHrP gene and the equally disseminated expression of the PTH/PTHrP receptor gene suggests that many of the “noncalcemic” actions of PTH may be carried out by PTHrP acting in an autocrine or paracrine manner.160

MEASUREMENT RADIOIMMUNOASSAYS Radiommunoassays (RIAs) developed for human PTH generally do not use human PTH (hPTH) 1–84 as a tracer because of its scarcity and because it lacks a tyrosine residue for convenient radioiodination. Instead, bovine PTH (bPTH) 1–84 is used, possibly contributing to the reduced sensitivity for hPTH. The antisera used in PTH RIAs are generally polyclonal and have been raised against bPTH 1–84 or hPTH 1–84. Nevertheless, populations of antibodies within polyclonal antisera directed against specific epitopes contained in the 1–84 molecule may predominate. Antibody populations recognizing the COOH-terminal region of PTH 1–84 usually are readily obtained after immunization with bPTH 1–84; antibodies recognizing the midregion also occur with high frequency.161 Sufficiently sensitive antisera containing antibodies predominantly interacting with the NH2-terminal region are more unusual but have been successfully raised.162 Attempts to direct antibody specificity have used several strategies. Although the development of monoclonal antibodies to PTH or PTH fragments would seem to be the most direct route to achieve specificity, this approach has not been successful for various technical and biochemical reasons. Instead, specificity has been achieved by using polyclonal antisera with predominant specificity for selected regions, as determined by their reactivity toward synthetic fragments of discrete regions of the molecule or by enhancing the specificity of such antisera by using synthetic fragments of the midregion or COOH-terminal end of PTH as tracers. An important advance in the measurement of PTH is the development of immunoradiometric assays (IRMAs) that use antisera directed against the NH2-terminal and midregion or COOH region of the molecule. These assays (i.e., “intact” assays) detect mainly intact PTH 1–84 and appear to be the most sensitive and specific.163 Defining the specificity of an RIA is important because of the complex metabolism of PTH, which results in a multiplicity of circulating molecular forms. The most abundant circulating forms are midregion and COOH fragments because of their longer half-life in the circulation; these become even more predominant in renal failure when clearance is further impaired.89,90 The presumed bioactive forms, intact PTH 1–84 and the NH2-terminal fragment, are cleared more readily and circulate at lower levels. Midregion and COOH-directed RIAs detect long-lived fragments and intact bioactive hormone. These RIAs generally do not require the high sensitivity of NH2-directed assays to be useful, because they measure higher absolute levels of PTH. Most material measured by these RIAs is thought to be inactive, but this has not greatly restricted their clinical usefulness in the diagnosis of primary or secondary hyperparathyroidism. For primary hyperparathyroidism, the midregion and COOH-directed RIAs may be especially useful in providing an index of integrated secretory function of the parathyroid glands.85 However, first it should be established that a given midregion or COOH-terminal RIA is clinically useful. Amino-terminal RIAs, which determine the level of short-lived PTH forms, can also be used to detect hypersecretion in primary hyperparathyroidism but may be especially useful in assessing acute changes in PTH secretion in physiologic studies. Two-site IRMAs, which also measure bioactive hormone, are generally more sensitive and can substitute for NH2-directed assays. For cases of hyperparathyroidism associated with renal failure, NH2-directed assays have the theoretical advantage of providing a better estimate of circulating bioactive PTH forms and differentiating decreased PTH clearance from hypersecretion. From a practical point of view, two-site IRMAs are preferred to NH2-directed assays. If PTH levels are determined serially during the progression of renal failure, COOH-directed assays also may provide useful estimates of the severity of the hyperparathyroidism, and hormone levels determined by COOH-directed assays in renal failure have correlated well with the extent of skeletal resorption.85 It has been suggested that midregion assays may detect intact PTH more readily than COOH-directed assays and may have a somewhat greater ability to measure a variety of PTH forms than the COOH-directed assays. However, midregion assays are subject to restrictions and advantages similar to those of the COOH-directed assays. Few RIAs, no matter the degree of specificity, completely discriminate between concentrations of immunoreactive PTH found in healthy persons and those detected in hyperparathyroidism, although best results appear to be obtained with two-site IRMAs. It has been suggested that PTH levels should be interpreted in conjunction with prevailing levels of blood calcium. This assists in the assay analysis; if a given PTH level is measured in a normocalcemic individual and that same PTH level is found in a patient with hypercalcemia, the level in the latter case can be considered inappropriately elevated. Another limitation of most RIAs is the inability to measure values in all healthy individuals. A lower limit of normality generally cannot be established, which makes it difficult to ascertain reduced PTH levels. Sensitive two-site IRMAs often approach a lower limit of normality. PTH RIAs are useful in differentiating hypocalcemia resulting from PTH deficiency (i.e., hypoparathyroidism) from hypocalcemia resulting from other causes. In the former situation, PTH is undetectable or is found in very low concentrations, but in the latter situation, the hypocalcemic stimulus is associated with increased PTH secretion and elevated levels of PTH (see Chap. 58 and Chap. 61). BIOASSAYS With the unraveling of the complex metabolism of PTH and the resulting heterogeneity of circulating hormone, the care that must be exercised in interpreting the results of PTH RIAs became apparent, and the possibility of using PTH bioassays to supplement the information obtained from RIAs arose. Some PTH bioassays estimate the biologic effects of the hormone in vivo, and some estimate levels of the hormone based on biologic effects in vitro. It has been estimated that the normal circulating levels of bioactive PTH are ~1 pmol/L. IN VIVO BIOASSAYS Measurements of in vivo phosphaturic effects (e.g., determination of Tm PO4/GFR) and renal calcium retention (e.g., fractional excretion of calcium) provide estimates of PTH effects in vivo and have been useful as adjunctive studies of PTH. Partly because of their lack of specificity, these assays have not achieved widespread popularity as indices of PTH concentrations in most clinical situations. Conversely, estimates of urinary cAMP excretion (measured by RIA) have been fairly widely used.164 Several hormones in addition to PTH (e.g., vasopressin) contribute to urinary cAMP levels, although the PTH-produced component is the major fraction. Greater specificity of urinary cAMP for PTH can be achieved by measuring the fraction of urinary cAMP that is of renal origin. This nephrogenous component can be determined from the clearance of cAMP relative to the GFR, an estimate requiring plasma cAMP measurements, or it can be determined by relating the urinary cAMP to 100 mL of glomerular filtrate. Nephrogenous cAMP is a specific and rather sensitive in vivo bioassay for PTH.164 Nephrogenous cAMP is often elevated in primary hyperparathyroidism and decreased in hypoparathyroidism. There is, however, overlap with the normal range. Most nephrogenous cAMP determinations accurately reflect circulating PTH levels as determined by RIA, but they do not have the sensitivity or specificity of the best two-site IRMAs. IN VITRO BIOASSAYS Several attempts have been made to develop clinically useful in vitro bioassays for the measurement of active PTH. The capacity of PTH to stimulate adenylate cyclase, with the addition of guanyl nucleotides, in purified renal membrane preparations or tumor cell lines and the availability of synthetic antagonistic PTH fragments have imparted useful specificity to such renal membrane bioassay preparations.165,166 Unfortunately, the relative insensitivity of this approach has precluded its use for anything other than experimental purposes. Another approach is the cytochemical PTH bioassay. The most widely used has been a renal cytochemical bioassay (CBA) performed in guinea pig kidney segments maintained in non-proliferative organ culture.4,91,167 The major drawbacks of the method are its technical difficulty and low throughput, which have precluded its use for routine clinical assay purposes.

CONCLUSIONS Considerable progress has been made in examining the biosynthesis of PTH, the regulation of PTH secretion, and the mechanism of PTH action. These advances include elucidation of the molecular genetics of PTH, discovery of the parathyroid calcium “sensor,” and determination of the structure of the PTH/PTHrP receptor. Discovery of the ODF/OPGL/RANKL system and the cloning of NPT-2 have greatly contributed to the understanding of the mechanism of PTH action in bone and kidney. The identification of PTHrP has disclosed a new member of the PTH gene family and provided possibilities for understanding the biologic effects of both molecules. This improved comprehension of the biochemistry and physiology of PTH should translate into exciting insights into the pathophysiology of disease states in which PTH is implicated. CHAPTER REFERENCES 1. Kronenberg HM, Igarashi T, Freeman MW, et al. Structure and expression of the human parathyroid hormone gene. Recent Prog Horm Res 1986; 42:641. 2. Cohn DV, Elting J. Biosynthesis, processing and secretion of parathormone and secretory protein-I. Recent Prog Horm Res 1983; 39:181. 3. Walter P, Ibrahimi I, Blobel G. Translocation of proteins across the endoplasmic reticulum. 1. Signal recognition protein (SRP) binds to in vitro–assembled polysomes synthesizing secretory protein. J Cell Biol 1981; 91:545. 4. Goltzman D, Bennett HPJ, Koutsilieris M, et al. Studies of the multiple forms of bioactive parathyroid hormone and parathyroid hormone-like substances. Recent Prog Horm Res 1986; 42:665. 5. Hendy GN, Bennett HPJ, Lazure C, et al. Proparathyroid hormone (ProPTH) is preferentially cleaved to parathyroid hormone (PTH) by the prohormone convertase furin: a mass spectrometric study. J Biol Chem 1995; 270:9517. 6. Canaff L, Bennett HPJ, Hou Y, et al. Proparathyroid hormone processing by the proprotein convertase PC7: comparison with furin and assessment of modulation of parathyroid convertase mRNA levels by calcium and 1,25-dihydroxy vitamin D3. Endocrinology 1999; 140:3633. 7. Rabbani SA, Kremer R, Bennett HPJ, Goltzman D. Phosphorylation of parathyroid hormone by human and bovine parathyroid glands. J Biol Chem 1984; 259:2949. 8. Hendy GN, Kronenberg HM, Potts JT Jr, Rich A. Nucleotide sequence of cloned DNAs encoding human preproparathyroid hormone. Proc Natl Acad Sci U S A 1981; 78:7365. 9. Vasicek TJ, McDevitt BE, Freeman MW, et al. Nucleotide sequence of genomic DNA encoding human parathyroid hormone. Proc Natl Acad Sci U S A 1983; 80:2127. 10. Weaver CA, Gordon DF, Kissil MS, et al. Isolation and complete nucleotide sequence of the gene for bovine parathyroid hormone. Gene 1984; 28:319. 11. Heinrich G, Kronenberg HM, Potts JT Jr, Habener JF. Gene encoding parathyroid hormone. Nucleotide sequence of the rat gene and deduced amino acid sequence of rat preproparathyroid hormone. J Biol Chem 1984; 259:3320. 12. Russell J, Sherwood LM. Nucleotide sequence of the DNA complementary to avian (chicken) preproparathyroid hormone mRNA and the deduced sequence of the hormone precursor. Mol Endocrinol 1989; 3:325. 13. Yasuda T, Banville D, Hendy GN, Goltzman D. Characterization of the human parathyroid hormone–like peptide gene: functional and evolutionary aspects. J Biol Chem 1989; 264:7720. 14. Mangin M, Ikeda K, Dreyer BE, Broadus AE. Isolation and characterization of the human parathyroid hormone–like peptide gene. Proc Natl Acad Sci U S A 1989; 86:2408. 15. Suva LJ, Mather KA, Gillespie MT, et al. Structure of the 5' flanking region of the gene encoding human parathyroid hormone–related protein (PTHrP). Gene 1989; 77:95. 16. Karaplis AC, Yasuda T, Hendy GN, et al. Gene encoding parathyroid hormone– like peptide: nucleotide sequence of the rat gene and comparison with the human homologue. Mol Endocrinol 1990; 4:441. 17. Naylor SL, Sakaguchi AY, Shows TB, et al. Human parathyroid hormone gene (PTH) is on the short arm of chromosome 11. Somatic Cell Mol Genet 1983; 9:609. 18. Seldin MF, Mattei M-G, Hendy GN. Localization of mouse parathyroid hormone-like peptide gene (Pthlh) to distal chromosome 6 using interspecific backcross mice and in situ hybridization. Cytogenet Cell Genet 1992; 60:252. 19. Pausova Z, Morgan K, Fujiwara TM, et al. Molecular characterization of an intragenic minisatellite (VNTR) polymorphism in the human parathyroid hormone–related peptide gene in chromosome region 12p12.1–p11.2. Genomics 1993; 17:243. 20. Keutmann HT, Sauer MM, Hendy GN, et al. Complete amino acid sequence of human parathyroid hormone. Biochemistry 1978; 17:5723. 21. Motokura T, Bloom T, Kim HG, et al. A novel cyclin encoded by bcl1-linked candidate oncogene. Nature 1991; 350:512. 22. Chandrasekharappa SC, Guru SC, Manickam P, et al. Positional cloning of the gene for multiple endocrine neoplasia-type 1. Science 1997; 276:404. 23. European Consortium on MEN1. Identification of the multiple endocrine neoplasia type 1 (MEN1) gene. Hum Mol Genet 1997; 6:1177. 24. Guru SC, Goldsmith PK, Burns AL, et al. MENIN, the product of the MEN1 gene, is a nuclear protein. Proc Natl Acad Sci U S A 1998; 95:1630. 25. Agarwal SK, Guru SC, Heppner C, et al. Menin interacts with the AP1 transcription factor JunD and represses JunD-activated transcription. Cell 1999; 96:143. 26. Friedman E, Sakaguchi K, Bale AE, et al. Clonality of parathyroid tumors in familial multiple endocrine neoplasia type 1. N Engl J Med 1989; 321:213. 27. Thakker RV, Bouloux P, Wooding C, et al. Association of parathyroid tumors in multiple endocrine neoplasia type 1 with loss of alleles on chromosome 11. N Engl J Med 1989; 321:218. 28. Arnold A, Horst SA, Gardella TJ, et al. Mutation of the signal peptide-encoding region of the preproparathyroid hormone gene in familial isolated hypoparathyroidism. J Clin Invest 1990; 86:1084. 29. Parkinson DB, Thakker RV. A donor splice site mutation in the parathyroid hormone gene is associated with autosomal recessive hypoparathyroidism. Nature Genet 1992; 1:149. 30. Simpson EL, Mundy GR, D'Souza SM, et al. Absence of parathyroid hormone messenger RNA in nonparathyroid tumors associated with hypercalcemia. N Engl J Med 1983; 309:325. 31. Nussbaum SR, Gaz RD, Arnold A. Hypercalcemia and ectopic secretion of parathyroid hormone by an ovarian carcinoma with rearrangement of the gene for parathyroid hormone. N Engl J Med 1990; 323:1324. 32. Fischer JA, Oldham SB, Sizemore GW, Arnaud CD. Calcium-regulated parathyroid hormone peptidase. Proc Natl Acad Sci U S A 1972; 69:2341. 33. Flueck JA, DiBella FP, Edis AJ, et al. Immunoheterogeneity of parathyroid hormone in venous effluent serum from hyperfunctioning parathyroid glands. J Clin Invest 1977; 60:1367. 34. Hanley DA, Takatsuki K, Sultan JM, et al. Direct release of parathyroid hormone fragments from functioning bovine parathyroid glands in vitro. J Clin Invest 1978; 62:1247. 35. Mayer GP, Keaton JA, Hurst JG, Habener JF. Effect of plasma calcium concentrations on the relative proportion of hormone and carboxyl fragments in parathyroid venous blood. Endocrinology 1979; 104:1778. 36. Kemper B, Habener JF, Rich A, Potts JT Jr. Parathyroid secretion: discovery of a major calcium-dependent protein. Science 1974; 184:167. 37. Cohn DV, Zangerle R, Fischer-Colbrie R, et al. Similarity of secretory protein-I from parathyroid gland to chromogranin A from adrenal medulla. Proc Natl Acad Sci U S A 1982; 79:6056. 38. Bhargava G, Russell J, Sherwood LM. Phosphorylation of parathyroid secretory protein. Proc Natl Acad Sci U S A 1983; 80:878. 39. Mouland AJ, Bevan S, White JH, Hendy GN. Human chromogranin A gene; molecular cloning, structural analysis and neuroendocrine cell-specific expression. J Biol Chem 1994; 269:6918. 40. Canaff L, Bevan S, Wheeler D, et al. Analysis of molecular mechanisms controlling neuroendocrine cell specific transcription of the chromogranin A gene. Endocrinology 1998; 139:1184. 41. el-Hajj Fuleihan G, Klerman EB, Brown EN, et al. The parathyroid hormone circadian rhythm is truly endogenous—a general clinical research center study. J. Clin Endocrinol Metab 1997; 82:281. 42. Samuels MH, Veldhais JD, Kramer P, et al. Episodic secretion of parathyroid hormone in postmenopausal women: assessment by deconvolution analysis and approximate entropy. J Bone Miner Res 1977; 12:616. 43. Mayer GP, Hurst JG. Sigmoidal relationship between parathyroid hormone secretion rate and plasma concentration in calves. Endocrinology 1978; 102:1036. 44. Shoback DM, Thatcher J, Leombruno R, Brown EM. Relationship between parathyroid hormone secretion and cytosolic calcium concentration in dispersed bovine parathyroid cells. Proc Natl Acad Sci U S A 1984; 81:3113. 45. Brown EM, Redgrave J, Thatcher J. Effect of the phorbol ester TPA on PTH secretion. Evidence for a role for protein kinase C in the control of PTH release. FEBS Lett 1984; 175:72. 46. Brown EM, Gambda G, Riccardi D, et al. Cloning and characterization of an extracellular Ca 2+-sensing receptor from bovine parathyroid. Nature 1993; 366:575. 47. Pollak MR, Brown EM, Chou Y-HW, et al. Mutations in the human Ca2+-sensing receptor gene cause familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Cell 1993; 75:1237. 48. Heath H III, Odelberg S, Jackson CE, et al. Clustered inactivating mutations and benign polymorphisms of the calcium receptor gene in familial benign hypocalciuric hypercalcemia suggest receptor functional domains. J Clin Endocrinol Metab 1996; 81:1312. 49. Cole DEC, Janicic N, Salisbury SR, Hendy, GN. Neonatal severe hyperparathyroidism, secondary hyperparathyroidism, and familial hypercalciuric hypercalcemia: multiple different phenotypes associated with an inactivating Alu insertion mutation of the calcium-sensing receptor gene. Am J Med Genet 1997; 71:202. 50. Pollak MR, Chou Y-HW, Marx SJ, et al. Familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Effects of mutant gene dosage on phenotype. J Clin Invest 1994; 93:1108. 51. Bai M, Pearce SHS, Kifor O, et al. In vivo and in vitro characterization of neonatal hyperparathyroidism resulting from a de novo, heterozygous mutation in the Ca 2+-sensing receptor gene: normal maternal calcium homeostasis as a cause of secondary hyperparathyroidism in familial benign hypocalciuric hypercalcemia. J Clin Invest 1997; 99:88. 52. Pearce SHS, Williamson C, Kifor O, et al. A familial syndrome of hypocalcemia with hypercalciuria due to mutations in the calcium-sensing receptor. N Engl J Med 1996; 335:1115. 53. Cole DEC, Peltekova VD, Rubin LA, et al. A986S polymorphism of the calcium-sensing receptor and circulating calcium concentrations. Lancet 1999; 353:112. 54. Nemeth EF, Steffey ME, Hammerland LG, et al. Calcimimetics with potent and selective activity on the parathyroid calcium receptor. Proc Natl Acad Sci U S A 1998; 95:4040. 55. Silverberg SJ, Bone GH III, Marriott TB, et al. Short-term inhibition of parathyroid hormone secretion by a calcium-receptor agonist in patients with primary hyperparathyroidism. N Engl J Med 1997; 337:1506. 56. Kovacs CS, Ho-Pao CL, Hunzelman JL, et al. Regulation of murine fetal-placental calcium metabolism by the calcium-sensing receptor. J Clin Invest 1998; 101:2812. 57. Brown AJ, Zhong M, Ritter C, et al. Loss of calcium responsiveness in cultured bovine parathyroid cells is associated with decreased calcium receptor expression. Biochem Biophys Res Comm 1995; 212:861. 58. Mithal A, Kifor O, Kifor I, et al. The reduced responsiveness of cultured bovine parathyroid cells to extracellular Ca2+ is associated with marked reduction in the expression of extracellular Ca(2+)-sensing receptor messenger ribonucleic acid and protein. Endocrinology 1995; 136:3087. 59. Russell J, Lettieri D, Sherwood LM. Direct regulation by calcium of cytoplasmic messenger ribonucleic acid coding for pre-proparathyroid hormone in isolated bovine parathyroid cells. J Clin Invest 1983; 72:1851. 60. Russell J, Sherwood LM. The effects of 1,25(OH)2 D3 and high calcium on transcription of the pre-proparathyroid hormone gene are direct. Trans Assoc Am Physicians 1987; 100:256. 61. Moallem E, Kilav R, Silver J, Naveh-Many T. RNA-protein binding and posttranscriptional regulation of parathyroid hormone gene expression by calcium and phosphate. J Biol Chem 1998; 273:5253. 62. Brown EM. Four-parameter model of the sigmoidal relationship between parathyroid hormone release and extracellular calcium concentration in normal and abnormal parathyroid tissue. Endocrinology 1983; 56:572. 63. Kifor O, Moore FD, Wang P, et al. Reduced immunostaining for the extracellular Ca2+-sensing receptor in primary and secondary hyperparathyroidism. J Clin Endocrinol Metab 1996; 81:1598. 64. Ho C, Conner DA, Pollack MR, et al. A mouse model of human familial hypocalciuric hypercalcemia and neonatal severe hyperparathyroidism. Nature Genet 1995; 11:389. 65. Habener JF, Potts JT Jr. Relative effectiveness of magnesium and calcium on the secretion and biosynthesis of parathyroid hormone in vitro. Endocrinology 1976; 98:197. 66. Rude RK, Oldham SB, Sharp CF Jr, Singer FR. Parathyroid hormone secretion in magnesium deficiency. J Clin Endocrinol Metab 1978; 47:800. 67. Cholst IN, Steinberg SF, Tropper PJ, et al. The influence of hypermagnesemia on serum calcium and parathyroid hormone levels in human subjects. N Engl J Med 1984; 310:1221. 68. Morrissey J, Rothstein M, Mayer G, Slatopolsky E. Suppression of parathyroid hormone by aluminum. Kidney Int 1983; 23:699. 69. Sherwood LM, Mayer GP, Ramberg DF, et al. Regulation of parathyroid hormone secretion: proportional control by calcium, lack of effect of phosphate. Endocrinology 1968; 83:1043. 70. Almaden Y, Canelejo A, Hernandez A, et al. Direct effect of phosphorous on PTH secretion from whole rat parathyroid glands in vitro. J Bone Miner Res 1996; 11:970. 71. Slatopolsky E, Finch J, Denda M, et al. Phosphorous restriction prevents parathyroid gland growth: high phosphorous directly stimulates PTH secretion in vitro. J Clin Invest 1996; 97:2534. 72. Abe M, Sherwood LM. Regulation of parathyroid hormone secretion by adenylate cyclase. Biochem Biophys Res Commun 1972; 48:396. 73. Brown EM, Gardner DG, Windek RA, Aurbach GD. Relationship of intracellular 3',5'-adenosine monophosphate accumulation to parathyroid hormone release from dispersed bovine parathyroid cells. Endocrinology 1978; 103:2323. 74. Heath H III. Biogenic amines and the secretion of parathyroid hormone and calcitonin. Endocr Rev 1980; 1:319. 75. Epstein S, Heath H III, Bell NH. Lack of influence of isoproterenol, propranolol and dopamine on immunoreactive parathyroid hormone and calcitonin in normal man. Calcif Tissue Int 1983; 35:32. 76. Fasciotto BH, Gorr S-U, De Franco DJ, et al. Pancreastatin, a presumed product of chromogranin-A (secretory protein-1) processing, inhibits secretion from porcine parathyroid cells in culture. Endocrinology 1989; 125:1617.

77. Drees BM, Rouse J, Johnson J, Hamilton JW. Bovine parathyroid glands secrete a 26-kDa N-terminal fragment of chromogranin-A which inhibits parathyroid cell secretion. Endocrinology 1991; 129:3381. 78. Silver J, Russell J, Sherwood LM. Regulation by vitamin D metabolites of messenger ribonucleic acid for preproparathyroid hormone in isolated bovine parathyroid cells. Proc Natl Acad Sci U S A 1985; 82:4270. 79. Hendy GN, Stotland MS, Grunbaum D, et al. Characteristics of secondary hyperparathyroidism in vitamin D deficient dogs. Am J Physiol 1989; 256:E765. 80. Mackey SL, Heymont JL, Kronenberg HM, Demay MB. Vitamin D receptor binding to the negative human parathyroid hormone vitamin D response element does not require the retinoid x receptor. Molec Endocrinol 1996; 10:298. 81. Kremer R, Bolivar I, Goltzman D, Hendy GN. Influence of calcium and 1,25-dihdroxycholecalciferol on proliferation and protooncogene expression in primary cultures of bovine parathyroid cells. Endocrinology 1989; 125:935. 82. Fucik RF, Krukeja SC, Hargis GK, et al. Effect of glucocorticoids on function of the parathyroid glands in man. J Clin Endocrinol Metab 1975; 40:152. 83. Silverman R, Yalow RS. Heterogeneity of parathyroid hormone: clinical and physiologic implications. J Clin Invest 1973; 52:1958. 84. Segre GV, Niall HD, Habener JF, Potts JT Jr. Metabolism of parathyroid hormone: physiological and clinical significance. Am J Med 1974; 56:774. 85. Arnaud CD, Goldsmith RS, Bordier PJ, et al. Influence of immunoheterogeneity of circulating parathyroid hormone on radioimmunoassays of serum in man. Am J Med 1974; 56:785. 86. Reiss E, Canterbury JM. Emerging concepts of the nature of circulating parathyroid hormones: implications for clinical research. Recent Prog Horm Res 1974; 30:391. 87. Neuman WF, Neuman MW, Lane K, et al. The metabolism of labeled parathyroid hormone V. Collected biological studies. Calcif Tissue Res 1975; 18:271. 88. Martin KJ, Hruska KA, Freitag JJ, et al. The peripheral metabolism of parathyroid hormone. N Engl J Med 1979; 301:1092. 89. Segre GV, D'Amour P, Hultman A, Potts JT Jr. Effects of hepatectomy, nephrectomy, and nephrectomy/uremia on the metabolism of parathyroid hormone in the rat. J Clin Invest 1981; 67:439. 90. Papapoulos SE, Hendy GN, Tomlinson S, et al. Clearance of exogenous parathyroid hormone in normal and uraemic man. Clin Endocrinol 1977; 7:211. 91. Goltzman D, Henderson B, Loveridge N. Cytochemical bioassay of parathyroid hormone: characteristics of the assay and analysis of circulating hormonal forms. J Clin Invest 1980; 65:1309. 92. DeLuca HF. Recent advances in the metabolism of vitamin D. Ann Rev Physiol 1981; 43:199. 93. Parsons JA, Rafferty B, Gray D, et al. Pharmacology of parathyroid hormone and some of its fragments and analogues. In: Talmage RV, Owen M, Parsons JA, eds. Calcium-regulating hormones. Amsterdam: Excerpta Medica, 1975:33. 94. Tregear GW, van Rietschoten J, Greene E, et al. Bovine parathyroid hormone: minimum chain length of synthetic peptide required for biological activity. Endocrinology 1973; 93:1349. 95. Kemp BE, Moseley JM, Rodda CP, et al. Parathyroid hormone–related protein of malignancy: active synthetic fragments. Science 1987; 23:1568. 96. Horiuchi N, Caulfield MP, Fisher JE, et al. Similarity of synthetic peptide from human tumor to parathyroid hormone in vivo and in vitro. Science 1987; 23:1566. 97. Rabbani SA, Mitchell J, Roy DR, et al. Influence of the amino-terminus on in vitro and in vivo biological activity of synthetic parathyroid hormone– like peptides of malignancy. Endocrinology 1988; 123:2709. 98. Goltzman D, Peytremann A, Callahan E, et al. Analysis of the requirements for parathyroid hormone action in renal membranes with the use of inhibiting analogues. J Biol Chem 1975; 250:3199. 99. Nussbaum SR, Rosenblatt M, Potts JT Jr. Parathyroid hormone renal receptor interactions: demonstration of two receptor binding domains. J Biol Chem 1980; 255:10183. 100. Abou-Samra A-B, Uneno S, Jüppner H, et al. Non-homologous sequences of parathyroid hormone and the parathyroid hormone related peptide bind to a common receptor on ROS 17/2.8 cells. Endocrinology 1989; 125:2215. 101. Caulfield MP, McKee RL, Goldman ME, et al. The bovine renal parathyroid hormone (PTH) receptor has equal affinity for two different amino acid sequences: The receptor binding domains of PTH and PTH-related protein are located within the 14-34 region. Endocrinology 1990; 127: 83. 102. Klaus W, Dieckmann T, Wray V, et al. Investigation of the solution structure of the human parathyroid hormone fragment (1-34) by 1H NMR spectroscopy, distance geometry, and molecular dynamics calculations. Biochemistry 1991; 30:6936. 103. Cohen FE, Strewler GJ, Bradley MS, et al. Analogues of parathyroid hormone modified at positions 3 and 6: Effects on receptor binding and activation of adenylyl cyclase in kidney and bone. J Biol Chem 1991; 266:1997. 104. Barden JA, Kemp BE. Stabilized NMR structure of the hypercalcemia of malignancy peptide PTHrP [Ala-26](1-34) amide. Biochim Biophys Acta 1994; 1208:256. 105. Wray V, Federau T, Gronwald W, et al. The structure of human parathyroid hormone from a study of fragments in solution using 1H NMR spectroscopy and its biological implications. Biochemistry 1994; 33:1684. 106. Neugebauer W, Barbier J-R, Sung WL, et al. Solution structure and adenylyl cyclase stimulating activities of C-terminal truncated human parathyroid hormone analogues. Biochemistry 1995; 34:8835. 107. Maretto S, Mammi S, Bissacco E, et al. Mono- and bicyclic analogs of parathyroid hormone-related protein. 2. Conformational analysis of antagonists by CD, NMR, and distance geometry calculations. Biochemistry 1997; 36:3300. 108. Barbier J-R, Neugebauer W, Morley P, et al. Bioactivities and secondary stuctures of constrained analogues of human parathyroid hormone: cyclic lactams of the receptor binding region. J Medicinal Chem 1997; 40:1373. 109. Vickery BH, Avnur Z, Cheng Y, et al. RS-66271, a C-terminally substituted analog of human parathyroid hormone-related protein (1-34), increases trabecular and cortical bone in ovariectomized, osteopenic rats. J Bone Miner Res 1996; 11:1943. 110. Juppner H, Abou-Samra AB, Freeman MW, et al. A G protein–linked receptor for parathyroid hormone and parathyroid hormone–related peptide. Science 1991; 254:1024. 111. Abou-Samra AB, Jüppner H, Force T, et al. Expression cloning of a parathyroid hormone/parathyroid hormone-related peptide receptor from rat osteoblast-like cells: a single receptor stimulates intracellular accumulation of both cAMP and inositol triphosphate and increases intracellular free calcium. Proc Natl Acad Sci U S A 1992; 89:2732. 112. Pausova Z, Bourdon J, Clayton D, et al. Cloning of a parathyroid hormone/parathyroid hormone-related peptide receptor (PTHR) cDNA from a rat osteosarcoma (UMR 106) cell line: chromosomal assignment of the gene in the human, mouse, and rat genomes. Genomics 1994; 20:20. 113. Schipani E, Harga H, Karaplis AC, et al. Identical complementary deoxyribonucleic acids encode a human renal and bone parathyroid hormone (PTH)/PTH-related peptide receptor. Endocrinology 1993; 132:2157. 114. Fraher LJ, Hodsman AB, Jonas K, et al. A comparison of the in vivo biochemical responses to exogenous parathyroid hormone-(1–34) [PTH-(1–34)] and PTH-related peptide-(1–34) in man. J Clin Endocrinol Metab 1992; 75:417. 114a. Fraher LJ, Avram R, Watson PH, et al. Comparison of the biochemical responses to human parathyroid hormone-(1-31)NH 2 and hPTH-(1-34) in healthy humans. J Clin Endocrinol Metab 1999; 84:2739. 115. McCuaig KA, Clarke JC, White JH. Molecular cloning of the gene encoding the mouse parathyroid hormone/parathyroid hormone related peptide receptor. Proc Natl Acad Sci U S A 1994; 91:5051. 116. Bettoun JD, Minagawa M, Hendy GN, et al. Developmental upregulation of human parathyroid hormone (PTH)/PTH-related peptide receptor gene expression from conserved and human-specific promoters. J Clin Invest 1998; 102:958. 116a. Bettoun JD, Kwan MY, Minagawa M, et al. Methylation patterns of human parathyroid hormone (PTH)/PTH-related peptide receptor gene promoters are established several weeks prior to onset of their function. Biochem Biophys Res Commun 2000; 267:482. 117. Urena P, Kong X-F, Abou-Samra A-B, et al. Parathyroid hormone (PTH)/PTH-related peptide receptor messenger ribonucleic acids are widely distributed in rat tissues. Endocrinology 1993; 133:617. 118. Usdin TB, Gruber C, Bonner TI. Identification and functional expression of a receptor selectively recognizing parathyroid hormone, the PTH2 receptor. J Biol Chem 1995; 270:15455. 119. Lee CW, Gardella TJ, Abou-Samra AB, et al. Role of the extracellular regions of the PTH/PTHrP receptor in hormone-binding. Endocrinology 1994; 135:1488. 120. Lee CW, Luck MD, Jüppner H, et al. Homolog-scanning mutagenesis of the parathyroid hormone (PTH) receptor reveals PTH-(1-34) binding determinants in the third extracellular loop. Mol Endocrinol 1995; 9:1269. 121. Iida KA, Guo J, Takemura M, et al. Mutations in the second cytoplasmic loop of the rat parathyroid hormone (PTH)/PTH-related protein receptor result in selective loss of PTH-stimulated phospholipase C activity. J Biol Chem 1997; 272:6882. 122. Huang Z, Chen Y, Nissenson RA. The cytoplasmic tail of the G protein-coupled receptor for parathyroid hormone and parathyroid hormone-related protein contains positive and negative signals for endocytosis. J Biol Chem 1996; 270:151. 123. Mahoney CA, Nissenson RA. Canine renal receptors for parathyroid hormone: down-regulation in vivo by exogenous parathyroid hormone. J Clin Invest 1983; 72:411. 124. Forte LR, Langeluttig SG, Poelling RE, Thomas ML. Renal parathyroid hormone receptors in the chick: downregulation in secondary hyperparathyroid animal models. Am J Physiol 1982; 242:E154. 125. Tomlinson S, Hendy GN, Pemberton DM, O'Riordan JLH. Reversible resistance to the renal action of parathyroid hormone in man. Clin Sci Mol Med 1976; 51:59. 126. Schipani E, Weinstein LS, Bergwitz C, et al. Pseudohypoparathyroidism type 1b is not caused by a defect in the coding exons of the human parathyroid hormone (PTH)/PTH-related peptide gene. J Clin Endocrinol Metab 1995; 80:1611. 127. Bettoun JD, Minagawa M, Kwan MY, et al. Cloning and characterization of the promoter regions of the human parathyroid hormone (PTH)/PTH-related peptide receptor gene: analysis of deoxyribonucleic acid from normal subjects and patients with pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 1997; 82:1031. 128. Jüppner H, Schipani E, Bastepe M, et al. The gene responsible for pseudohypoparathyroidism type 1b is paternally imprinted and maps in four unrelated kindreds to chromosome 20q13.3. Proc Natl Acad Sci U S A 1998; 95:11798. 129. Yu S, Yu D, Lee E, et al. Variable and tissue-specific hormone resistance in heterotrimeric Gs protein a-subunit (Gs a) knockout mice is due to tissue-specific imprinting of the G s a gene. Proc Natl Acad Sci U S A 1998; 95:8715. 130. Hayward BE, Moran V, Strain L, Bonthron DT. Bidirectional imprinting of a single gene: GNAS1 encodes maternally, paternally, and biallelically derived proteins. Proc Natl Acad Sci U S A 1998; 95:15475. 131. Jansen M. über atypische Chondrodystrophie (Achondroplasie) und über eine noch nicht beschriebene angeborene Wachstumsstö;rung des Knochen-systems: Metaphysäre Dysostosis. Orthop Chir 1934; 61:253. 132. Schipani E, Langman CB, Parfitt AM, et al. Constitutively activated receptors for parathyroid hormone-related peptide in Jansen's metaphyseal chondrodysplasia. N Engl J Med 1996; 335:708. 133. Blomstrand S, Claesson I, Save-Soderbergh J. A case of lethal congenital dwarfism with accelerated skeletal maturation. Pediatr Radiol 1985; 15:141. 134. Zhang P, Jobert A-S, Couvineau A, Silve C. A homozygous inactivating mutation in the parathyroid hormone/parathyroid hormone-related peptide receptor causing Blomstrand chondrodysplasia. J Clin Endocrinol Metab 1998; 83:3365. 135. Jobert A-S, Zhang P, Couvineau A, et al. Absence of functional receptors for parathyroid hormone and parathyroid hormone-related peptide in Blom-strand chondrodysplasia. J Clin Invest 1998; 102:34. 135a. Usdin TB, Hoare SR, Wang T, et al. TIP39: a new neuropeptide and PTH2-receptor agonist from hypothalamus. Nature Neuroscience 1999; 2:941. 136. Brenza HL, Kimmel-Jehan C, Jehan F, et al. Parathyroid hormone activation of the 25-hydroxyvitamin D3-1 alpha-hydroxylase gene promoter. Proc Natl Acad Sci U S A 1998; 95:1387. 137. Livesey SA, Kemp BE, Re CA, et al. Selective hormonal activation of cyclic AMP-dependent protein-kinase isoenzymes in normal and malignant osteoblasts. J Biol Chem 1982; 257:14983. 138. Pearman AT, Chou WY, Bergman KD, et al. Parathyroid hormone induces c-fos promoter activity in osteoblastic cells through phosphorylated cAMP response element (CRE)-binding protein to the major CRE. J Biol Chem 1996; 271:25715. 139. Stern PH. Cationic agonists and antagonists of bone resorption. In: Cohn DV, Fujita T, Potts JT Jr, Talmage RV, eds. Endocrine control of bone and calcium metabolism. Amsterdam: Excerpta Medica, 1984:109. 140. Hruska KA, Moskowitz D, Esbrit P, et al. Stimulation of inositol triphosphate and diacylglycerol production in renal tubular cells by parathyroid hormone. J Clin Invest 1987; 79:230. 141. Bingham P, Brazell I, Owen M. The effect of parathyroid extract on cellular activity and plasma calcium levels in vivo. J Endocrinol 1969; 45:387. 142. Holtrop ME, Raisz LG, Simmons HA. The effect of parathyroid hormone, colchicine and calcitonin on the ultrastructure and activity of osteoblasts in organ culture. J Cell Biol 1974; 60:346. 143. Rouleau MF, Mitchell J, Goltzman D. In vivo distribution of parathyroid hormone receptors in bone. Evidence that a predominant osseous target cell is not the mature osteoblast. Endocrinology 1988; 123:187. 144. Rouleau MF, Warshawsky H, Goltzman D. Parathyroid hormone binding in vivo to renal, hepatic and skeletal tissues of the rat using a radioautographic approach. Endocrinology 1986; 118:919. 145. Rizzoli RE, Somerman M, Murray TM, Aurbach GD. Binding of radioiodinated parathyroid hormone to cloned bone cells. Endocrinology 1983; 113:1832.

146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167.

Amizuka N, Karaplis AC, Henderson JE, et al. Haploinsufficiency of parathyroid hormone-related peptide (PTHrP) results in abnormal postnatal bone development. Dev Biol 1996; 175:166. Takahashi N, Akatsu T, Udagawa N, et al. Osteoblastic cells are involved in osteoclast formation. Endocrinology 1988; 123:2600. Lacey DL, Timms E, Tan HL, et al. Osteoprotegerin ligand is a cytokine that regulates osteoclast differentiation and activation. Cell 1998; 93:165. Kong Y-Y, Yoshida H, Sarosi L, et al. OPGL is a key regulator of osteoclastogenesis, lymphocyte development and lymphnode organogenesis. Nature 1999; 397:315. Simonet WS, Lacey DL, Dunstan CR, et al. Osteoprotegerin: A novel secreted protein involved in the regulation of bone density. Cell 1997; 89:309. Parsons JA, Potts JT Jr. Physiology and chemistry of parathyroid hormone. In: MacIntyre I, ed. Clinics in endocrinology and metabolism, vol 1. Calcium metabolism and bone disease. London: WB Saunders, 1972:33. Tam CS, Heersche JNM, Murray TM, Parsons JA. Parathyroid hormone stimulates the bone apposition rate independently of its resorptive action: differential effects of intermittent and continual administration. Endocrinology 1982; 110:506. Corral DA, Amling M, Priemel M, et al. Dissociation between bone resorption and bone formation in osteopenic transgenic mice. Proc Natl Acad Sci U S A 1998; 95:13835. Bijvoet OLM. Kidney function in calcium and phosphate metabolism. In: Avioli LV, Krane SM, eds. Metabolic bone disease, vol 1. New York: Academic Press, 1977:48. Hartmann CM, Hewson AS, Kos CH, et al. Structure of murine and human renal type II Na+-phosphate cotransporter genes (Npt2 and NPT2). Proc Natl Acad Sci U S A 1996; 93:7409. Azarani A, Goltzman D, Orlowski J. Structurally diverse N-terminal peptides of parathyroid hormone (PTH) and PTH-related peptide (PTHRP) inhibit the Na +/H+ exchanger NHE3 isoform by binding to the PTH/PTHRP receptor type I and activating distinct signalling pathways. J Biol Chem 1996; 271:14931. Morel F. Regulation of kidney functions by hormones: a new approach. Recent Prog Horm Res 1983; 39:271. Agus ZS, Wasserstein A, Goldfarb S. PTH, calcitonin, cyclic nucleotides and the kidney. Annu Rev Physiol 1981; 43:583. Nordin BEC, Peacock M. Role of the kidney in regulation of plasma calcium. Lancet 1969; 2:1280. Amizuka N, Warshawsky H, Henderson JE, et al. Parathyroid hormone-related peptide–depleted mice show abnormal epiphyseal cartilage development and altered endochondral bone formation J Cell Biol 1994; 126:1611. Marx SJ, Sharp ME, Krudy A, et al. Radioimmunoassay for the middle region of human parathyroid hormone: studies with a radioiodinated synthetic peptide. J Clin Endocrinol Metab 1981; 53:76. Papapoulos SE, Manning RM, Hendy GN, et al. Studies of circulating parathyroid hormone in man using a homologous amino-terminal specific immunoradiometric assay. Clin Endocrinol 1980; 13:57. Nussbaum SR, Zahnadnik RJ, Labigne JR, et al. A highly sensitive two-site immunoradiometric assay of parathyrin (PTH) and its clinical utility in evaluating patients with hypercalcemia. Clin Chem 1987; 33:1364. Broadus AE. Nephrogenous cyclic AMP. Recent Prog Horm Res 1981; 37:667. Nissenson RA, Abbott SR, Teitelbaum AP, et al. Endogenous biologically active human parathyroid hormone: measurement by a guanyl nucleotide-amplified renal adenylate cyclase assay. J Clin Endocrinol Metab 1981; 52:840. Sato K, Han DC, Ozawa M, et al. A highly sensitive bioassay for PTH using ROS 17/2.8 subclonal cells. Acta Endocrinol (Copenh) 1987; 116:113. Chambers DJ, Dunham J, Zanelli JM, et al. A sensitive bioassay of parathyroid hormone in plasma. Clin Endocrinol 1978; 9:375.

CHAPTER 52 PARATHYROID HORMONE-RELATED PROTEIN Principles and Practice of Endocrinology and Metabolism

CHAPTER 52 PARATHYROID HORMONE-RELATED PROTEIN GORDON J. STREWLER Structure and Biologic Properties Structure and Processing of Parathyroid Hormone-Related Protein and its Gene Secreted Forms: Peptide Heterogeneity Immunoassays Parathyroid Hormone-Related Protein Receptors Biologic Properties: Action on Bone and Kidney Role of Parathyroid Hormone- Related Protein in Hypercalcemia Tumors Secreting Parathyroid Hormone-Related Protein Pathogenetic Role of Parathyroid Hormone-Related Protein Use of Parathyroid Hormone Related Protein Assays in the Differential Diagnosis of Hypercalcemia Role of Parathyroid Hormone Related Protein in Physiology Cartilage Breast Skin Teeth Smooth Muscle Uterus and Placenta Chapter References

Parathyroid hormone–related protein (PTHrP), sometimes referred to as parathyroid hormone–like protein, is the sister of PTH. Originally identified as the cause of humoral hypercalcemia in malignancy, PTHrP has a distinct set of physiologic functions that are unrelated to the regulation of systemic calcium homeostasis but rival the actions of PTH in their importance. It has been recognized since Fuller Albright's time that in some ways patients with malignant tumors causing hypercalcemia resemble patients with primary hyperparathyroidism (pHPT). Malignancy-associated hypercalcemia is characterized not only by humorally mediated bone resorption but also by diminished tubule resorption of phosphate with consequent phosphaturia and hypophosphatemia. When it was recognized that the disorder also produced an increase in nephrogenous cyclic adenosine monophosphate (cAMP), which is the component of urinary cAMP secreted into the urine from the renal tubule,1 it became clear that a humoral factor in patients with malignancy-associated hypercalcemia was mimicking PTH at the kidney. Increased nephrogenous cAMP was previously thought to be unique to hyperparathyroidism, reflecting increased secretion of cAMP from the renal tubule, where cAMP is the intracellular second messenger for PTH. Once it was understood that increased cAMP concentrations in kidney or bone cells could be used as a bioassay to detect the humoral factor secreted by malignant tumors, the substance responsible for malignancy-associated hypercalcemia was purified and shown to be a protein that was structurally related to PTH.2,3 and 4

STRUCTURE AND BIOLOGIC PROPERTIES STRUCTURE AND PROCESSING OF PARATHYROID HORMONE-RELATED PROTEIN AND ITS GENE PTHrP is a protein of 139 to 173 amino acids.5,6 and 7 It resembles PTH in primary sequence only at the aminoterminus, where 8 of the first 13 amino acids in the two peptides are identical (Fig. 52-1). Although this is a limited region of homology, it is a critical region of both peptides; it is not required for binding but it is necessary for receptors occupied by either hormone to activate adenylate cyclase and hence to produce cAMP as a second messenger. Thus, the homology of PTH and PTHrP at the aminoterminal end is responsible for their shared biologic properties (see later in this chapter).

FIGURE 52-1. Primary structures of the aminoterminal part of parathyroid hormone–related protein and of parathyroid hormone. Compared are the human sequences 1–34. Identical amino acids are highlighted.

The gene for PTHrP is located on human chromosome 12. It is considerably more complex than the PTH gene, which is located on chromosome 11, with three promoters and four alternatively spliced exons upstream of the coding sequence.8 The complexity of the promoter region allows for considerable flexibility in the regulation of PTHrP gene transcription. There is evidence that the promoters are used differentially in different tissues8 and by the viral transactivating protein tax.9 Most of the prepro sequence of PTHrP is encoded on one exon, with the last 2 amino acids of the prepro sequence and 139 amino acids of the mature peptide encoded on a second exon. The splice junction between these two coding exons is precisely conserved between PTHrP and PTH. Downstream of the main coding sequence are two additional exons, which are alternatively spliced to contribute distinct carboxyterminal ends to the isoforms of PTHrP and distinct 3' noncoding sequences. The protein isoforms encoded by these exons are identical through amino acid 139 but comprise 139, 141, or 173 amino acids in toto. This downstream complexity of the PTHrP gene may also be used for regulation. The three transcript families with different 3' untranslated sequences have markedly different halflives, ranging from 20 minutes to ~7 hours. In addition, their stability is differentially regulated. Exposure of cells to transforming growth factor-b specifically increases the halflife of the most unstable transcript, which encodes the 139 amino acid form of the protein. It is clear from conservation of sequence, gene structure, and chromosomal localization that PTHrP and PTH arose from a common ancestral gene. In addition to the striking resemblance of their aminoterminal sequence and the conservation of the intron–exon boundaries between the two major coding exons, each gene is flanked by duplicated genes for lactate dehydrogenase. It is believed that chromosome 12, on which the PTHrP gene is located, and chromosome 11, where the PTH gene resides, arose by an ancient duplication in which the ancestral PTH/PTHrP gene was evidently copied along with its neighbors. SECRETED FORMS: PEPTIDE HETEROGENEITY The three mature PTHrP peptides predicted by cDNA are 139, 141, and 173 amino acids long. It is not clear whether any of them are secreted as intact proteins, although it seems likely they are. AMINOTERMINAL FRAGMENTS The molecular size of PTHrP containing the bioactive aminoterminal part found in plasma and in tumor extracts ranges anywhere from 6 to 18 kDa,2,3 and 4,10 and, apparently, several different aminoterminal fragments of PTHrP are secreted.11 Three different cell types (renal carcinoma cells, PTHrP-transfected RIN-141 cells, and normal keratinocytes) use prohormone convertases to cleave PTHrP after arginine-37 to produce an aminoterminal fragment that is probably processed to PTHrP(1–36)12 (Fig. 52-2). This aminoterminal PTHrP fragment contains the PTH-like region and could account completely for the PTH-like effects of PTHrP, as

discussed later.

FIGURE 52-2. To p, Structural features of PTHrP. The PTH-homologous domain PTHrP(1–13) is delineated by hatched lines ( ), and potential cleavage sites are shown as vertical lines or crosshatched regions ( ). Middle, Peptides known or postulated to be derived from PTHrP. Bottom, Proven or postulated sites of action of individual PTHrP peptides are shown beneath each peptide.

MIDREGIONAL FRAGMENTS The intracellular cleavage step at arginine-37 also produces midregional PTHrP fragments of ~50 to 70 amino acids beginning with alanine-3812 (see Fig. 52-2). The carboxytermini of these fragments are located at amino acids 94, 95, and 101 in the highly basic region PTHrP(88–106).13 Midregional fragments of PTHrP generated intracellularly can apparently enter the regulated pathway of peptide secretion and become packaged in dense neurosecretory granules, suggesting that they may be under separate secretory control from aminoterminal fragments.12 Midregional PTHrP fragments of similar size are the predominant circulating forms in the plasma of patients withhumoral hypercalcemia of malignancy.14 These midregional PTHrPs play a physiologic role in transplacental calcium transport (see later), and synthetic midregion peptides also appear to have bioactivities in squamous carcinoma cells. CARBOXYTERMINAL FRAGMENTS The size of the midregional fragments suggests the existence of additional carboxyterminal PTHrP fragments resulting from cleavage at the multibasic amino acid clusters at amino acids 88–106. Because the most carboxyterminal cleavage site is at amino acid 101, which precedes the lysine-arginine cluster at PTHrP(102–106), it is likely that carboxyterminal peptides such as PTHrP(107-139) are produced and secreted. Peptides with PTHrP(109–138) but not PTHrP(1–36) immunoreactivity have been found in sera of patients with humoral hypercalcemia of malignancy, and a peptide with PTHrP(109–138) immunoreactivity circulates as a separate peptide that accumulates in patients with chronic renal failure.10,15,16 The high degree of conservation between human, rat, and chicken PTHrP in the portion up to amino acid 111 suggests some biologic relevance of such peptides. The very carboxyterminal part of the conserved region, PTHrP(107–111), was found in several studies to be a potent inhibitor of osteoclastic bone resorption in vitro,17,18 and has been given the name osteostatin, although this result has not been confirmed in all bone resorption assays.19 The corresponding synthetic peptides (PTHrP[107–111] and PTHrP[107–139]) induce calcium transients in hippocampal neurons.20 PTHrP is posttranslationally modified in additional ways21; for example, PTHrP secreted by keratinocytes is glycosylated posttranslationally.22 Whether this is also the case with other cell types and with PTHrP forms circulating in plasma is unknown. The findings that cells process PTHrP to multiple peptide fragments, that some of these fragments may be under separate secretory control, and that one or more (in addition to its PTH-like aminoterminus) has its own bioactivity indicate that PTHrP is, like the adrenocorticotropic hormone-melanocyte-stimulating hormone–endorphin precursor pro-opiomelano-cortin, a polyprotein precursor for multiple bioactive peptides.23 The pattern of posttranslational modification may be tissue-specific, and this would add another dimension of specificity to the secreted forms of PTHrP. IMMUNOASSAYS Because the known PTH-like bioactivity of PTHrP is located within the first 34 amino acids, most immunoassays have been directed against the aminoterminal sequence 1–34 to 1–40 of PTHrP in radioimmunoassays (RIAs)24,25 and 26 or against larger fragments containing aminoterminal PTHrP in immunoradiometric assays.16,27 Because the amounts of aminoterminal PTHrP extractable from normal serum are very low (–2.5 but 15 mg/dL).48 Approximately 25% of patients reported with this form of primary hyperparathyroidism have a previous history of mild hypercalcemia. Patients with acute primary hyperparathyroidism have extremely high serum PTH levels. In one review, the serum calcium value for 43 patients was 17.5 ± 2.1 mg/dL (mean ± standard deviation). In marked contrast to patients with asymptomatic primary hyperparathyroidism, virtually all patients were symptomatic. Unlike primary hyperparathyroidism, acute primary hyperparathyroidism involved both bone and kidney in one-half of the patients. More than two-thirds of patients had nephrolithiasis or nephrocalcinosis. The clinician must recognize that life-threatening hypercalcemia may be the result of acute primary hyperparathyroidism, because this is a curable disease. Another clinical presentation of primary hyperparathyroidism is that associated with a limited set of complications. As expected, the skeleton, kidneys, or neuromuscular system is most commonly involved. Primary hyperparathyroidism may also occur in a hereditary syndrome in association with MEN1 or MEN2A or simply as FIPH. Patients with FIPH do not have any manifestations of MEN. Patients from these kindreds often present at a younger age than the typical individual with sporadic primary hyperparathyroidism. Their disease generally affects more than one gland and is often recurrent. Because of the high recurrence rate after removal of a single adenoma, subtotal parathyroidectomy is the preferred approach in such individuals.49,50 Neonatal primary hyperparathyroidism also occurs (see Chap. 70). Another familial form of primary hyperparathyroidism is characterized by distinctive cystic pathologic features.51 Patients from these kindreds typically develop recurrent primary hyperparathyroidism resulting from another cystic adenoma years after successful removal of the first adenoma. No other endocrine tumors have been described in this syndrome, but several fibrous maxillary or mandibular tumors have been seen. Results of the physical examination of patients with primary hyperparathyroidism are not noteworthy. The clinical sign of calcium deposition in the cornea, band keratopathy, is rarely seen (see Chap. 215). Similarly, palpation of an enlarged parathyroid gland is unusual unless parathyroid carcinoma is present.

LABORATORY EVALUATION The diagnosis of primary hyperparathyroidism depends on laboratory tests. The serum calcium concentration is virtually always elevated. The serum phosphorus value may be decreased, but it is usually normal, although typically in the lower range of normal. Newer markers of bone metabolism are useful to evaluate the presence of underlying bone involvement in cases of primary hyperparathyroidism. Markers of bone formation and resorption tend to be elevated, even in asymptomatic primary hyperparathyroidism. Whether quantification of these markers is useful to predict the extent of PTH-dependent skeletal tumors or whether they are useful predictors of the course of the skeletal involvement is not yet certain, however. The bone formation marker, total alkaline phosphatase activity, which is readily available on routine multichannel panels, continues to be a useful marker in primary hyperparathyroidism,52 and its bone-specific isoenzyme is often clearly elevated in patients with mild primary hyperparathyroidism.53,54 Osteocalcin is also generally increased in patients with primary hyperparathyroidism.54,55 and 56 Bone resorption is reflected in a product of osteoclast activity, tartrate-resistant acid phosphatase (TRAP), and collagen breakdown products such as hydroxyproline, hydroxypyridinium cross-links of collagen, and the small N- and C-telopep-tides of collagen metabolism. The oldest known marker of bone resorption,57,58 urinary hydroxyproline excretion, is frankly elevated in patients with osteitis fibrosa cystica. However, in patients with mild, asymptomatic primary hyperparathyroidism, values are often entirely normal. The hydroxypyridinium cross-links of collagen have replaced hydroxyproline excretion as markers of bone resorption used in evaluating primary hyperparathyroidism (see Chap. 56). These cross-links of collagen (pyridinoline and deoxypyridinoline) are mildly elevated.59 Studies of TRAP levels in primary hyperparathyroidism are limited, although levels have been shown to be elevated.60 Bone sialoprotein is a phosphorylated glycoprotein that makes up b5% to 10% of the noncollagenous protein of bone. It is elevated in high-turnover states and reflects, in part, bone resorption. In primary hyperparathyroidism, bone sialoprotein levels are elevated and correlate with urinary pyridinoline and deoxypyridinoline.61 Although their measurement is not mandatory for the evaluation of the patient with primary

hyperparathyroidism, bone marker levels can provide useful information about the extent to which the skeleton is involved in the hyperparathyroid process. PARATHYROID HORMONE MEASUREMENT The diagnosis of primary hyperparathyroidism is substantiated by measurement of the concentration of PTH. For most patients with primary hyperparathyroidism, assays show frankly elevated levels of PTH associated with the sine qua non of the disease, hypercalcemia. Fragment assays for circulating midregion or COOH-terminal forms of PTH reliably show elevations by radioimmunoassay. Circulating intact PTH can be measured by immunoradiometric or immunochemiluminometric assays (Fig. 58-4; see Chap. 51).62,63 Elevated levels of circulating, intact PTH occur in b90% of patients with primary hyperparathyroidism. The remaining 10% of patients have levels that are in the upper range of normal, concentrations that are inappropriately high in the face of hypercalcemia. If these patients with “normal” levels of PTH are monitored over time, they invariably show elevated levels.

FIGURE 58-4. Comparison of midregion radioimmunoassay and immunoradiometric assay for parathyroid hormone (PTH) in primary hyperparathyroidism and hypercalcemia associated with malignancy. (From Nussbaum SR, Zahradnik RJ, Lavigne JR, et al. Highly sensitive two-site immunoradiometric assay of parathyrin and its clinical utility in evaluating patients with hypercalcemia. Clin Chem 1987; 33:1364.)

In addition to helping to establish the diagnosis of primary hyperparathyroidism, the assays for PTH help to differentiate it from the other most common cause of hypercalcemia, malignancy-associated hypercalcemia. In malignancy-associated hypercalcemia, even that associated with the production of parathyroid hormone–related protein (PTHrP), PTH levels are suppressed. This is because PTHrP is not recognized in assays for PTH (see Chap. 52). In virtually all other hypercalcemic states, the parathyroid glands are normally suppressed, and levels of PTH are not detected in the circulation. OTHER TESTS As a reflection of the increased concentration of biologically active PTH in the circulation of patients with primary hyperparathyroidism, total and nephrogenous urinary cyclic adenosine monophosphate (cAMP) levels are elevated.64 However, with clinically proven assays for PTH, urinary cAMP measurement does not offer any additional information. Urinary calcium excretion is not useful for diagnostic purposes but does help in the overall assessment of the patient and in the process of decision making for surgery. Serum 1,25-dihydroxyvitamin D [1,25(OH)2D] may be elevated in primary hyperparathyroidism, reflecting an action of PTH to stimulate the conversion of 25-hydroxyvitamin D [25(OH)D] to 1,25(OH)2D. Further insight into the functional abnormality of primary hyperparathyroidism derives from the observation that patients with this disorder convert more 25(OH)D to 1,25(OH)2D than do normal people.65 The level of 1,25(OH)2D in primary hyperparathyroidism may reflect the actions of PTH on the renal 1a-hydroxylase enzyme. However, the elevated 1,25(OH)2D level in primary hyperparathyroidism is not specific. Many patients with primary hyperparathyroidism do not show elevations in this vitamin D metabolite, and other disorders associated with hypercalcemia (e.g., certain granulomatous diseases, lymphomas, sarcoidosis, vitamin D toxicity) may be associated with elevations in the 1,25(OH)2D concentration. Patients with primary hyperparathyroidism demonstrate characteristic histopathologic changes of the hyperparathyroid state on bone biopsy (see Chap. 55). Presently, bone biopsy has little clinical utility in management of this disease.

DIAGNOSIS The diagnostic hallmarks of primary hyperparathyroidism are hypercalcemia and an elevated level of PTH. Just as a normal circulating PTH concentration is a relatively unusual finding in primary hyperparathyroidism, a normal level of serum calcium is also most unusual in primary hyperparathyroidism. When this occurs, the circulating serum proteins may be low, and the “normal” serum calcium value may be associated with an elevated ionized calcium concentration (see Chap. 49). Only in this unusual situation of suspected primary hyperparathyroidism—elevated PTH and normal serum calcium—may ionized serum calcium values be a valuable adjunct to other diagnostic tests. Nonetheless, primary hyperparathyroidism is seen occasionally with a truly normal serum calcium concentration. The name normocalcemic primary hyperparathyroidism was given to this clinical presentation; over time, however, the serum calcium level does become elevated.66 Presumably, this situation reflects increased parathyroid glandular activity that has not yet caused frank hypercalcemia. Rarely can primary hyperparathyroidism be shown to be present in an individual for whom serum calcium determinations are always normal. In a small number of patients with primary hyperparathyroidism, the circulating PTH concentration is not frankly elevated. Values are instead in the upper range of normal. The PTH level is, however, inappropriately high given the patient's hypercalcemia. Any patient with hypercalcemia should have suppressed levels of PTH unless primary hyperparathyroidism is present. Exceptions to this rule include patients with FHH and those who are taking thiazide diuretics or lithium. Some data also suggest that younger individuals (i.e., younger than 45 years of age) may normally have lower levels of PTH than their older counterparts.67 Thus, a value at the upper end of the normal range in a 40-year-old patient with hypercalcemia is probably indicative of hyperparathyroidism. If the serum calcium level is consistently normal when the PTH concentration is elevated, the physician must consider a secondary cause for the hyperparathyroid state. The distinction between primary and secondary hyperparathyroidism is made readily. Parathyroid glandular overactivity in secondary hyperparathyroidism has a nonparathyroid cause, a condition associated with a depressed serum calcium level. Parathyroid tissues respond to this hypocalcemic stimulus with an increased secretory rate. The increased PTH concentration returns the serum calcium level toward normal. Secondary hyperparathyroidism has many causes (see Chap. 61 and Chap. 63), but usually it is related to renal insufficiency or to gastrointestinal tract disorders in which malabsorption of vitamin D, calcium, or both occurs. All four parathyroid glands respond to hypocalcemia with the development of hyperplasia. In renal disease, if the hypocalcemic signal is uncontrolled and prolonged, the parathyroid tissue may become relatively autonomous, a condition called tertiary hyperparathyroidism, in which extreme elevations of PTH, sometimes >10 times normal, occur with hypercalcemia. Before determination of PTH became the most useful test in the differential diagnosis of hypercalcemia, several other tests and measurements were used that now have only historic interest. These tests include the decreased basal tubular reabsorption of phosphate, the lack of effect of exogenous PTH on the tubular reabsorption of phosphate, an increased chloride/bicarbonate ratio, and the failure to suppress serum calcium with prednisone. None of these maneuvers is of diagnostic value and they are not used to establish the diagnosis of primary hyperparathyroidism.

DIFFERENTIAL DIAGNOSIS The differential diagnosis of hypercalcemia is covered in Chapter 59, but several points are particularly relevant to the discussion of primary hyperparathyroidism. If previous medical records are available, the patient may be found to have had serum calcium levels at the upper limits of normal before frank hypercalcemia became evident. LITHIUM USE Lithium administration has been associated with hypercalcemia and an apparent hyperparathyroid state. In vitro studies suggest that lithium may alter the calcium

setpoint for calcium-mediated inhibition of PTH secretion.68 Anecdotal clinical reports describe reversible hyperparathyroidism in patients treated with lithium. Confounding variables, such as diuretic therapy or renal failure, preclude definite conclusions in many of these case reports. Lithium treatment has been associated with hypercalcemia, hypermagnesemia, and reduced urinary calcium.69 A controlled study involving normal volunteers assessed lithium effects on PTH secretion. No significant difference was found in serum calcium, plasma PTH, or nephrogenous cAMP measurements after calcium infusion in normal volunteers, off or on lithium therapy.70 These data suggest that clinically relevant lithium levels do not alter the calcium setpoint of PTH release, but evidence suggests that the setpoint for calcium may be altered short term and that PTH levels may rise over time.71 Practically, the diagnosis of primary hyperparathyroidism is on firmer grounds if lithium can be withdrawn and the patient shown to have persistent hypercalcemia over the ensuing several months. THIAZIDE USE Another medication in common use, thiazide diuretics, may be associated with hypercalcemia.72 The mechanism for hypercalcemia is related to several factors: reduced plasma volume, increased proximal tubular reabsorption of calcium, and perhaps activity at the level of the parathyroid glands themselves. Some patients who develop hypercalcemia in association with thiazide diuretic therapy are ultimately shown to have primary hyperparathyroidism. In these patients, the hypercalcemia persists after thiazides are withdrawn. Other patients who develop hypercalcemia during thiazide therapy show a return to normal of the serum calcium and serum PTH levels. The diagnosis of primary hyperparathyroidism cannot be made with confidence in hypercalcemic patients receiving thiazides unless the calcium and parathyroid levels are still elevated 2 to 3 months after the diuretic is discontinued. COEXISTENCE OF TWO CAUSES OF HYPERCALCEMIA Because primary hyperparathyroidism is a relatively common disorder, it may coexist with another disorder associated with hypercalcemia, such as malignancy.

THERAPY SURGERY In view of the modern clinical profile of primary hyperparathyroidism, not all patients have features that would prompt a recommendation for surgery. The existence of asymptomatic patients with primary hyperparathyroidism has engendered controversy about the need for parathyroidectomy and surgical guidelines. At the Consensus Development Conference on the Diagnosis and Management of Asymptomatic Primary Hyperparathyroidism, a set of surgical guidelines was endorsed73 (Table 58-1). The criteria for surgery include serum calcium concentration >12 mg/dL; any complication of primary hyperparathyroidism (e.g., overt bone disease, nephrolithiasis, nephrocalcinosis, classic neuromuscular disease); marked hypercalciuria (>400 mg per day); reduction in bone density more than two standard deviations below normal at the site of cortical bone, as in the forearm; an episode of acute hyperparathyroidism; and age younger than 50 years (see Table 58-1). Approximately 50% of patients with primary hyperparathyroidism meet one or more of these criteria. This is a significantly greater percentage of patients that have symptomatic primary hyperparathyroidism (20–30%). Some patients with asymptomatic primary hyperparathyroidism do meet surgical criteria and are candidates for parathyroid surgery, and in individual cases, surgical guidelines may be altered according to clinical judgment.

TABLE 58-1. Criteria for Surgery in Primary Hyperparathyroidism

Parathyroidectomy leads to the rapid resolution of the biochemical abnormalities of primary hyperparathyroidism.74 Surgery has also been documented to be of clear benefit in reducing the incidence of recurrent nephrolithiasis75,76 and in leading to an improvement in bone mineral density.74 Parathyroidectomy leads to a 12% rise in bone density mainly at the cancellous lumbar spine and femoral neck (Fig. 58-5). This increase is sustained over at least four years after surgery.

FIGURE 58-5. Mean (±SE) bone mineral density at three sites in two groups of patients with primary hyperthyroidism. Cumulative percentage change (mean ±SEM) from baseline by site at year 1, year 4, year 7, and year 10 of follow-up, reported in patients followed with no intervention (hatched bars) and after parathyroidectomy (solid bars). Differences between parathyroidectomy and no intervention groups are shown. (From Silverberg SJ, Gartenberg F, Jacobs TP, et al. Increased bone mineral density following parathyroidectomy in primary hyperparathyroidism. J Clin Endocrinol Metab 1995; 80:729.

A few patients show a densitometric profile that is unusual for patients with asymptomatic primary hyperparathyroidism. These individuals do not show the usual sparing of cancellous bone; rather, they have low bone density at the spine. In these patients, the postoperative increase in vertebral bone density is even more dramatic than that seen in the average patient; this has led to the recommendation that patients with low vertebral bone density also should be considered for parathyroidectomy.77 Special mention should be made of surgery in postmenopausal women with primary hyperparathyroidism.78 Many postmenopausal women with primary hyperparathyroidism have bone mass of the lumbar spine that is not below normal. Bone loss in the cancellous spine, typical of the postmenopausal state, may not be apparent. In these patients, the hyperparathyroid state could conceivably afford a relative protection, and the physician could use this information to proceed with a more conservative approach to management, especially if other guidelines for surgery are not met. On the other hand, the increase in bone density after surgery (at the spine and hip) is also seen in postmenopausal women.74 Issues related to parathyroid surgery and to preoperative localization of parathyroid tissue are also covered in Chapter 62.79 NONSURGICAL MEDICAL APPROACHES In patients who are not candidates for surgery at the time primary hyperparathyroidism is recognized, the course over the next 10 years often is stable (see Fig.

58-5).74,80 Serum calcium and PTH levels, urinary calcium excretion, and other biochemical indices do not appear to show any changes or trends over time. Bone mineral densitometry at all three sites (i.e., forearm, hip, lumbar spine) does not appear to show any unusual changes or trends over time.81,82 Fractures might be expected to be increased in patients followed conservatively, because of the reduction in bone mineral density typically seen in this disease. In fact, an increased incidence of distal radial fractures would be consistent with the known selective effects of PTH to reduce cortical bone mass. Despite publication of a few reports, no conclusion can be drawn as to whether fracture incidence is increased in primary hyperparathyroidism or whether these potential adverse events show site or time dependence.83,84 Relatively little is known about survival in patients who develop primary hyperparathyroidism. The indolent nature of the disease, as well as the difficulty in long-term follow-up, account, in part, for lack of pertinent information about mortality in this disease. However, the Mayo Clinic's review of the records of >400 patients with primary hyperparathyroidism for 30 years indicates that survival, on average, is indistinguishable from the expected longevity from life tables.85 Patients with hypercalcemia in the highest quartile (Ca2+ of 11.2–16.0 mg/dL) may have had higher mortality; however, this becomes apparent only after 15 years. These findings are different from those of earlier studies in which an increased risk of death from cancer and cardiovascular events was reported.86,87 These differences may be attributed to the apparently much milder disease observed in the Mayo Clinic population, very much like that usually seen in the United States today. Although knowledge of the natural history of primary hyperparathyroidism managed without parathyroid surgery is still incomplete, medical approaches to the management of primary hyperparathyroidism should be considered in nonsurgical patients. Patients with acute primary hyperparathyroidism should be treated the same as any patient with severe hypercalcemia88,89 (see Chap. 59). Long-term management of chronically elevated, mild hypercalcemia centers on adequate hydration and ambulation. If possible, diuretics should be avoided; in particular, thiazide diuretics may worsen the hypercalcemia in some patients. Other diuretics, such as furosemide, may place the patient at risk for dehydration and electrolyte imbalances. General recommendations for diet are not yet certain, and rationales exist for both low- and highcalcium diets in patients with hyperparathyroidism. Diets high in calcium may suppress levels of endogenous PTH. However, high-calcium diets may lead to greater absorption of calcium because of the elevated levels of 1,25(OH)2D in some patients.90,91 The recommendation for a low-calcium diet is based on the notion that less calcium is available for absorption. However, low-calcium diets may predispose patients to further stimulation of endogenous PTH levels. One study showed no effect of dietary calcium on biochemical indices or bone densitometry in patients with primary hyperparathyroidism.92 A normal calcium intake can be followed without adverse effects, except in those patients with elevated 1,25-dihydroxyvitamin D levels. Such patients are advised to be more moderate in their calcium intake to prevent hypercalciuria. Other approaches to the medical management of primary hyperparathyroidism have been considered. Attempts to block PTH secretion with b-adrenergic inhibitors or H2-receptor antagonists have not been successful.7 Oral phosphate, which has been used for many years in primary hyperparathyroidism, lowers serum calcium by 0.5 to 1.0 mg/dL in most patients. The average dosage is 1 to 2 g daily in divided doses. Phosphate appears to have several mechanisms of action. It may inhibit calcium absorption from the gastrointestinal tract; it prevents calcium mobilization from bone; and it may also impair the production of 1,25(OH)2D. The risks of the long-term use of phosphate in the management of patients with primary hyperparathyroidism are unknown. One concern is the possibility of ectopic calcification in soft tissues when the normal solubility product of Ca2+ × PO43– is exceeded (normally ~40). Phosphate therapy is contraindicated when renal insufficiency or hyperphosphatemia is present. If phosphate is to be used, the serum levels of calcium and phosphate should be monitored at regular intervals. Moreover, the long-term use of phosphate in patients with primary hyperparathyroidism promotes a further increase in PTH.93 The possibility that some of the symptoms and signs of primary hyperparathyroidism are caused by PTH itself and not by hypercalcemia raises questions about further increasing PTH levels in conjunction with phosphate therapy. If sufficient concern exists about lowering the serum calcium level in patients with asymptomatic primary hyperparathyroidism, parathyroid surgery remains the treatment of choice. Estrogen therapy has been proposed as a means of lowering serum calcium, especially because prevalence of primary hyperparathyroidism among women is increased in the post-menopausal years. Estrogens have well-known but not clearly understood antagonist actions on PTH-induced bone resorption. The serum calcium does tend to fall by ~0.5 to 1.0 mg/dL in women receiving estrogens.94,95 and 96 However, PTH and phosphorus levels do not change. More studies are required to better delineate the role of estrogen therapy. To date, no available data exist on the role of selective estrogen-receptor modulators (SERMs) in the treatment of primary hyperparathyroidism. Because the antiresorptive actions of these drugs are similar to those of estrogens, SERMs might be predicted to lower serum calcium levels in postmenopausal women with primary hyperparathyroidism in a fashion similar to that seen with estrogen. This hypothesis remains to be tested. Calcitonin may have a potential use in treating primary hyperparathyroidism, but no controlled trial has been conducted to test the efficacy of calcitonin for management of this condition. Data from the use of calcitonin in other states of increased bone turnover indicate that it may never become a useful long-term therapy for primary hyperparathyroidism. The bisphosphonates represent a class of important antiresorbing agents that may emerge as a useful approach to the medical management of primary hyperparathyroidism. Oral etidronate is not useful, and the effect of oral clodronate and alendronate is limited in duration of effect.97 Of the third-generation bisphosphonates, risedronate has been shown to have some efficacy in preliminary studies of patients with primary hyperparathyroidism.98 Finally, specifically targeted medical therapy for primary hyperparathyroidism is under active investigation. Calcimimetic agents that target the calcium-sensing receptor on the parathyroid cell99 have shown early promise in animal and in vitro studies.100,101 Data from human investigations are also encouraging. In early studies, one such calcimimetic has been shown to lower serum calcium and PTH levels both in a patient with parathyroid carcinoma and in a group of postmenopausal women with mild primary hyperparathyroidism.102,103

PRIMARY HYPERPARATHYROIDISM DURING PREGNANCY Rarely, primary hyperparathyroidism becomes evident during pregnancy.104 Hyperparathyroidism during pregnancy used to be associated with an increased incidence of fetal death. Perinatal and neonatal complications were also thought to be increased in the hypocalcemic infant whose endogenous PTH production is suppressed by maternal hypercalcemia. Neonatal hypocalcemia and tetany can be the first sign of primary hyperparathyroidism in the mother. Although systematically collected and controlled data are unavailable, experience has indicated that primary hyperparathyroidism during pregnancy can be managed successfully without resorting to surgery.105,106 The management of primary hyperparathyroidism during pregnancy is controversial.107 Some clinicians advocate surgery during the second trimester to reduce fetal risk of chronic hypercalcemia during the gestational period. Others advocate a much more conservative approach.

PARATHYROID CARCINOMA Parathyroid carcinoma is a rare form of primary hyperparathyroidism.108,109 These patients often have hypercalcemia that is more severe than that usually seen in primary hyperparathyroidism. Serum calcium values >14 mg/dL may suggest a parathyroid malignancy. However, parathyroid carcinoma is uncommon, occurring in 95% of the time, which makes it the most powerful single test for differential diagnosis in Cushing syndrome.27 In previously untreated patients, the procedure correctly lateralizes the disease ~85% of the time. Untoward effects are largely confined to the catheterization process. A few patients have had transient signs and symptoms of brainstem dysfunction or other neurological findings.27a These have not led to permanent disability. Other untoward effects include bleeding, bruising, and minor infections at the venipuncture site.

FIGURE 74-7. Anatomy of venous draining of the pituitary gland.

COMPUTED TOMOGRAPHY CT has become the procedure of choice in localizing adrenal disease (see Chap. 88). Most adrenal cancers are large, and CT, for practical purposes, identifies all these, as well as nearly all the smaller adrenal adenomas. Other useful findings revealed by CT include adrenal hyperplasia, usually bilateral and symmetric, and an abnormal intraabdominal fat distribution that supports the clinical diagnosis of Cushing syndrome.28,29,30 and 31 MAGNETIC RESONANCE IMAGING Magnetic resonance imaging offers only a single advantage over CT in the differential diagnosis of adrenal disease: the classification of adrenal masses into probably benign versus probably malignant categories on the basis of the T2-weighted image.32 Normal adrenal glands are dark on T1-weighted magnetic resonance imaging scans. Benign adenomas remain dark when the scan is shifted to T2-weighting, but malignant lesions become bright. A hyperintense rim on fat-saturated spin-echo magnetic resonance imaging is often seen in adrenal adenomas and may aid in their differentiation from adrenal metastases.33 Adrenal cysts and pheochromocytomas also are bright, but usually present no problem in diagnosis when viewed in the context of other endocrine tests and the ultrasound examination for cystic structure (see Chap. 88). IODOCHOLESTEROL SCAN The iodocholesterol scan is an imaging procedure that depends on the function of the adrenal gland. It is rarely required in the differential diagnosis of hypercortisolism. Circulating cholesterol serves as the precursor for most of the steroids produced by the adrenal gland. Not surprisingly, therefore, circulating cholesterol is concentrated in the adrenal glands. Iodocholesterol, a g-emitter, can be used to examine this physiologic process. Normal and hyperactive glands image bilaterally. Adenomas causing Cushing syndrome image, but the contralateral side does not, because the resultant hypercortisolism suppresses ACTH and diminishes the function of the normal gland. Adrenal carcinomas causing Cushing syndrome fail to produce an image because of an inefficient cholesterol-concentrating mechanism. The contralateral normal side also fails to image because ACTH secretion is suppressed. Thus, if both sides are visualized, an ACTH-dependent process is implied. If one side is visualized, an adrenal adenoma is implied. If neither side is seen, adrenal carcinoma or factitious disease is suggested.34,35 and 36 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.

Ruder HJ, Guy RL, Lipsett MB. Radioimmunoassay for cortisol in plasma and urine. J Clin Endocrinol Metab 1972; 35:219. Murphy BEP. Clinical evaluation of urinary cortisol determinations by competitive protein binding radioassay. J Clin Endocrinol Metab 1968; 28:343. Hsu TH, Bledsoe T. Measurement of urinary free corticoids by competitive protein binding radioassay in hypoadrenal states. J Clin Endocrinol Metab 1970; 30:443. Beardwell CG, Burke CW, Cope CL. Urinary free cortisol measured by competitive protein binding. J Endocrinol 1968; 42:79. Silber RH, Porter CC. The determination of 17-21-dihydroxy, 20-ketosteroids in urine and plasma. J Biol Chem 1954; 210:923. Borushek S, Gold JJ. Commonly used medications that interfere with routine endocrine laboratory procedures. Clin Chem 1964; 10:41. Werk EE, MacBee T, Sholiton LJ. Effect of diphenylhydantoin on cortisol metabolism in man. J Clin Invest 1964; 43:1824. Burstein S, Klaiber EL. Phenobarbital-induced increase in 6-hydroxycortisol excretion: clue to its significance in human urine. J Clin Endocrinol Metab 1965; 25:293. Franks RC. Urinary 17-hydroxycorticosteroid and cortisol excretion in childhood. J Clin Endocrinol Metab 1973; 36:702. Krieger DT, Allen W, Rizzo F, Krieger HP. Characterization of the normal temporal pattern of plasma corticosteroid levels. J Clin Endocrinol Metab 1971; 32:266. Doe RP, Zimmerman HH, Flink EB. Significance of the concentration of nonprotein bound plasma cortisol in normal subjects, Cushing's syndrome, pregnancy, and during estrogen therapy. J Clin Endocrinol Metab 1960; 20:1484. Liddle GW, Island DP, Meador CK. Normal and abnormal regulation of corticotropin secretion in man. Recent Prog Horm Res 1962; 18:125. Flack MR, Loriaux DL, Nieman LK. Urinary free cortisol excretion: a better measure of response to the dexamethasone suppression test in Cushing's syndrome? (Abstract). New Orleans: Endocrine Society, 1988:1. Liddle GW, Estep HL, Kendall TW. Clinical application of a new test of pituitary reserve. J Clin Endocrinol Metab 1959; 19:875. Meikle AW, Jubiz W, Hutchings MD, et al. A simplified metyrapone test with determination of plasma 11-deoxycortisol (metyrapone test with plasma S). J Clin Endocrinol Metab 1969; 29:985. Jubiz W, Meikle AW, West CD, Tyler FH. A single dose metyrapone test. Arch Intern Med 1970; 125:472. Cushman P. Hypothalamic-pituitary-adrenal function in thyroid disorders: effects of metyrapone infusion on plasma corticosteroids. Metabolism 1968; 17:263. Meikle AW, Jubiz W, Matsukura S, et al. Effect of estrogen on the metabolism of metyrapone and release of ACTH. J Clin Endocrinol Metab 1970; 30:259. Jubiz W, Levinson RA, Meikle AW. Absorption and conjugation of metyrapone during diphenylhydantoin therapy: mechanism of abnormal response to oral metyrapone. Endocrinology 1970; 86:328. Chrousos GP, Schulte HM, Oldfield EH, et al. The corticotropin releasing factor stimulation test: an aid in the evaluation of patients with Cushing's syndrome. N Engl J Med 1984; 310:622. Nieman LK, Chrousos GP, Oldfield EH, et al. The corticotropin-releasing hormone stimulation test and the dexamethasone suppression test in the differential diagnosis of Cushing's syndrome. Ann Intern Med 1986; 105:862. Kehlet H, Binder C. Value of an ACTH test in assessing hypothalamic-pituitary-adrenocortical function in glucocorticoid treated patients. BMJ 1973; 2:147. Lindholm J, Kehlet H, Blichert-Toft M. Reliability of the 30-minute ACTH test in assessing hypothalamic-pituitary-adrenal function. J Clin Endo-crinol Metab 1978; 47:272. Schulte HM, Chrousos GP, Avgerinos P. The CRF test: a possible aid in the evaluation of patients with adrenal insufficiency. J Clin Endocrinol Metab 1984; 58:1064. Slyper AH, Findling JW. Use of a two-site immunoradiometric assay to resolve a factitious elevation of ACTH in primary pigmented nodular adrenocortical disease. J Pediatr Endocrinol 1994; 7:61. Oldfield EH, Chrousos GP, Schulte HM, et al. Preoperative localization of ACTH secreting pituitary microadenomas by bilateral and simultaneous inferior petrosal sinus sampling. N Engl J Med 1985; 312:100. Zovickian J, Oldfield EH, Doppman JL, et al. Usefulness of inferior petrosal sinus venous endocrine markers in Cushing's disease. J Neurosurg 1988; 68:205.

27a.Lefournier V, Gatta B, Martinie M, et al. One transient neurological complication (sixth nerve palsy) in 166 consecutive interior petrosal sinus samplings for the etiological diagnosis of Cushing's syndrome. J Clin Endocrinol Metab 1999; 84:3401. 28. 29. 30. 31. 32. 33. 34. 35. 36.

Vincent JM, Morrison ID, Armstrong P, Reznek RH. The size of normal adrenal glands on computed tomography. Clin Radiol 1994; 49:453. Ganguly A, Pratt JH, Yune HY, et al. Detection of adrenal tumors by computerized tomographic scan in endocrine hypertension. Arch Intern Med 1979; 139:590. Kelestimur F, Unlu Y, Ozesmi M, Tolu I. A hormonal and radiological evaluation of adrenal gland in patients with acute or chronic pulmonary tuberculosis. Clin Radiol 1994; 49:453. White FE, White MC, Drury PL, et al. Value of computed tomography of the abdomen and chest in investigation of Cushing's syndrome. BMJ 1982; 284:771. Heinz-Peer G, Honigschabel S, Schneider B, et al. Characterization of adrenal masses using MR imaging with histopathologic correlation. AJR Am J Roentgenol 1999; 173:15. Ichikawa T, Ohtamo K, Uchiyama G, et al. Adrenal adenomas: characteristic hyperintense rim sign on fat-saturated spin-echo MR images. Radiology 1994; 193:247. Moses DC, Schteingart DS, Sturman MF, et al. Efficacy of radiocholesterol imaging of the adrenal glands in Cushing's syndrome. Surg Gynecol Obstet 1974; 139:201. Chatal JR, Charbonnel B, Le Mevel BP, et al. Uptake of 131 I-19-iodocholes-terol by an adrenal cortical carcinoma and its metastases. J Clin Endocrinol Metab 1976; 43:248. Beierwaltes WH, Sisson JC, Shapiro B. Diagnosis of adrenal tumors with radionuclide imaging. Spec Top Endocrinol Metab 1984; 6:1.

CHAPTER 75 CUSHING SYNDROME Principles and Practice of Endocrinology and Metabolism

CHAPTER 75 CUSHING SYNDROME DAVID E. SCHTEINGART Pathophysiology Corticotropin-Dependent Cushing Syndrome Pituitary Corticotropin-Dependent Cushing Syndrome Ectopic Corticotropin Syndrome Ectopic Corticotropin-Releasing Hormone Syndrome Corticotropin-Independent Cushing Syndrome Primary Bilateral Micronodular Hyperplasia Adrenal Adenoma and Carcinoma Clinical Presentations Corticotropin-Dependent Cushing Disease Ectopic Corticotropin and Corticotropin-Releasing Hormone Syndromes Corticotropin-Independent Types of Cushing Syndrome Diagnosis Standard Diagnostic Evaluation Difficulties in Interpretation of the Diagnostic Tests Anatomic Localization Detection of Pituitary Lesions Detection of Adrenal Lesions Localization of Ectopic Corticotropin-Secreting Lesions Pseudo–Cushing Syndrome Obesity Major Depressive Disorder Alcohol-Induced Pseudo–Cushing Syndrome Differential Diagnosis of Cushing-Type Hypercortisolism and Pseudo–Cushing Syndrome Corticotropin-Releasing Hormone in the Pathophysiology of Pseudo–Cushing Syndrome Treatment of Cushing Disease Pituitary Surgery Pituitary Irradiation Adrenal Surgery Drug Therapy Treatment of Ectopic Adrenocorticotropin Syndrome Treatment of Corticotropin-Independent Cushing Syndrome Chapter References

Cushing syndrome is the clinical expression of the metabolic effects of persistent, inappropriate hypercortisolism. Since the disorder was first described by Harvey Cushing in 1932, great progress has been made in understanding its pathophysiology and the spectrum of its clinical presentation. The diagnosis can now be precisely established by biochemical means, by noninvasive imaging of the pituitary and adrenal lesions, and by selective identification of the source of corticotropin (ACTH or adrenocorticotropic hormone) in the ACTH-dependent types. The clinical manifestations of Cushing syndrome involve many organ systems and metabolic processes.1 Hypertension, obesity, diabetes, androgen-type hirsutism, and acne are commonly seen in these patients and are prevalent in patients who do not have primary cortisol excess. When present, the following findings increase the suspicion of Cushing syndrome: symptoms and signs of protein catabolism (e.g., ecchymoses, myopathy, osteopenia); truncal obesity; lanugal hirsutism; cutaneous lesions (i.e., wide, purple striae; tinea versicolor; verruca vulgaris); hyperpigmentation; and psychiatric manifestations (i.e., impairment of affect, cognition, and vegetative functions).2,3,4,5 and 6 In some instances, the diagnosis of Cushing syndrome can be strongly suspected on clinical grounds alone, but for most patients, the diagnosis is established by laboratory studies.7

PATHOPHYSIOLOGY Cushing syndrome can be exogenous, resulting from the administration of glucocorticoids or ACTH, or endogenous, resulting from a primary increased secretion of cortisol or ACTH. An etiologic classification of Cushing syndrome is given in Figure 75-1.

FIGURE 75-1. Classification of Cushing syndrome. (ACTH, corticotropin; CRH, corticotropin-releasing hormone.)

The changes in ACTH and cortisol secretion observed in patients with ACTH-dependent and ACTH-independent Cushing syndrome are illustrated in Figure 75-2. In the ACTH-dependent types, ACTH is secreted by the pituitary gland or by an extrapituitary neoplasm capable of producing peptide hormones, including ACTH and corticotropin-releasing hormone (CRH). In pituitary-dependent Cushing disease, the feedback relationship between the pituitary and the adrenal glands is maintained but is abnormal. In the ectopic ACTH syndrome, pituitary ACTH release is suppressed by excessive production of cortisol, which is stimulated by an unsuppressible overproduction of ACTH by the ectopic source. In ectopic CRH syndrome, tumor CRH stimulates ACTH secretion by the pituitary corticotropes; this causes an excessive production of cortisol that is not sufficient to suppress ACTH secretion. In patients with ACTH-independent types of Cushing syndrome, pituitary ACTH secretion is suppressed by excessive cortisol production that originates in adrenocortical tumors (i.e., adenomas or carcinomas) or in autonomous nodular hyperplastic glands.

FIGURE 75-2. Corticotropin (ACTH) and cortisol levels in ACTH-dependent and ACTH-independent types of Cushing syndrome. (CRH, corticotropin-releasing

hormone.)

CORTICOTROPIN-DEPENDENT CUSHING SYNDROME PITUITARY CORTICOTROPIN-DEPENDENT CUSHING SYNDROME Approximately 85% of patients with pituitary ACTH-dependent Cushing syndrome, also called Cushing disease, have pituitary microadenomas. The origin, nature, and biochemical characteristics of these microadenomas are not clear.8,9 In most cases, the microadenomas are located in the periphery of the gland and can be identified by imaging techniques or transsphenoidal pituitary exploration. Occasionally, they are located deep in the central wedge of the pituitary and can be missed on imaging or at surgery.8 The suggestion has been made that excessive ACTH secretion may be caused by neurohypothalamic stimulation, which results in corticotrope hyperplasia.10,11 However, fewer than 17% of cases of Cushing disease are associated with ACTH hypersecretion by nonneoplastic corticotrope cells.12 Whether pituitary ACTH-dependent Cushing disease develops from a primary pituitary disorder or is the result of corticotrope stimulation by CRH has been unclear. In the case of a pituitary adenoma, the nonadenomatous corticotropes are usually suppressed, and removal of the adenoma is followed by ACTH and cortisol insufficiency. In the rare cases of corticotrope hyperplasia, the possibility exists that these corticotropes have been stimulated from a hypothalamic source. Clonal analysis of the pituitary tissue obtained from patients with Cushing disease shows that ACTH-secreting adenomas have a monoclonal pattern consistent with a monoclonal proliferation of a genetically altered cell. Thus, a spontaneous somatic mutation in pituitary corticotropes is the primary pathogenetic mechanism in this disorder. In contrast, the corticotrope hyperplasia seen in patients with CRH-secreting bronchial carcinoids is polyclonal.13 Direct measurement of CRH levels in patients with Cushing disease secondary to a pituitary adenoma has not been performed. In the few patients in whom CRH levels have been measured in the spinal fluid, low levels have been found, suggesting suppressed CRH secretion. The manifestations of abnormal ACTH secretion in Cushing disease are a loss of normal negative feedback and a blunted circadian rhythm of ACTH and cortisol secretion. Although the relative sensitivity of ACTH release to corticosteroid suppression remains in most of these patients, as many as 40% exhibit resistance to suppression and resemble patients with the ectopic ACTH syndrome. Unusual abnormal patterns of release of ACTH and cortisol have been described, and episodes of hypersecretion have persisted for hours or weeks and are interrupted by periods of normal secretion.14,15 This periodic hormonogenesis has been found in patients with and without ACTH-secreting pituitary adenomas. ECTOPIC CORTICOTROPIN SYNDROME The association of Cushing syndrome with nonadrenal neoplasms has been recognized for decades and is called the ectopic ACTH syndrome.16 Typically, a rapidly developing malignancy causes the autonomous secretion of high levels of ACTH and cortisol, with clinical manifestations of Cushing syndrome and pigmentation. Approximately 50% of these ectopic ACTH-secreting tumors are in the lung, and the remainder occur in a variety of tissues. Table 75-1 describes the type and frequency of these tumors.

TABLE 75-1. Tumors Causing Ectopic Corticotropin Syndrome

Pancreatic islet cell tumors have been associated with Cushing syndrome in fewer than 5% of patients. In addition to ACTH, these tumors sometimes contain other hormones, such as insulin, gastrin, and glucagon, which are not produced in sufficient amounts to cause a metabolic syndrome. Less frequent causes of the ectopic ACTH syndrome include thymic carcinoids; gastric and appendicular carcinomas; pheochromocytomas; adrenal medullary paragangliomas; cystadenomas of the pancreas; epithelial carcinomas of the thymus; medullary thyroid carcinomas; prostatic carcinomas; carcinomas of the esophagus, ileum, and colon; ovarian tumors; squamous cell carcinoma of the lung, larynx, and cervix; and melanomas. Although the relationship between the more commonly occurring tumors and ectopic ACTH production has been well established, the production of ACTH by the less frequently occurring tumors has not been absolutely proved.17 Several mechanisms of ectopic elaboration of peptide hormones have been postulated. Neoplastic transformation of dormant stem cells results in expression by the tumor of a specific hormone gene that would not otherwise be expressed by the mature cells of that organ. An alternative hypothesis is that cells with endocrine potential migrate to various organs during embryogenesis, where they remain dormant. These cells have the capacity to initiate amine precursor uptake and decarboxylation (APUD) and synthesize a variety of peptides.18 This hypothesis does not account for the large number of ectopic hormones elaborated by non-APUD cells. An enhanced expression of genes normally encoding for proopiomelanocortin (POMC) in extrapituitary tissues is a more likely explanation for the development of ectopic hormone syndromes. These tissues include stomach, adrenal medulla, pancreas, brain, and placenta.19 The tumor cells in patients with the ectopic ACTH syndrome appear to have developed a more efficient mechanism for translation of messenger RNA (mRNA) to immunoreactive POMC-derived peptides or may have lost the mechanism for preventing the continuous expression of the POMC gene.20 Unique transcription factors may be constitutively expressed in tumor cells and cause unregulated secretion of POMC-derived peptides.21 An unusual cause of ectopic ACTH production is an ectopic pituitary adenoma. In some cases, normal pituitary tissue occurs in an ectopic location and not in continuity with intrasellar pituitary tissue. Occasionally, adenomas may develop in this ectopic pituitary tissue, causing excessive ACTH production.22 From the biochemical standpoint, these patients resemble those with pituitary ACTH-dependent or ectopic ACTH-dependent Cushing syndrome. The possibility of an ectopic pituitary adenoma in a patient presenting with an ACTH-dependent Cushing syndrome should be considered if the patient exhibits an atypical hormonal profile in the absence of an ectopic ACTH-secreting, extracranial neoplasm or in the absence of a pituitary abnormality at the time of transsphenoidal exploration. A computerized tomographic (CT) study of the parasellar region and the sphenoid sinus may help to identify such ectopically located pituitary adenomas. ECTOPIC CORTICOTROPIN-RELEASING HORMONE SYNDROME Ectopic CRH production can cause Cushing syndrome.23,24 Tumors with a capacity to secrete CRH ectopically include bronchial carcinoids, medullary thyroid carcinoma, and metastatic prostatic carcinoma. In patients with ectopic ACTH production, the pituitary ACTH content is low and only Crooke changes (i.e., hyaline degeneration) are observed in pituitary corticotropes. In contrast, patients with ectopic CRH secretion demonstrate corticotrope hyperplasia. The presence of an ectopic CRH-secreting tumor should be suspected in patients who have disease that behaves biochemically like pituitary ACTH-dependent disease but who have a clinical course of short duration and high plasma ACTH levels.

CORTICOTROPIN-INDEPENDENT CUSHING SYNDROME Primary nodular hyperplasia, adrenocortical adenomas, and carcinomas as well as rare cases of ectopic cortisol production by tumor are found in 30% of patients presenting with Cushing syndrome.25 PRIMARY BILATERAL MICRONODULAR HYPERPLASIA Primary bilateral micronodular hyperplasia is characterized by the presence in the adrenal cortex of one or more yellow nodules visible to the naked eye and 2 to 3 cm

in diameter. Its pathogenesis is unknown. Whether this type of hyperplasia is of primary adrenal origin or the result of a combined mechanism by which the pituitary predominates initially but the adrenal eventually becomes autonomous, is not clear.26 Increased adrenocortical cell sensitivity to ACTH has also been suggested.27 Histologically, the micronodules consist of clear cells arranged in acini and cords, and are occasionally encapsulated and surrounded by simple or micronodular cortical hyperplasia. The frequency of mitotic figures is low. In one form of ACTH-independent micronodular adrenal disease, darkly pigmented micronodules are seen in the presence of atrophy of the perinodular adrenal tissue, disorganization of normal zonation of the cortex, and small glands28 (Fig. 75-3). Microscopically, the nodules are composed predominantly of large globular cortical cells with granular eosinophilic cytoplasm that often contains lipofuscin.

FIGURE 75-3. Appearance of an adrenal gland with pigmented micronodular hyperplasia removed from a patient with familial, corticotropin-independent Cushing syndrome. The patient was also found to have an atrial myxoma and a Sertoli cell testicular tumor.

A condition associated with pigmented micronodular hyperplasia is Carney complex, a form of multiple endocrine neoplasia. Patients with this condition have, in addition to the adrenal cortical disease, growth hormone–secreting pituitary tumors, Sertoli cell testicular tumors, atrial or ventricular myxomas, single or multiple mammary fibroadenomas, and cutaneous myxomas. A prominent feature in these patients is spotty skin pigmentation similar to that observed in several lentiginosis syndromes. The genetic defect(s) responsible for Carney complex map(s) to the short arm of chromosome 2 (2p16). This region has been found to exhibit cytogenic aberrations in atrial myxomas associated with the complex and has been characterized by microsatellite instability in human neoplasias.29 ACTH-independent macronodular adrenal cortical hyperplasia is a rare cause of Cushing syndrome in which the adrenal glands become very enlarged and occupied by multiple large cortical nodules. In some patients, the masses can be extremely large and weigh close to 100 g each. Histologically, the nodules are composed mostly of clear cells, and no cortical architecture has been observed in the internodular regions.30 Several interesting pathogenic mechanisms have been described in patients with ACTH-independent nodular adrenal cortical hyperplasia. In one case, the secretion of cortisol was stimulated by food. The patient had elevated postprandial cortisol levels that could not be suppressed with dexamethasone. Cortisol increased in response to oral glucose, a lipid-rich meal, or a protein-rich meal but not to the intravenous administration of glucose. In response to test meals, plasma glucose-dependent insulinotropic polypeptide (GIP) increased in parallel to the plasma cortisol. The ability of GIP to stimulate adrenal cortisol production could be demonstrated in cell suspensions of adrenal tissue from the patient. Nodular adrenal cortical hyperplasia and Cushing syndrome were postulated to be food dependent and to result from abnormal responsiveness of adrenal cells to the physiologic secretion of GIP. The suggestion was made that ectopic expression of GIP receptors on the adrenal cells mediate this disorder.31 A case of Cushing syndrome secondary to ACTH-independent bilateral adrenal hyperplasia was also described in which cortisol secretion responded to catecholamines acting through ectopic adrenal b receptors. The increased cortisol secretion was inhibited by b-blockade with propranolol. High-affinity binding sites compatible with b1 or b2 receptors were detected in the adrenal tissue from this patient but not from control subjects. The ectopic expression of b receptors in the adrenal cortex of this patient led to nodular hyperplasia, hypersecretion of cortisol and aldosterone, and feedback suppression of ACTH and angiotensin.32 ADRENAL ADENOMA AND CARCINOMA Adrenal adenomas are unilateral, well-circumscribed, brown to yellow tumors with a homogeneous appearance, weighing 100 g, they may be palpable on the abdominal examination, and hepatomegaly may be noted if hepatic metastases are present. Although the discovery of an adrenocortical adenoma is usually made while a patient is being investigated for the possibility of Cushing syndrome, the adenoma occasionally is discovered incidentally during the investigation of abdominal complaints. Silent adrenal masses are found in 2% to 3% of patients who are undergoing CT scans for investigation of abdominal symptoms.34a Although these patients may not have obvious clinical manifestations of Cushing syndrome and their baseline urinary cortisol levels are normal, cortisol secretion is autonomous. Evidence of autonomous cortisol secretion includes a blunted circadian rhythm, suppressed ACTH levels, and a lack of response to dexamethasone. Adrenal scintigraphy with iodocholesterol shows increased uptake on the side of the tumor but suppression of uptake in the contralateral adrenal gland. Although these patients may have minimal manifestations of Cushing syndrome, these manifestations improve after surgical removal of the lesion. Patients with primary nodular adrenocortical hyperplasia are clinically indistinguishable from patients with benign adrenocortical adenomas or pituitary ACTH-dependent disease.

DIAGNOSIS

STANDARD DIAGNOSTIC EVALUATION A biochemical evaluation of Cushing syndrome is necessary for confirming the clinical diagnosis and for determining the presence of the syndrome in patients with an equivocal clinical presentation.35 Preliminary testing includes the measurement of urinary free cortisol and of the ability to suppress serum cortisol levels with a low dose of dexamethasone (Table 75-2). In patients with Cushing syndrome, the urinary free cortisol usually exceeds 90 µg per day. Serum cortisol levels obtained 9 hours after the oral administration of 1 mg of dexamethasone at 23:00 hours fail to decrease below 10 µg/dL, and plasma ACTH levels (assessed by immunoradiometric assay) fail to decrease to 50 known physiologic actions. These include vasoconstriction, salt retention through aldosterone, increased thirst and secretion of antidiuretic hormone resulting in retention of water, increased cardiac output, and stimulation of sympathetic nervous system activity (Table 79-3). Also, there are direct effects of A-II on early proximal tubule Na+ reabsorption that are mediated in part through the Na+, H+ antiporter; this accounts for 40% to 50% of sodium and water reabsorption.58 In AGT gene-null mutant mice, chronic volume depletion causes a sharp decrease in glomerular filtration rate (GFR) and a marked concentration defect, which is insensitive to AVP.59 Furthermore, the renin-angiotensin system is essential for two fundamental homeostatic functions in the kidney, stabilizing the GFR and maintaining urine diluting and concentrating capacity.

TABLE 79-3. Angiotensin II Actions in Various Target Tissues

Systemic Vasoconstriction. A-II produces arteriolar vaso-constriction and elevates vascular resistance and blood pressure. The signal transduction for this A-II–mediated vasoconstriction includes an increase of intracellular calcium and formation of protein kinase C.60 The facilitated release of norepinephrine61 and endothelin-I62 may also contribute to the vasoconstriction that is induced by A-II. A-II stimulates the release of arachidonic acidproducts in the vasculature (e.g., lipoxygenase products and P450 epoxides), which have vasoconstrictive properties.63 Regulation of Gomerular Filtration Rate. A-II regulates the GFR and renal blood flow by constricting the efferent and afferent glomerular arterioles, and the interlobular artery.64 A-II differentially increases efferent resistance to levels that are two to three times those of the afferent resistance; this elevates glomerular capillary pressure to maintain the single nephron GFR. In AGT gene-null mutant mice, chronic volume depletion causes a sharp decrease in GFR and a marked concentration defect, which is insensitive to AVP.59 These physiologic changes are paralleled by changes in renal structure at the glomerular level as well as at the renal papilla and underscore the key role of angiotensins in the development of the mammalian kidney. As mentioned previously, A-II can also stimulate prostaglandins (e.g., thromboxane) that mediate renal hemodynamics.65 A-II controls the GFR by constricting the mesangium and altering tubuloglomerular feedback. In sodium-replete women, A-II infusion elicits a blunted renal microcirculatory response compared with the response in sodium-replete men; this demonstrates that sex steroids modulate the effects of A-II.66 A-II also has nonhemodynamic effects on the kidney (e.g., stimulation of cytokines in diabetic nephropathy).67 Cardiac Effects. Cardiac Effects. The cardiac effects of the tissue angiotensin system are multiple.68 A-II is involved in cardiac remodeling and, through the AT1 receptor, promotes extracellular production of matrix as well as a fibrous tissue response and increasing collagen levels. In addition, the vasoconstriction caused by A-II and oxidative stress further contribute to ischemia and left ventricular hypertrophy with a resultant greater stiffness of the myocardial wall.69 Elevations in plasma renin activity are associated with increased risk for myocardial infarction.70 This risk may be related to renin-angiotensin-system–induced changes in the fibrinolytic system, because A-II inhibits fibrinolysis (thereby promoting clot formation) and stimulates excess production of plasminogen activator inhibitor.71 Also, the ACE-induced breakdown of bradykinin leads to lower tissue type plasminogen activator levels and decreases nitric oxide concentrations. A low-salt diet increases plasminogen activator inhibitor activity by increasing the renin-angiotensin system and raising the aldosterone levels.72 These findings help to explain why blockade of the renin-angiotensin system by ACE-I or by AT1 receptor antagonists protect against thrombotic cardiac events and reduce the risk of recurrent myocardial infarctions. Brain Effects. There are several actions of the angiotensins in the brain, an organ in which all three receptor subtypes exist.52 In addition, A-II exerts a central pressor activity and causes blunting of the baroreceptor; it also stimulates dipsogenic behavior and the synthesis of AVP. Other Effects. Other actions of A-II include platelet adhesion, glycogenolysis, and inhibition of prolactin release. A-II also has interactions with prostaglandins, nitric oxide, and endothelin. More diverse effects include contraction of the uterus, blockade of the effects of glucagon, stimulation of cate-cholamines, stimulation of endothelial cell growth, stimulation of ovulation, and regulation of steroidogenesis. Moreover, A-II selectively stimulates transforming growth factor-b and stimulates mRNA in the heart and kidneys (thereby contributing to cardiac and renal fibrosis), and it suppresses clusterin, a glyco-protein that is expressed in response to tissue injury.73 Functional analysis of A-II and other components of the renin-angiotensin system can now be achieved by transgenic and gene-targeting technology.74,75 The direct transfection of the human renin gene using gene-transfer techniques in hepatocytes in vivo elevates blood pressure. In vivo transfection of antisense oligodeoxynucleotides for AGT decreases plasma AGT levels and blood pressure in spontaneously hypertensive rats.74 Transfection of antisense to AGT or AT1 receptors decreases blood pressure in spontaneously hypertensive rats.76 Vascular injury is associated with abundant expression of AGT and ACE. This process of vascular hypertrophy can be duplicated by transfecting ACE vector locally into intact rat carotid arteries.74 In vivo transfer of antisense ACE oligodeoxynucleotides inhibits neointimal formation in balloon-injured rat carotid arteries.74 Antisense technology has been used to examine the role of the renin-angiotensin system in the central nervous system (see earlier).76 Conceivably, a gene therapy for hypertension could use an antisense inhibition with adeno-associated viral vector delivery to target the A-II type 1 receptor messenger ribonucleic acid.76 Prolonged reductions in blood pressure in hypertension may be achieved with a single dose of adeno-associated viral vectors to deliver antisense oligodeoxynucleotides to inhibit

AT1 receptors.76

ALDOSTERONE Aldosterone is the most potent of several mineralocorticoid steroid hormones that act on the epithelium, especially in the kidney, to produce Na+ reabsorption and K+ excretion.77 Other steroids that have weaker mineralocorticoid effects include deoxycorticosterone (DOC), 18-oxycortisol, 19-nor-deoxycorti-costerone (19-nor-DOC), and 18-hydroxydeoxycorticosterone (18-OH-DOC). SYNTHESIS The adrenal gland is partitioned into different zones, which contain specific enzymes that produce either glucocorticoids (i.e., cortisol), mineralocorticoids (i.e., aldosterone), or weak adrenal androgens.78,79 Aldosterone is synthesized in the outer region of the adrenal gland (zona glomerulosa). Cortisol is synthesized in the inner regions in the zona fasciculata and reticularis. The same adrenal enzymes share most of the early synthetic steps of these steroids, but they differ markedly in their terminal enzymes. The zona glomerulosa is unique in lacking 17a-hydroxylase, the essential enzyme for cortisol formation. However, the zona glomerulosa is rich in the enzyme aldosterone synthase, which is required for the terminal steps in aldosterone synthesis.78,79 and 80 Aldosterone synthase is a single multifunctional P450 enzyme on the inner mitochondrial membrane that converts corticosterone to aldosterone. It catalyzes three successive hydroxylation steps: (a) conversion of deoxycorticosterone to corticosterone; (b) addition of a hydroxyl group to corticosterone (18-OH-corticosterone); and (c) conversion of 18-OH-corticosterone to an aldehyde, with consequent formation of aldosterone. Two cytochrome P450 enzymes have been found, including 11b-hydroxylase and aldosterone synthase.78 Their genes are located in close proximity (40 kb) on chromosome 8 (8q21-q22) and display 95% homology. In the familial disorder glucocorticoid-remediable hyperaldosteronism, there is hypertension and variable hypokalemia as a result of the formation of a chimeric gene that makes aldosterone synthesis ACTH-dependent78(see Chap. 80). MINERALOCORTICOID RECEPTORS The mineralocorticoid receptor (steroid receptor type 1 [SR1]) binds both aldosterone and cortisol with equal affinity, whereas the cortisol receptor (steroid receptor type 2 [SR2]) binds only cortisol with high affinity.81,82 Both receptors are part of the steroid/vitamin D/retinoic acid superfamily of transcription regulators. Both receptors have high amino acid homology and may overlap the actions of progesterone and androgen receptors. In humans, there are two isoforms of the SR1; the a and b isoforms.83,84 The SR1s are found mainly on epithelial cells of the renal tubule, the parotid gland, and the colon. SR1s have been found in vascular smooth muscle, in the heart, and in areas of the brain. These receptors can undergo down-regulation, as seen with sodium loading in which the b isoform expression is decreased.84 In contrast, SR2s are found in most cells of the body. Glucocorticoids and mineralocorticoids bind to SR1 with equal affinity, and glucocorticoid levels are 1000-fold higher than those of mineralocorticoids; thus, one would expect the receptor to be saturated with cortisol. However, specificity is conferred by the enzyme 11b-hydroxysteroid dehydrogenase (HSD),78,79,85 which oxidizes cortisol to inactive corticosterone, thus allowing aldosterone to bind to its receptor. Thus, coexpression of 11b-HSD and SR1 in tissues appears to be a mechanism by which specificity is conferred to aldosterone action. There are multiple forms of this enzyme in the mammalian kidney,86 including a high density form in the principal and intercalated cells.87 In the syndrome of apparent mineralocorticoid excess, the renal isoform for 11b-HSD2 is missing, thus allowing the normally high concentrations of cortisol in the kidney to bind to SR1.78 This effect creates an excess Na+ reabsorption with resultant hypertension and hypokalemia (see Chap. 80). The active ingredient of licorice, glycyrrhetinic acid, also inhibits 11b-HSD2 causing an, acquired syndrome of mineralocorticoid excess in humans. However, several questions remain regarding mineral corticoid receptors.87a ALDOSTERONE ACTION The major site of aldosterone action is on the distal nephron, where it increases Na+ reabsorption and K+ and H+ secretion. Aldosterone initiates its action by diffusing across the plasma membrane of the cell and binding avidly to specific receptors, which are located mostly in the cytoplasm.82 In the nucleus, the aldosterone-receptor complex acts on chromatin to increase mRNA and ribosomal RNA transcription.87,88 The binding of steroid to the carboxyl terminal of the receptor causes release of heat shock proteins, after which the receptor changes conformation to allow dimerization and binding to the regulatory end of specific target genes (hormone-responsive elements).82 Mineralocorticoids affect electrolyte balance by inducing and activating proteins in epithelial cells (e.g., aldosterone-induced protein may act as a regulatory unit for the luminal membrane Na+ channel to allow luminal Na+ to enter the cell).89,90 Another aldosterone-induced protein may be a subunit of the Na+, K+ pump. This strengthens the argument that aldosterone not only has an indirect effect through ionic changes but also has a direct action on the Na+, K+ pump.88 The major sites for aldosterone action are in the connecting segment and in the collecting tubules.78 Aldosterone promotes NaCl reabsorption and K+ secretion in the connecting segments and in the principal cells in the cortical collecting tubules. In addition, it may act on Na+ in the papillary (inner medullary) collecting tubule and is thought to increase ionic transport by increasing the number of open Na+ and K+ channels in the luminal membrane.89, 90 Aldosterone also stimulates the Na+, K+ pump in the basolateral membrane. Na+ then diffuses into the tubular cells from which the Na+ is removed by the Na+, K+ pump, thus creating an electrogenic negative potential difference within the lumen. K+ is also secreted from the cell as Cl–moves in to maintain electroneutrality. Another mechanism for K+ efflux is through the Na+, K+ pump. It is believed that aldosterone mediates Na+ channel movement by the methylation of channel proteins that, in turn, activate silent or inoperative Na+ channels.90 Because amiloride blocks luminal Na+ permeability, it is likely that increasing luminal permeability is one of the primary actions of aldosterone.91 Intracellular calcium changes may serve as second messenger for this response by increasing the number of open Na+ channels.92 The intercalated cells in the renal cortex and the tubular cells of the outer medullary region seem to be the sites for aldosterone action on H+ secretion.93 These cells do not activate Na+ reabsorption; therefore, aldosterone-mediated H+ secretion and Na+ reabsorption occur in different regions of the kidney. However, there is a modest Na+/H+ exchange process mediated by aldosterone in the principal cells. Although aldosterone acts almost exclusively on the kidney, there are other epithelial sites (e.g., the colon, sweat and salivary glands) where the hormone reduces the Na+ and raises the K+ content of secretions.78 In very end-stage renal failure, the colonic secretion of K+ can become an important route of elimination. Aldosterone may be implicated in the fibrosis of left ventricular hypertrophy and congestive heart failure.94,95,96 and 97 In fact, the heart could be capable of de novo synthesis. The hormone may have effects on vascular remodeling and collagen formation as well as modifying endothelial cell function, and the cardiac angiotensin AT1 receptor may be one of its targets. 94,95,96 and 97 A local or tissue aldosterone, which has been found in vascular smooth muscle, may contribute to A-II–induced vascular hypertrophy.96 Thus, hypertension, cardiac fibrosis, and hypertrophy by nonepithelial cells may, in part, be consequences of high aldosterone levels, in either the primary or the secondary forms of hyperaldosteronism.95,96 Evidence suggests that many of the effects of aldosterone on blood pressure and the resultant pathologic changes could be mediated through specific miner-alocorticoid receptors in the AV3V region of the brain and in the heart.98 Some of these nonepithelial effects are blocked either by aldosterone-receptor antagonists (e.g., spironolactone) or by selective aldosterone-receptor agonists (e.g., eplerenone). In the Randomized Aldactone Evaluation Study (RALES) trial, the addition of low-dose spironolactone (25 mg) to standard therapy for congestive heart failure further reduced total mortality and/or the hospitalization rate for heart failure by 31%.99 Hyperaldosteronism occurs in renal failure despite normal renin levels. Furthermore, aldosterone appears to contribute to the progression of renal disease by causing fibrosis and glomerular damage, as seen in the rat remnant kidney model.100 More details of aldosterone action may be seen in a review.100a CONTROL OF ALDOSTERONE SECRETION The Renin-Angiotensin System. A-II is the major stimulator of aldosterone secretion. The proximal site of action of A-II is in enhancing the conversion of cholesterol to pregnenolone, but it also can modulate the distal regulatory site influencing the conversion of corticosterone to aldosterone by acting on aldosterone synthase.80,101,102 In general, the renin-angiotensin system is the major mediator of the volume-induced changes in aldosterone. Thus, volume and/or salt depletion will stimulate renin release, producing more A-II to act on the adrenal gland. The resulting Na+ retention then restores the volume to normal and shuts off renin release. The converse effect is seen with volume overload or with a high-salt diet, in which the renin-angiotensin system and aldosterone secretion are suppressed, allowing excess Na+ to be excreted. With volume expansion, there also is release of the ANPs and of dopamine, which, as inhibitors of aldosterone secretion, may act in concert with a low A-II level to limit the release of this mineralocorticoid.

Plasma Potassium Concentration. There is a direct relationship between increases in plasma K+ and increases in aldosterone secretion.103 Studies of K+ infusion in humans show that a 0.1- to 0.2-mEq/L change in plasma K+ can alter aldosterone levels, suggesting that the zona glomerulosa is exquisitely sensitive to K+ .104 In the feedback cycle, an increasing plasma K+ level stimulates aldosterone, which in turn increases K+ secretion and restores it to normal. This mechanism protects against excessive K+ in the body; it represents the single pathway for handling K+ overload. K+ has a direct effect on the zona glomerulosa cells to stimulate aldosterone synthesis, thereby activating the conversion of cholesterol to pregnenolone.105 Although both A-II and K+ act on the zona glomerulosa through different and independent effects, their cumulative action can be synergistic.106 Thus, A-II is more effective in releasing aldosterone when the subject is on a high K+ diet. This effect may be a result of adrenal renin, because in vitro studies show that K+ enhances adrenal renin and adrenal A-II. Both young and elderly subjects have similar basal serum potassium levels and similar responses to a K+ infusion; however, during a K+ infusion, hyperkalemia is more pronounced in elderly subjects because of a blunted adrenal response.107 Corticotropin. As with cortisol release, ACTH, acting through adenylate cyclase, acutely stimulates aldosterone secretion in a dose-dependent manner; however, this effect is not sustained.108 Thus, ACTH is not considered an important chronic regulator of aldosterone. ACTH initially stimulates and then suppresses the aldosterone-synthase gene.108 ACTH also directs steroid biosynthesis through the enzyme 17a-hydroxylase, which diverts synthesis away from the mineralocorticoid pathway and into the glucocorticoid and androgen pathways. Other Factors. The plasma Na+ concentration is a weak regulator of aldosterone; hyponatremia increases aldosterone levels, and hypernatremia possibly has the opposite effect.109 In humans, extreme changes in plasma Na+ are required to alter aldosterone secretion independently from more powerful stimuli (e.g., A-II and K+ ). Hyponatremia usually involves water retention, which (by suppression of A-II) overrides any direct action of Na+ on adrenal gland aldosterone synthesis.110 There is a modest effect of systemic acidosis on adrenal aldosterone release; this increases the excretion of acid and reduces systemic acidemia.111 The effect of aldosterone is thought to be through the H+ -ATPase pumps on the luminal membrane of the collecting tubules. Aldosterone escape occurs when, after several days of administration of this hormone, there is a loss of normal sodium- and fluid-retaining capacity, and a spontaneous diuresis occurs. This escape mechanism, which is caused by increases in ANPs and by a rise in renal perfusion pressure that limits Na+ retention, is observed in primary hyperaldosteronism and perhaps in the inherited forms of mineralocorticoid hypertension, conditions in which edema is not seen, despite a volume retention state.112 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

Corvol P, Jeunemaitre X. Molecular genetics of human hypertension: role of angiotensinogen. Endocr Rev 1997; 18(5):662. Kim HS, Krege JH, Kluckman KD, et al. Genetic control of blood pressure and the angiotensinogen locus. Proc Natl Acad Sci U S A 1995; 92:2735. Brasier AR, Li J. Mechanisms for inducible control of angiotensinogen gene transcription. Hypertension 1996; 27:465. Klett CP, Printz MP, Bader M, et al. Angiotensinogen messenger RNA stabilization by angiotensin II. J Hypertens 1996; 14(Suppl 1):S25. Wielbo D, Sernia C, Gyurko R, Phillips MJ. Antisense inhibition of hypertension in the spontaneously hypertensive rat. Hypertension 1995; 25:314. Naftilan AJ, Zuo WM, Ingelfinger J, et al. Localization and differential regulation of angiotensinogen mRNA expression in the vessel wall. J Clin Invest 1991; 88:1300. Schunkert H, Ingelfinger JR, Hirsch AT, et al. Evidence for tissue specific activation of renal angiotensinogen mRNA expression in chronic stable experimental heart failure. J Clin Invest 1992; 90:1523. lnoue I, Nakajima T, Williams CS, et al. A nucleotide substitution in the promoter of human angiotensinogen is associated with essential hypertension. J Clin Invest 1997; 99:1786. Corvol P, Persu A, Gimenez-Roqueplo AP, Jeunemaitre X. Seven lessons from two candidate genes in human essential hypertension: angiotensinogen and epithelial sodium channel. Hypertension 1999; 33:1324. Freire MBS, Ji L, Onuma T, et al. Gender-specific association of M235T polymorphism in angiotensinogen gene and diabetic nephropathy in NIDDM. Hypertension 1998; 31:896. Kunz R, Kreutze R, Beige J, et al. Association between the angiotensinogen 235T variant and essential hypertension in whites: a systematic review and methodological appraisal. Hypertension 1997; 30:1331. Staessen JA, Kuznetsova T, Wang JG, et al. M235T angiotensinogen polymorphism and cardiovascular renal risk. J Hypertens 1999; 17:9. Atwood LD, Kammerer CM, Samollow PB. Linkage of essential hypertension to the angiotensinogen locus in Mexican Americans. Hypertension 1997; 30:326. Niu T, Xu X, Rogus J, et al. Angiotensinogen gene and hypertension in Chinese. J Clin Invest 1998; 101:188. Rakugi H, Jacob HJ, Krieger JE, et al. Vascular injury induces angiotensinogen gene expression in the media and neointima. Circulation 1993; 87:283. Umemura S, Kihara M, Sumida Y, et al. Endocrinological abnormalities in angiotensinogen-gene knockout mice: studies of hormonal responses to dietary salt loading. J Hypertens 1998; 16:285. Navar LG, Inscho EW, Majid SA, et al. Paracrine regulation of the renal microcirculation. Physiol Rev 1996; 76:425. Skott O, Jensen BL. Cellular and intrarenal control of renin secretion. Clin Sci 1993; 84:1. Toffelmire EB, Slater K, Corvol P, et al. Response of prorenin and active renin to chronic and acute alterations of renin secretion in normal humans. Studies using a direct immunoradiometric assay J Clin Invest 1989; 83:679. Hosoi M, Kim S, Takada T, et al. Effects of prorenin on blood pressure and plasma renin concentrations in stroke-prone spontaneously hypertensive rats. Am J Physiol 1992; 262:E234. Gomez RA, Chevalier RL, Everett AD. Recruitment of renin gene-expressing cells in adult rat kidney. Am J Physiol 1990; 259:F660. Sigmund CD, Gross KW. Structure, expression, and regulation of the murine ren genes. Hypertension 1991; 18:446. Kopp U, DiBona GF. Interaction of renal b1-adrenoreceptors and prosta-glandins in reflex renin release. Am J Physiol 1983; 244:F418. Wagner C, Jensen BL, Kramer BK, Kurtz A. Control of the renal renin system by local factors. Kidney Int 1998; 54(Suppl 67):S78. Lee MA, Bohm M, Paul M, et al. Physiologic characterization of the hypertensive transgenic rat TGR(mREN2)27. Am J Physiol 1996; 270(Endocrinol Metab 33):E919. Erdos EG, Skidgel RA. Metabolism of bradykinin by peptidases in health and disease. In: Farmer SG, ed. The kinin system: handbook of immuno-pharmacology. London: Academic Press, 1997:112. Linz W, Wiemer G, Gohlke P, et al. Contribution of the kinins to the cardiovascular actions of angiotensin-converting enzyme inhibitors. Pharmacol Rev 1995; 47:25. Ramchandran R, Kasturi S, Douglas J, et al. Metalloprotease-mediated cleavage secretion of pulmonary ACE by vascular endothelial and kidney epithelial cells. Am J Physiol 1996; 271:H744. Soubrier F, Wei L, Hubert C, et al. Molecular biology of the angiotensin I-converting enzyme, II: structure function: gene polymorphism and clinical implications. J Hypertens l993; 11:599. Wei I, Clauser F, Athenc-Gelas F, Corval P. The two homologous domains of human angiotensin I-converting enzyme interact differently with competitive inhibitors. J Biol Chem 1992; 267:13398. Azizi M, Rousseau A, Ezan E, et al. Acute angiotensin-converting enzyme inhibition increases the plasma level of the natural stem cell regulator N-acetylseryl-lysyl-proline. J Clin Invest 1996; 97:839. Rigat B, Hubert C, Alhenc-Gelas F, et al. An insertion/deletion polymorphism in the angiotensin I-converting enzyme gene accounting for half the variance of serum enzyme levels. J Clin Invest 1990; 86:1343. Van der Kliej FGH, Schmidt A, Navis GJ, et al. Angiotensin converting enzyme insertion/deletion polymorphism and short-term renal response to ACE inhibition: role of sodium status. Kidney Int 1998; 53(Suppl 63):S23. Marre M, Bernadet P, Gallois Y, et al. Relationship between angiotensin I converting enzyme gene polymorphism, plasma levels, and diabetic retinal and renal complications. Diabetes 1994; 43:384. Biege J, Scherer S, Weber A, et al. Angiotensin-converting enzyme genotype and renal allograft survival. J Am Soc Nephrol 1997; 8:1319. Kimura H, Geiyo F, Suzuki S, et al. Polymorphisms of angiotensin converting enzyme and plasminogen activator inhibitor-1 genes in diabetes and microangiopathy. Kidney Int 1998; 54:1659. Urata H, Nishimura H, Ganten D. Chymase-dependent angiotensin II forming system in humans. Am J Hypertens 1996; 9:277. Harding JW, Wright JW, Swanson GN, et al. AT4 receptors: specificity and distribution. Kidney Int 1994; 46:1510. Chappel MC, Pirro NT, Sykes A, Ferrario CM. Metabolism of angiotensin (1-7) by angiotensin converting enzyme. Hypertension 1998; 31:362. Ferrario CM, Chappel MC, Tallant EA, et al. Counterregulatory actions of angiotensin (1-7). Hypertension 1997; 30:535. Unger T, Gohlke P. Tissue renin-angiotensin systems in the heart and vasculature: possible involvement in the cardiovascular actions of converting enzyme inhibitors. Am J Cardiol 1990; 65:31. Dostal DE, Baker KM. Evidence for a role of an intracardiac renin-angiotensin system in normal and failing hearts. Trends Cardiovasc Med 1993; 3:67. Martin P, Wagner J, Dzau VJ. Gene expression of the renin-angiotensin system in human tissues: quantitative analysis by polymerase chain reaction. J Clin Invest 1993; 91:2058. Wang DH, Du Y, Zhao H, et al. Regulation of angiotensin I receptor and its gene expression: role in renal growth. J Am Soc Nephrol 1997; 8:193. Yanagawa N. Potential role of local luminal angiotensin II in proximal tubule sodium transport. Kidney Int Suppl 1991; 32:S33. Kamide K, Hori M, Tuck ML, et al. Insulin-mediated growth in aortic smooth muscle and the vascular renin-angiotensin system. Hypertension 1998; 32:482. Nemeth G, Pepperell JR, Yamada Y, et al. The basis and evidence of a role for the ovarian renin-angiotensin system in health and disease. J Soc Gynecol Investig 1994; 1:118. Speth RC, Daubert DL, Grove KL. Angiotensin II: a reproductive hormone too? Regul Pept 1999; 79(l):25. Hageman JR, Moyer JS, Bachman ES, et al. Angiotensin-converting enzyme and male fertility. Proc Natl Acad Sci U S A 1998; 95:2552. Jones BH, Standbridge MK, Taylor JW, et al. Angiotensinogen gene expression in adipose tissue, analysis of obese models and hormonal and nutritional control. Am J Physiol 1997; 273(Regulatory Integrative Comp Physiol 42):236. Kifor I, Moore TJ, Fallo F, et al. Potassium-stimulated angiotensin release from superfused adrenal capsules and enzymatically digested cells of the zona glomerulosa. Endocrinology 1991; 129:823. Phillips MI, Speakman EA, Kimura B. Levels of angiotensin and molecular biology of the tissue renin angiotensin system. Regul Pept 1993; 43:1. Baltatu O, Nishimura H, Hoffman S, et al. High levels of human chymase expression in the pineal and pituitary gland. Brain Res 1997; 759:269. Griendling KK, Lassegue B, Alexander RW. The vascular angiotensin (AT1) receptor. Thromb Haemost 1993; 70(1):188. Coffman TM. Gene targeting in physiological investigations: studies of the renin-angiotensin system. Am J Physiol 1998; 274 (Renal Physiol):F999. Sadoshima J. Versatility of the angiotensin II type 1 receptor. Circ Res 1998; 82:1352.

56a. Schiffrin EL, Park JB, Intengan HD, Touyz RM. Correction of arterial structure and endothelial dysfunction in human essential hypertension by the angiotensin receptor antagonist losartan. Circulation 2000; 101(14):1653. 57. 58. 59. 60. 61. 62. 63. 64.

Horiuchi M, Akishita M, Dzau VJ. Recent progress in angiotensin II type 2 receptor research in the cardiovascular system. Hypertension 1999; 33:613. Cogan MG. Angiotensin II: a powerful controller of sodium transport in the early proximal tubule. Hypertension 1990; 15:451. Okubo S, Numura F, Matsusaka T, et al. Angiotensinogen gene null-mutant lacks homeostatic regulation of glomerular filtration and tubular reabsorption. Kidney Int 1998; 53:617. Scholz H, Kurtz A. Role of protein kinase C in renal vasoconstriction caused by angiotensin II. Am J Physiol 1990; 259:C421. Clemson B, Gaul L, Gubin SS, et al. Prejunctional angiotensin II receptors. Facilitation of norepinephrine release in the human forearm. J Clin Invest 1994; 93:684. Moreau P, d'Uscio LV, Shaw S, et al. Angiotensin II increases tissue endothelin and induces vascular hypertrophy: reversal by ET(A)-receptor antagonist. Circulation 1997; 96:1593. Golub MS, Hori MT, Tuck ML. Arachidonic acid metabolites in the vascutature. In: Sowers JR, ed. Endocrinology of the vasculature. Totowa, NJ: Humana Press, 1996:357. Ichikawa I, Harris RC. Angiotensin actions in the kidney: renewed insight into the old hormone. Kidney Int 1991; 40:583.

65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87.

Wilcox CS, Welch WJ, Snellen H. Thromboxane mediates renal hemodynamic response to infused angiotensin II. Kidney Int 1991; 40:1090. Miller JA, Anacata LA, Cattran DC. Impact of gender on the response to angiotensin II. Kidney Int 1999; 55:278. Wolf G, Ziyadeh FN. The role of angiotensin II in diabetic nephropathy: emphasis on nonhemodynamic mechanisms. Am J Kidney Dis 1997; 29:153. Dzau VJ, Re R. Tissue angiotensin system in cardiovascular medicine: a paradigm shift. Circulation 1994; 89:493. Gibbons GH. The pathophysiology of hypertension. The importance of angiotensin II in cardiovascular remodeling. Am J Hypertens 1998; 11:177S. Alderman MH, Madhaven S, Ooi WL, et al. Association of renin-sodium profile with risk of myocardial infarction in patients with hypertension. N Engl J Med 1991; 324:1098. Vaughan DE. The renin-angiotensin system and fibrinolysis. Am J Cardiol 1997; 79(5A):12. Brown NJ, Agirbasli MA, Williams GH, et al. Effect of activation and inhibition of the renin-angiotensin system on plasma PAI-1. Hypertension 1998; 32:965. . Hwan K, Thornhill B, Wolstenholme T, Chevalier RL. Tissue-specific regulation of growth factors and clusterin by angiotensin II. Am J Hypertens 1998; 11:715. Morishita R, Ogihara T. In vivo evaluation of in vitro hypothesis using “gain or loss” of function: functional analysis of the renin-angiotensin system. Am J Hypertens 1998; 11:507. Tomita N, Morishita R, Higaki J, et al. Transient decrease in high blood pressure by in vitro transfer of antisense oligodeoxynucleotides against angiotensinogen. Hypertension 1995; 26:131. Phillips MI. Antisense inhibition and adeno-associated viral vector delivery for reducing hypertension. Hypertension 1997; 29:177. Marver D, Kokko JP. Renal target sites and the mechanism of action of aldosterone. Min Elect Metab 1983; 9:1. White PC. Mechanisms of disease. Disorders of aldosterone biosynthesis and action. N Engl J Med 1994; 331:250. Connell JM, Fraser R. Adrenal cortical synthesis and hypertension. J Hypertens 1991; 9:97. Holland OB, Carr B. Modulation of aldosterone synthase messenger ribonucleic acid levels by dietary sodium and potassium and by adrenocorticotropin. Endocrinology 1993; 132:2666. Arriza JL, Weinberger C, Cerelli C, et al. Cloning of human mineralocorticoid receptor complementary DNA: structural and functional kinship with the glucocorticoid receptor. Science 1987; 237:268. Funder JW. Mineralocorticoids, glucocorticoids, receptors and response elements. Science 1993; 259:1132. Bastl CP, Hayslett. The cellular action of aldsterone in target epithelia. Kidney Int 1992; 42:250. Zennaro M-C, Farman N, Bonvalet J-P, et al. Tissue-specific expression of alpha and beta messenger ribonucleic acid isoforms of the human mineral-ocorticoid receptor in normal and pathological states. J Clin Endocrinol Metab 1997; 82:1345. Funder JW. Enzymes and the regulation of sodium balance. Kidney Int 1992; 37(Suppl):S114. Kenouch S, Coutry N, Farman N, Bonvalet J-P. Multiple patterns of 11b-hydroxysteroid dehydrogenase catalytic activity along the mammalian nephron. Kidney Int 1992; 42:56. Naray-Fejes-Toth A, Rusvai E, Fejes-Toth G. Mineralocorticoid receptors and 11b-hydroxysteroid dehydrogenase activity in renal principal and intercalated cells. Am J Physiol 1994; 266:F76.

87a. Funder JW. Aldosterone and mineralocorticoid receptors: orphan questions. Kidney Int 2000; 57(4):1358. 88. 89. 90. 91. 92.

Horisberger JD, Rossier BC. Aldosterone regulation of gene transcription leading to control of ion transport. Hypertension 1992; 19:221. Szerlip H, Palevsky P, Cox M, Blazer-Yost B. Relationship of the aldoster-one-induced protein GP70 to the conductive channel. J Am Soc Nephrol 1991; 2:1108. Schafer JA, Hawk CT. Regulation of Na+ channels in the cortical collecting duct by AVP and mineralocorticoids. Kidney Int 1992; 41:255. Hayhurst RA, O'Neil RG. Time-dependent actions of aldosterone and amiloride-blockable sodium channels in A6 epithelia. Am J Physiol 1988; 254:F689. Petzel D, Ganz MB, Nestler EJ, et al. Correlates of aldosterone-induced increases in Ca i 2+ and Isc suggest that Cai 2+ is the second messenger for stimulation of apical membrane conductance. J Clin Invest 1992; 89:150. 93. Hays SR. Mineratocorticoid modulation of apical and basolateral membrane H+/OH–/HCO3–, transport processes in rabbit inner strip of outer medullary collecting duct. J Clin Invest 1992; 90:180. 94. Brilla CG, Weber KT. Mineralocorticoid excess, dietary medium and myocardial fibrosis. J Lab Clin Med 1992; 120:893. 95. Young M, Fullerton M, Dilley P, Funder J. Mineralocorticoids, hypertension and cardiac fibrosis. J Clin Invest 1994; 93:2578. 95a. Falkenstein E, Christ M, Feuring M, Wehling M. Specific nongenomic actions of aldosterone. Kidney Int 2000; 57(4):1390. 96. 97. 98. 99.

Hakeyama H, Miramuri I, Fujita T, et al. Vascular aldosterone, biosynthesis and a link to angiotensin II-induced hypertrophy of vascular smooth muscle. J Biol Chem 1994; 269:24316. Robert V, Heymes C, Silvestre J-B, et al. Angiotensin AT1 receptor subtype as a cardiac target of aldosterone; role in aldosterone-salt-induced fibrosis. Hypertension 1999; 33:981. Gomez Sanchez EP. What is the role of the central nervous system in mineralocorticoid hypertension? Am J Hypertension 1991; 4:374. The RALES Investigators. Effectiveness of spironolactone added to an angiotensin converting enzyme inhibitor and a loop diuretic for severe chronic congestive heart failure (The Randomized Aldactone Evaluation Study [RALES]). Am J Cardiol 1996; 78:902. 100. Ibrahim HN, Rosenberg ME, Greene EL, et al. Aldosterone is a major factor in the progression of renal disease. Kidney Int 1997; 52(Suppl 63):S115. 100a. Farman N, Verrey F. Forefronts in nephrology news in aldosterone action. Kidney Int 2000; 57:1239. 101. Brann DW, Hendry LB, Mahesh VB. Emerging diversities in the mechanism of action of steroid hormones. J Steroid Biochem Molec Biol 1995; 52:113. 102. Adler GK, Chen R, Menachery AI, et al. Sodium restriction increases aldosterone synthesis by increased late pathway, not early pathway, messenger ribonucleic acid levels and enzyme activity in normal rats. Endocrinology 1993; 133:2235. 103. Young DB. Quantitative analysis of aldosterone's role in potassium regulation. Am J Physiol 1988; 255:F811. 104. Himathongan T, Dluhy R, Williams GH. Potassium-aldosterone-renin interrelationships. J Clin Endocrinol Metab 1975; 41:153. 105. Kifor I, Moore TJ, Fallo F, et al. Potassium-stimulated angiotensin release from superfused adrenal capsules and enzymatically digested cells of the zona glomerulosa. Endocrinology 1991; 129:823. 106. Young DB, Smith MJ Jr, Jackson TE, Scott RE. Multiplicative interaction between angiotensin II and K concentration in stimulation of aldosterone. Am J Physiol 1984; 247:E328. 107. Mulkerrin E, Epstein FH, Clark BA. Aldosterone responses to hyperkalemia in healthy elderly humans. J Am Soc Nephrol 1995; 6:1459. 108. Braley LM, Adler GK, Mortensen RM, et al. Dose effect of adrenocorticotropin on aldosterone and cortisol biosynthesis in cultured bovine adrenal glomerulosa cells. In vitro correlate of hyperreninemic hypoaldosterone. Endocrinology 1992; 131:187. 109. Merrill DC, Ebert TJ, Skelton MM, Cowley AW Jr. Effect of plasma sodium on aldosterone during angiotensin II stimulation in normal humans. Hypertension 1989; 14:164. 110. Taylor RE, Glass GT, Radke KJ, Schneider EG. Specificity of effect of osmolality on aldosterone secretion. Am J Physiol 1987; 252:E118. 111. Jones GV, Wall BM, Williams HH, et al. Modulation of plasma aldosterone by physiologic changes in hydrogen ion concentration. Am J Physiol 1992; 262:R269. 112. White PC. Inherited forms of mineralocorticoid hypertension. Hypertension 1996; 28:927.

CHAPTER 80 HYPERALDOSTERONISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 80 HYPERALDOSTERONISM JOHN R. GILL, JR. Disorders of Mineralocorticoid Excess Physiology of aldosterone Synthesis Secretion Actions “Primary Aldosteronism” Aldosterone-Producing Adenomas Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Idiopathic Hyperaldosteronism Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Dexamethasone-Suppressible Hyperaldosteronism Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Adrenocortical Carcinoma Clinical Features and Pathophysiology Diagnosis and Treatment Syndromes of Real or Apparent Mineralocorticoid Excess not Caused by Aldosterone Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Secondary Hyperaldosteronism Bartter Syndrome Clinical Features and Pathophysiology Diagnosis and Differential Diagnosis Treatment Chapter References

DISORDERS OF MINERALOCORTICOID EXCESS Mineralocorticoid excess, whether primary or secondary, and disorders that mimic it are commonly manifested clinically as hypokalemia, which usually but not always is accompanied by alkalosis. Thus, a useful strategy for diagnosing hypokalemia is to approach the evaluation as a differential diagnosis of disorders of mineralocorticoid excess. Clinical features that accompany hypokalemia and may be used to determine the specific disorders that need to be considered are illustrated in Table 80-1 and in the flow chart in Figure 80-1. This basic framework is expanded in the course of this review to include the disorders appropriate to each of the four listed patterns of abnormalities of renin and aldosterone. Ideally, evaluation of the renin and aldosterone systems should be done in the absence of medication during a standardized sodium intake (e.g., 100 mEq per day of sodium as sodium chloride). If this is not possible, then at least the use of medications that affect the production of renin and aldosterone (i.e., diuretics and angiotensin converting enzyme inhibitors) should be discontinued and the blood pressure should be controlled with a calcium channel blocking agent. The level of sodium intake immediately preceding the sampling of blood for plasma renin activity and plasma aldosteroneshould be estimated by collecting a 24-hour aliquot of urine for the determination of sodium excretion. The blood for the plasma renin activity and plasma aldosterone evaluations should be drawn after overnight bedrest (in the hospital) or after 30 minutes of bedrest (in the clinic), and then drawn again after 2 hours of standing and walking.

TABLE 80-1. Disorders Characterized by Hypokalemia and Alkalosis

FIGURE 80-1. Flow chart for evaluation of patients with hypokalemic alkalosis. (PRA, plasma renin activity; ALDO, aldosterone.) (See Table 80-1 for continuation of evaluation.)

PHYSIOLOGY OF ALDOSTERONE SYNTHESIS The zona glomerulosa cells of the adrenal cortex synthesize aldosterone from corticosterone in a two-step, mixed function oxidation reaction that is catalyzed by the enzyme corticosterone methyl oxidase1 (Fig. 80-2). Normally, the adrenal glands secrete 60 to 200 µg aldosterone each day, depending on the state of sodium chloride and water balance of the body; the steroid circulates in the blood either free or loosely bound to albumin. Although some of the circulating aldosterone may be excreted in the urine in the free form, most of it is metabolized by the liver to tetrahy-droaldosterone and by the kidneys to aldosterone-18-glucuronide. Measurement of the excretion of the 18-glucuronide metabolite has been used as an index of aldosterone production; values obtained by this method range from 1 to 15 µg per day in normal persons receiving an unrestricted sodium chloride intake.

FIGURE 80-2. Proposed mechanism and structure of intermediates in the conversion of corticosterone to aldosterone. (From Ulick S. Diagnosis and nomenclature of the disorders of the terminal portion of the aldosterone biosynthetic pathway. J Clin Endocrinol Metab 1976; 43:92.)

SECRETION Aldosterone biogenesis is regulated primarily by the renin-angiotensin system, with potassium, atrial natriuretic hormone (ANH), anddopamine also making important contributions. Corticotropin (ACTH) also may stimulate aldosterone production, but its effects are self-limited except in certain disease states. In addition, the factors that determine aldosterone secretion have direct effects on the kidney and vasculature and, along with aldosterone, contribute to the maintenance of the volume of extracellular fluid, the concentration of extracellular potassium, and, ultimately, blood pressure. Thus, when the intake of sodium chloride is restricted and blood volume is contracted, an increase in the formation of angiotensin II stimulates the secretion of aldosterone. Angiotensin II and aldosterone, in turn, both increase the reabsorption of sodium by the renal tubule to curtail sodium loss and restore extracellular volume. Conversely, when sodium chloride intake is increased and blood volume is expanded, the formation of angiotensin II decreases, ANH is secreted by the heart, and the formation of dopamine by the adrenal glands and kidneys increases. ANH and dopamine inhibit aldosterone biogenesis; they also increase sodium excretion through direct effects on the kidney. Potassium exerts differential effects on aldosterone biosynthesis. Hyperkalemia, secondary to potassium administration or retention, stimulates aldosterone secretion, whereas hypokalemia, secondary to potassium depletion, inhibits it. A more detailed discussion of the renin-angiotensin system and its control of aldosterone secretion is presented in Chapter 79 and Chapter 183; ANH is discussed in Chapter 178. ACTIONS Aldosterone exerts its biologic actions by crossing the plasma membrane of target cells and binding to its receptor in the cytosol.2 The aldosterone-receptor complex then migrates to the nucleus, where it binds, thereby initiating transcription, translation, and synthesis of proteins responsible for expression of the action of aldosterone on the functions of that target cell (see Chap. 4). Collecting tubule cells of the kidney and epithelial cells of other transporting tissues (i.e., sweat, salivary gland, and intestine) are target tissues in which the physiologic effects of aldosterone are most discernible, although effects on vascular3 and other tissues may occur. In epithelial cells for which the physiologic actions of aldosterone have been extensively characterized, the steroid facilitates passage of sodium across the apical membrane into the cell and accelerates extrusion of sodium and uptake of potassium by Na+,K+-adenosine triphosphatase located in the basolateral membrane.4 In the renal collecting tubule, these actions of aldosterone increase the reabsorption of sodium and the secretion of potassium, leading to retention of sodium by the body and loss of potassium (see Chap. 206). As the foregoing discussion indicates, the renin-angiotensin-aldosteronesystems regulate the volume of extracellular fluid and contribute to the maintenance of the extracellular potassium concentration. As a result of disease, the production of aldosterone may become supernormal and cease to be responsive to the physiologic stimuli that regulate it. The overproduction of aldosterone may result from overactivity of the renin-angiotensin system (secondary hyperaldosteronism) or from other abnormalities (primary hyperaldosteronism). The consequences of continual overproduction of aldosterone, or hyperaldosteronism, depend in part on the disorder with which it is associated, and may include excessive sodium chloride and water retention, potassium loss, alkalosis, and hypertension.

“PRIMARY ALDOSTERONISM” Autonomous overproduction of aldosterone occurs in ~2% of patients with hypertension and was recognized initially as being associated with either a unilateral adenoma (aldosterone-producing adenoma, or APA) or bilateral hyperplasia of the zona glomerulosa. Because patients with bilateral adrenal disease tended to have a clinical presentation similar to that of patients with an APA (hypertension and hypokalemia) and, like the patients with unilateral disease, they showed suppression of the renin-angiotensin system, they also were assumed to have a primary adrenal disorder. Subsequent observations have not proved this to be true; they suggest, instead, that bilateral hyperplasia of the zona glomerulosa probably is not a primary adrenal disorder in most cases and may have more than one cause. Unilateral adrenalectomy usually cures or ameliorates the hypertension in patients with an APA, whereas subtotal or total adrenalectomy, with few exceptions, has little effect on the blood pressure of patients with bilateral hyperplasia. In the case of those few patients who have primary adrenal hyperplasia, unilateral adrenalectomy may be curative.5 In some patients with bilateral hyperplasia, the hyper-aldosteronism was dependent on ACTH (so-called glucocorticoid-remediable hyperaldosteronism). The hypothesis that a trophic hormone, possibly of pituitary origin, may be responsible for bilateral hyperplasia not caused by ACTH has received considerable support, although it has not been definitively proved. As a result of advances in knowledge, the inclusive designation of “primary hyperaldosteronism” for those patients with hyperal-dosteronism associated with both adenoma and bilateral hyperplasia has tended to be replaced by the more specific designations of aldosterone-producing adenoma, idiopathic hyperaldosteronism, primary adrenal hyperplasia, and glucocorticoid-remediable hyper-aldosteronism. Rarely, adrenocortical carcinoma may overproduce aldosterone. These five disorders present as hypokalemic alkalosis, hypertension, low or suppressed plasma renin activity, and high aldosterone concentrations (see Table 80-1).

ALDOSTERONE-PRODUCING ADENOMAS CLINICAL FEATURES AND PATHOPHYSIOLOGY Aldosterone-producing adrenocortical adenomas usually are small (0.5–2.5 cm), unilateral, solitary, and associated with hypoplastic zona glomerulosa. Occasionally, however, these lesions may be multiple and associated with hyperplasia of the zona glomerulosa. This benign tumor of the adrenal cortex occurs most frequently in the third through the fifth decades of life but also has been observed in prepubertal children6 and older persons (range, 13–66 years). Aldosterone-producing adenomas occur equally in men and women and do not appear to have a racial predisposition; they tend to develop more frequently in the left adrenal gland than in the right, but this difference is small. (The occurrence of primary aldosteronism resulting from APA in two or more members of the same family has been reported,7 as well as the familial occurrence of idiopathic aldo-steronism. In one of the families, one member had bilateral nodular hyperplasia and another had APA. This familial form of primary aldosteronism was labeled familial hyperaldosteronism type II to distinguish it from glucocorticoid-remediable hyperaldosteronism. The clinical features of familial hyperal-dosteronism type II do not differ from those of sporadic patients with APA or idiopathic hyperaldosteronism.) The symptoms reported most frequently by patients with an APA are headache, easy fatigability, and weakness, although high blood pressure may be the only symptom in many cases. Hypokalemia, suppressed plasma renin activity, and high plasma and urinary aldosterone values are the findings usually associated with APA, but they may not always be demonstrable.8 Aldosterone production by an APA is autonomous and is affected only minimally by angiotensin II because of the decreased formation of this peptide as well as the decreased responsiveness of the adenoma. An APA is unresponsive to the infusion of dopamine, although endogenous dopamine may exert tonic inhibitory effects.9 ANH, levels of which appear to be elevated in patients with this disorder,10 does not inhibit aldosterone biogenesis by APA cells in vitro, as it does in normal glomerulosa cells.11 Therefore, this hormone may have little effect on aldosterone secretion by an APA in vivo. In contrast, changes in serum potassium or ACTH levels may exert important regulatory effects on aldosterone secretion in patients with this tumor: aldosterone production may be blunted by the potassium loss and the hypokalemia that it induces. The circadian rhythm of aldosterone in patients with APA tends to mirror that of ACTH. As a consequence of sustained overproduction of aldosterone, retention of sodium chloride and water occurs, but this usually is limited; “escape” occurs when extracellular fluid has been expanded by ~500 mL.12 Initially, the renal excretion of potassium increases. As hypokalemia develops, potassium excretion decreases to reflect intake of this cation. Hydrogen ion excretion, which increases principally because of an increase in urinary ammonium, is responsible for the development and

maintenance of metabolic alkalosis. Urinary calcium and magnesium levels are high; presumably, this is a reflection of the effects of expansion of extracellular fluid on the tubular reabsorption of these ions. The urinary concentrating ability is impaired because of potassium depletion. Although sodium and water retention may contribute to the increase in blood pressure that occurs, the precise mechanisms responsible for an increase in peripheral resistance and the associated hypertension are unclear. Receptors for aldosterone have been found in blood vessels and in the brain, and the effects of aldosterone on ion transport in these tissues may be important mediators of the increase in blood pressure. Preliminary studies in dogs indicate that an amount of aldosterone that is too small to exert an effect when infused peripherally increases blood pressure when it is infused into the third ventricle of the brain.13 DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS Patients with hypertension who have borderline or low serum potassium concentrations, low stimulated plasma renin activity (14 ng/dL) should be evaluated further for autonomous overproductionof aldosterone. The evaluation should be done before antihypertensive treatment is instituted or after it has been discontinued for 2 weeks. Determining aldosterone suppressibility, by combining a high sodium intake and treatment with a mineralocorticoid such as 9-afluorohydrocortisone for 3 or more days, or by intravenous saline infusion, is a useful first step. Of these two interventions, the infusion of 2 L normal saline over 4 hours with the measurement of plasma aldosterone levels before and at the conclusion of the infusion is simpler and may be readily adapted to the evaluation of outpatients.14 At the conclusion of the infusion, the plasma aldosterone level normally is 5 ng/dL or less; in patients with hyperaldosteronism, it usually exceeds 10 ng/dL. Determination of the plasma aldosterone to plasma renin activity ratio also has been used to detect aldosteronism. Single determinations of this ratio tend to be of limited use because values from patients with hyperaldosteronism frequently are the same as those from patients with normal adrenal function. A solution to the problem of overlap that appears to improve specificity considerably is to collect blood samples every 30 minutes for 6 hours, so that an integrated plasma aldosterone/plasma renin activity ratio can be determined.15 A somewhat simpler but probably equally effective alternative is to collect blood 90 minutes after the administration of 25 to 50 mg captopril, which is a converting enzyme inhibitor, for determination of the plasma aldosterone/plasma renin activity ratio16 (Fig. 80-3). Both procedures are safe and do not require dietary preparation or hospitalization. It has been suggested that all patients with hypertension undergo captopril screening tests,17 because determination of the serum potassium concentration and plasma renin activity may be inadequate measures for screening the hypertensive population for hyperaldosteronism.8 The normogram in Figure 80-3 indicates the degree of separation of patients with aldosterone excess from those with essential hypertension.16 Subsequent evaluation of the captopril test indicates a false-negative rate of 6.3% and a false-positive rate of 0.6% in patients with primary hyperaldosteronism.18 The problem of identifying those patients with hyperaldosteronism who harbor an adenoma has been greatly facilitated by technologic advances. Computed tomography (CT) appears to have the capacity to discern an adrenal mass in ~80% of patients with an aldosterone-producing adenoma.19,20 Magnetic resonance imaging also has been used to evaluate the adrenal glands in patients with hyperaldosteronism; apart from confirming the findings of CT, however, it does not appear to offer any advantage over CT. These new techniques for imaging adrenal masses have replaced the more cumbersome and time-consuming procedure of131 Iiodocholesterol imaging in most centers (see Chap. 88). A limitation of all the imaging techniques, however, is that an adenoma may be poorly visualized or not detected at all. These problems result from the small size of some adenomas, which may be less than 1 cm in diameter. The problem of an inconclusive CT scan may be partially resolved by determining the plasma 18-hydroxycorticosterone level after overnight bedrest.21 In a series of 50 patients with aldosteronism, all but 1 of the 24 patients with an adenoma had 18-hydroxycorticosterone values that exceeded 50 ng/dL.8 Thus, a plasma 18-hydroxycorticosterone value >50 ng/dL in a patient with hyperaldosteronism suggests that the diagnosis is more likely to be an aldosterone-producing adenoma than bilateral hyperplasia, even though the adrenal glands may appear normal (Fig. 80-4). If the plasma aldosterone level were to show a “paradoxic” fall paralleling the circadian fall in plasma cortisol during 2 hours of ambulation after early morning sampling, this also would suggest that the diagnosis is more likely to be an aldosterone-producing adenoma than bilateral hyperplasia.22 So many patients with an adenoma, like those with bilateral hyperplasia, show a rise rather than a fall in the plasma aldosterone level that the usefulness of the procedure is limited.8

FIGURE 80-3. Values for the plasma aldosterone (ALDO)/plasma renin activity (PRA) ratio plotted as a function of plasma aldosterone for normal and hypertensive control subjects (hatched area), patients with aldosterone-producing adenoma (APA), and patients with idiopathic hyperaldosteronism (IHA) 2 hours after the administrationof 25 mg captopril. The plasma aldosterone value decreased to 90% in the lateralization of an adenoma, making the distinction between an APA and those disorders associated with bilateral overproduction of aldosterone. The observations that APA may be associated with other abnormalities seen on CT scans such as ipsilateral and contralateral nonfunctioning nodules and thickening of an adrenal limb indicate the important role of adrenal vein sampling in reaching a correct diagnosis.20 Only occasionally will a patient with idiopathic aldosteronism have a gradient between the two adrenal glands with an aldosterone/cortisol ratio that approaches 4 to 1, the lower limit of the range of the ratio of APA to normal adrenal.24

TABLE 80-2. Adrenal Venous Sampling

TREATMENT The preferred therapy for patients with an aldosterone-producing adenoma is operative removal of the adrenal gland containing the tumor.25 Adenomectomy alone is not sufficient.25a This can be done laparoscopically with minimal morbidity26 (see Chap. 89). The cure rate (correction of hypertension as well as aldosteronism) ranges from 60% to 75%. Because the prevalence of a family history of hypertension in patients with an aldosterone-producing adenoma may range from 40% to 60%, this, along with the duration of the hypertension, may account for the persistence of an elevated blood pressure despite the correction of hyperaldosteronism. Most of the patients who remain hypertensive after the correction of hyperaldosteronism become more responsive to antihypertensive medication. Spironolactone is probably the single most effective drug for treating patients with an APA. It acts as a competitive inhibitor, competing with aldosterone for its cytosol receptor, thereby antagonizing the action of the mineralocorticoid in target tissue. Spironolactone also is both an antiandrogen and a progestagen, and this explains many of its distressing side effects;2,4 decreased libido, mastodynia, and gynecomastia may occur in 50% or more ofmen, and menometrorrhagia and mastodynia may occur in an equally large number of women taking the drug.27 The problem of side effects may preclude the long-term use of spironolactone, particularly in younger men and women, and is more likely to occur when the dosage exceeds 100 mg per day. The usual dosage of spironolactone ranges from 50 mg once daily to 100 mg twice daily. Drugs such as amiloride and triamterene also may oppose the effect of aldosterone on the renal tubule, but these agents act by inhibiting sodium reabsorption and potassium secretion through a direct effect on the tubule cell, not by competing with aldosterone for its receptor. This may explain why triamterene and amiloride tend to be less effective than spironolactone as antihypertensive agents in patients with hyperaldosteronism. Calcium channel blocking drugs, such as nifedipine, also may be useful therapeutic agents in patients with hyperaldo-steronism. In addition to its antihypertensive action, nifedipine may decrease aldosterone production.28 Thus, the combined use of nifedipine and a potassium-sparing diuretic may serve as an alternative to spironolactone in the medical treatment of patients with an APA.

IDIOPATHIC HYPERALDOSTERONISM CLINICAL FEATURES AND PATHOPHYSIOLOGY Idiopathic bilateral adrenocortical hyperplasia of the zona glomerulosa (idiopathic hyperaldosteronism) is characterized by features similar to those associated with APAs, although they are less pronounced in many patients.29 The reported prevalence of idiopathic hyperaldosteronism has ranged from a high of 70% of those patients with primary hyperaldosteronism14 to a low of 8%.20 The true prevalence probably is somewhere between these two extremes and may be ~30%. Although the etiology of idiopathic aldosteronism has not been established, the belief that a circulating stimulatory factor is responsible for hyperfunction of the zona glomerulosa is generally accepted. Extracts of urine from normal persons contain a protein fraction that selectively stimulates the production of aldosterone and produces hypertension when injected into rats.30 Subsequent studies of one of these extracts have shown that (a) it is a peptide of pituitary origin, (b) it is increased in patients with idiopathic hyperaldosteronism but not in those with an APA, and (c) it is not suppressed by dexamethasone.31 The specific peptide has not been identified. In other studies, plasma concentrations of g-melanotropin32 and of b-endorphin33 were increased in patients with idiopathic hyperaldosteronism but not in those with an aldosterone-producing adenoma. These two peptides are fragments of a larger peptide, pro-opiomelanocortin, which is formed in the pituitary; they may stimulate the secretion of aldosterone or increase the sensitivity of the adrenal gland to angiotensin II, or both. The report of an enlargement of the intermediate lobe of the pituitary in a patient with idiopathic hyperaldosteronism,34 together with the observations that central serotoninergic stimulation of aldosterone secretion may occur35 and that cypro-heptadine, which is an inhibitor of central serotonin release, may decrease plasma aldosterone in patients with idiopathic hyperaldosteronism,36 is consistent with a possible role for the pituitary in the pathogenesis of idiopathic hyperaldosteronism. It should be noted, however, that a pathogenetic schema for idiopathic hyperaldosteronism has to account for the hypertension as well as the hyperaldosteronism because adrenalectomy rarely lowers blood pressure. Therefore, a putative aldoster-one-stimulating factor of pituitary or other origin, orchanges proximal to its release, must be responsible for the hypertension in idiopathic hyperaldosteronism, unless the hypertension is a separate process. The effects of the putative aldosterone-stimulating factors on blood pressure are not known. As in an APA, the overproduction of aldosterone in idiopathic hyperal-dosteronism causes increased sodium reabsorption and potassium secretion by the renal tubule, expansion of the volume of extracellular fluid, suppression of plasma renin activity, and hypokalemia. DIAGNOSIS AND DIFFERENTIAL DIAGNOSIS Although the features that characterize prolonged aldosterone excess may be as prominent in patients with idiopathic hyperal-dosteronism as in patients with an APA, they usually are less pronounced and more subtle.29 (It has been suggested that approximately half the patients with hypertension and suppressed plasma renin activity who show a decrease in plasma aldosterone to values between 5 and 10 ng/dL in response to saline infusion may have idiopathic hyperaldosteronism, even though baseline plasma aldosterone and serum potassium values may be normal.14 When these diagnostic criteria are used, it may be difficult to distinguish between idiopathic hyperaldo-steronism and low-renin essential hypertension. Such a sharp distinction between idiopathic hyperaldosteronism and low-renin essential hypertension may not be made with certainty until the pathogenesis of both disorders is more clearly delineated.) More important distinctions are those between idiopathic hyperaldosteronism and an APA, and between idiopathic hyperaldosteronism and primary bilateral adrenal hyperplasia, because therapy for these disorders is different. The response to saline infusion may be helpful because the aldosterone/cortisol ratio increases in an APA but remains unchanged or decreases, with a value of 2.2 or less, in idiopathic hyperaldosteronism37 (Fig.80-5). A value for plasma 18-hydroxycorticosterone less than 50 ng/dL at 8:00 a.m. after overnight bedrest is the usual finding in idiopathic hyperaldosteronism.8,21 In patients with an APA, as well as in most patients with primary hyperplasia, values for 18-hydroxycorticosterone are >50 ng/dL and may exceed 100 ng/dL.5,21,38 An increase in plasma aldosterone levels associated with a decrease in plasma cortisol levels after 2 hours of ambulation in the morning favors a diagnosis of idiopathic hyperaldosteronism but also may occur in association with an APA.22 An increase in plasma aldosterone levels is more difficult to interpret than is a paradoxic decrease. Because of the number of false-positive and false-negative results, the test, as originally described, is not sufficiently dependable to be a reliable tool.8 A refinement in the interpretation of the postural stimulation test (i.e., correction of the percentage increase in aldosterone by subtraction of the percentage increase in cortisol with standing) and acceptance of an increase in levels of plasma aldosterone with standing of 1.5 cm disparity in pole-to-pole diameter) and a unilateral delay in the appearance of contrast medium in the 1- to 5-minute exposures.3 False-negative results often occur in cases of renovascular hypertension because of bilateral renal artery disease.3,21 The hypertensive urogram has low sensitivity (a negative test does not exclude renovascular hypertension) and specificity and, importantly, has the risk of nephrotoxicity because of contrast agents. Plasma Renin Activity. Although hypoperfusion in the stenotic kidney activates the renin-angiotensin system, renin secretion in the nonstenotic kidney is often normal or low, so that the baseline or unstimulated circulating PRA levels are normal in most cases. Thus, when measured in peripheral venous samples, PRA is normal in 20% to 50% of cases of documented renovascular hypertension.7,13,20 The interpretation is confounded by the fact that 16% to 20% of subjects with essential hypertension have PRA values that are moderately elevated. However, in renovascular hypertension there are exaggerated PRA responses to procedures that stimulate the renin-angiotensin system, such as upright posture, sodium restriction, and the administration of angiotensin-converting enzyme inhibitors (ACEI).22 ACEI interrupt the conversion of A-I to A-II; the negative feedback effect of A-II on renin release is diminished, resulting in an increase in secretion. The administration of the ACEI captopril to patients with renovascular hypertension produces a marked increase in PRA compared to responses in other forms of hypertension.20,22,23 and 24 In the usual test procedure, captopril, 25 to 50 mg, is given orally, and PRA is obtained after 1 hour. The test has 75% to 100% sensitivity and 60% to 90% specificity.22,24 Furthermore, it is safe and inexpensive, and can be done on an outpatient basis or simultaneously with captopril renography. The utility of the test is hampered by the need for strict standardization, the need to stop antihypertensive medications, its reduced accuracy in renal failure, and low sensitivity and poor predictive value compared to a renogram. Renal vein renin levels, sampled by percutaneous catheterization, determine the functional significance of the stenotic lesion and predict the BP response to therapy.3 Renin production is increased in the stenotic kidney and suppressed in the contralateral kidney; asymmetry of renin production yields a renal vein renin ratio of ³1.5:1 in renovascular hypertension. This procedure has predictive value for surgical curability, but use has declined because it is invasive, and up to 60% of patients do not lateralize but are nevertheless improved by surgery. 25 Radionuclide Renography. The renogram relies on a disparity of renal function and of perfusion between the stenotic and nonstenotic kidneys. Isotopic compounds such as technetium-99m–diethylenetetraminepentaacetic acid (DTPA) accurately measure glomerular filtration rate, and compounds such as 99m Tc-mercapto acetyl triglycine (99mTc-MAG3) measure both glomerular filtration rate and tubular secretion and are good in renal insufficiency.7 The addition of computer quantitation has improved their accuracy.3,4,5,6 and 7,13,26 Patients with renovascular hypertension show decreased uptake, delayed peak values, and prolonged excretion of these agents. Because of the low-risk and noninvasive nature of renography, it has many advantages. However, renography used alone has a high incidence of false-negative and false-positive results. Captopril administration during renography has improved its sensitivity and specificity.5,6 and 7,22,24,26,27 Captopril (25–50 mg), which is given 1 hour before the isotope injection, reduces A-II–mediated vasoconstriction in the efferent arterioles and lowers glomerular pressure and filtration rate. In the stenotic kidney, captopril reduces renal perfusion more than in the nonstenotic kidney, thereby revealing the asymmetric reduction in renal function in renovascular hypertension. Scintiphotographs and time-activity curves are analyzed to assess renal perfusion, function, and size. In high-risk populations, sensitivity and specificity exceed 90% for high-grade stenosis. The predictive value of the ACE-inhibitor renogram is less in low-risk populations and in bilateral disease. Duplex Doppler Ultrasonography. Ultrasonography has been improved by the use of low-frequency transducers (B-mode imaging) for better visualization of the renal arteries, and Doppler for the measurement of differential blood flow velocities.28,29,30 and 31 It has the advantage of determining both the anatomic and the functional significance of a stenotic lesion. A renal-aortic flow velocity ratio of ³3.5 suggests renovascular hypertension. Duplex Doppler ultrasonography can detect unilateral and bilateral disease, can detect recurrent stenosis in previously treated patients, and can be enhanced with ACE inhibition30 and color coding.31 It has a 99% positive predictive value and a 97% negative predictive value.29 (Disadvantages: time-consuming and highly operator-dependent.) Magnetic Resonance Angiography. Magnetic resonance (MR) angiography is promising and noninvasive, and may become the method of choice.32,33,34,35 and 36 In comparison with arteriography, MR angiography showed 100% sensitivity and 96% specificity for finding stenosis of the main renal arteries.32,33,34,35 and 36 A sensitivity of 100% and a specificity of 71% were found for lesions of the proximal renal arteries with 50% to 75% stenosis.34 Significant advances in techniques include breath-holding MR angiography and paramagnetic contrast material. Thus, the procedure may be able to visualize accessory arteries.34 Phase-contrast MR angiography may enable determination of the hemodynamic significance of a stenotic lesion.35 The procedure cannot be used in individuals with metallic implants such as a pacemaker or a clip for an aneurysm. Spiral (Helical) Computed Tomographic Scan and Computed Tomographic Angiography. This offers great promise as the best noninvasive screening test.37 In a comparison of spiral computed tomography (CT) to arteriography in subjects investigated for renovascular hypertension, the sensitivity was 98% and the specificity 94%.28 However, renal insufficiency reduces the accuracy of this procedure due to diminishing renal blood flow. It identifies lesions of the main renal arteries but has more difficulty in finding branch lesions. Although it is minimally invasive, the risk of radiocontrast-induced nephrotoxicity still exists. Angiography. Despite the increase in types of procedures and technical advances in detecting renovascular hypertension, renal angiography by the percutaneous, transfemoral route remains the “gold standard” for detecting renal artery stenosis, as well as its location and pathology. Atherosclerotic lesions most often involve the proximal segment of the renal artery and have a circular configuration. By contrast, fibromuscular disease is in the more distal portions of the renal artery and is either localized or diffuse and bilateral. Angiography is costly and invasive and has the risk of nephrotoxicity. Because angiography does not always predict the functional significance of renal artery stenosis or its treatment outcome, many centers apply simultaneous angioplasty, with the BP response to percutaneous renal angioplasty or stenting serving as the indicator of functional significance. Digital subtraction angiography avoids some of the complications of percutaneous catheterization, because computer enhancement of the renal area allows the injection of smaller amounts of contrast material. 3 Injection has been in a peripheral vein, but the specificity and sensitivity of the procedure are both 90% or less when compared to arterial procedures. Because 150 to 200 mL of dye is injected, there is still risk for nephrotoxicity. An intraaortic injection site provides better visualization. The procedure can be performed safely in azotemic patients, and it has been useful in hypertension after renal transplantation.4 Thus, the renal arteriogram best establishes the presence of renal artery stenosis but does not distinguish functional from nonfunctional lesions. Tests such as the intravenous pyelogram and radionucleotide renogram lack sensitivity for detecting renovascular hypertension and, if used alone, cannot consistently diagnose renovascular hypertension. The newer tests (e.g., captopril renography, MR angiography, duplex ultra-sonography, and spiral CT) show great promise. TREATMENT OF RENOVASCULAR HYPERTENSION The goals of therapy of renovascular hypertension are the control of BP and the preservation of renal function. The three therapies for treatment of renovascular hypertension are medical, percutaneous transluminal renal angioplasty, and surgery. Each approach has advantages and disadvantages, and the selection also depends on cause, severity, and age. Few trials have compared the three therapies. Often no single test can help in this decision, and experience and clinical judgment are paramount. There are several differences in the approach to unilateral versus bilateral renal artery stenosis. Unilateral Renal Artery Stenosis

SURGERY. Often, surgery is reserved for patients for whom hypertension cannot be controlled on medical therapy or for those who fail angioplasty and in whom renal function is deteriorating. However, some believe that surgery should be considered in patients 2000 ng/dL), with only modest elevation of DHEAS. Nonclassic congenital adrenal hyperplasia may have plasma androgen levels in the adrenarchal range and normal baseline 17-hydroxyprogesterone levels. Adrenal tumors are associated with extremely high levels of DHEAS (>800 µg/dL) and very little elevation in plasma 17-hydroxyprogesterone. The differential diagnosis of extreme elevation of plasma DHEAS also includes the 3b-hydroxysteroid dehydrogenase deficiency form of congenital adrenal hyperplasia and steroid sulfatase deficiency. Gonadal tumors produce relatively large amounts of androstenedione, which causes a modest elevation in excretion of urinary 17-ketosteroids. The dexamethasone suppression test is helpful in distinguishing among the causes of hyperandrogenemia (Fig. 83-4). Dexamethasone, 1 mg/m2 (in three or four divided daily doses), is administered for 4 days (7 days if the patient weighs ³100 kg). This will selectively suppress adrenal androgenic hyperfunction to normal levels in functional adrenal hyperandrogenism, including congenital adrenal hyperplasia. Normally, in prepubertal children, this regimen suppresses plasma androgens to preadrenarchal levels and suppresses cortisol levels to 5 cm) are surgically removed? The prevalence of benign lesions of this size is difficult to determine, but in one series of 12,000 autopsies, four such lesions were found, giving an incidence of 1 in 3000.7 The appropriate control population for this group would be the incidence of adrenal cancer found at autopsy, which is ~1 in 600,000. Thus, if we assume that all cancers are >5 cm, the ratio of benign to malignant lesions in this group is 200:1. If the surgeon is operating to cure cancer, the operation should probably proceed by an anterior abdominal approach (see Chap. 89). This will have a higher mortality, perhaps 2%. In this analysis, four people with a benign lesion will die for every one cured of a cancer. If, on the other hand, the mortality is 5 to 6 cm in largest diameter. The opinion of the author is that the efficacy of “curative” surgery in this setting is equivocal, and that in the absence of pain, absences of torsion and necrosis, and absence of clear-cut signs of malignancy such as invasion of the renal vein, the kidney, or adjacent structures, surgery should be avoided. On the other hand, it is quite clear that surgery prolongs life and improves the quality of life for patients with documented adrenocortical cancer (see Chap. 89). Studies have shown an average longevity of 48 months from diagnosis, and most of that gain can be attributed to surgical intervention; it is the only treatment known to be effective for adrenocortical cancer and should be applied with vigor at this stage of the disease.9 CHAPTER REFERENCES 1. US Department of Health Services National Cancer Institute 1981 Surveillance, Epidemiology, and End Results; Incidence and Mortality Data, 1973–1977. NIH Publication No. 81-2330. Washington, DC: US Government Printing Office, 1981. 2. Terzolo M, Ali A, Osella G, Mazza E. The prevalence of adrenal carcinoma among incidentally discovered adrenal masses. Arch Surg 1997; 132:914. 3. Belldegrun A, Hussain S, Seltzer SE, et al. Incidentally discovered mass of the adrenal gland. Surg Gynecol Obstet 1986; 163:203. 4. Kloos RT. Incidentally discovered adrenal masses. Endocr Rev 1995; 16:460. 5. Cook DM, Loriaux DL. The incidental adrenal mass. Am J Med 1996; 101:88. 6. Bernardino ME, Walther MM, Phillips VM, et al. AJR 1985; 144:67. 7. Shin SJ, Hoda RS, Ying L, DeLellis RA. Diagnostic utility of the monoclonal antibody A103 in fine-needle aspiration biopsies of the adrenal. Am J Clin Pathol 2000; 113:295. 8. Mitchell DG, Crovello M, Matteucci T, et al. Benign adrenocortical masses: diagnosis with chemical shift MR imaging. Radiology. 1992; 185:345. 9. Vassilopoulou-Sellin R, Guinee VF, Klein MJ, et al. Impact of adjuvant mitotane on the clinical course of patients with adrenocortical cancer. Cancer 1993; 71:319.

CHAPTER 85 PHYSIOLOGY OF THE ADRENAL MEDULLA AND THE SYMPATHETIC NERVOUS SYSTEM Principles and Practice of Endocrinology and Metabolism

CHAPTER 85 PHYSIOLOGY OF THE ADRENAL MEDULLA AND THE SYMPATHETIC NERVOUS SYSTEM DAVID S. GOLDSTEIN Integration and Adaptation Embryology Functional Anatomy Catecholamine Synthesis Catecholamine Release Catecholamine Removal Presynaptic Modulation Adrenergic Receptors Effects of Catecholamines Circulation Metabolism Visceral Effects Circulating Catecholamines and Activities of the Sympathoneural and Adrenomedullary Systems Direct Measurement of Sympathetic Neural Activity Other Methods Patterns of Catecholaminergic Activation Chapter References

In responding to stressors, whether physical or psychological, trivial or mortal, the autonomic nervous system plays a crucial role. This system influences cardiovascular, metabolic, and visceral activity in the resting organism and determines the physiologic concomitants of virtually every motion and emotion. According to Langley's conceptualization dating from approximately the turn of the twentieth century, the autonomic nervous system consists of the parasympathetic nervous system and the sympathetic nervous system. The former uses acetylcholine (ACh) as the main neurotransmitter, and the latter (with exceptions) uses the catecholamine norepinephrine (NE, noradrenaline). The adrenomedullary hormonal system uses the catecholamine epinephrine (EPI, adrenaline) as the main hormone secreted into the bloodstream. Walter B. Cannon considered the adrenal gland to act in concert with the sympathetic nervous system to maintain homeostasis during emergencies, leading to the current use of the term sympathoadrenal system. As discussed in this chapter, however, recent evidence has indicated separate regulation of sympathoneural and adrenomedullary outflows during different forms of stress.1 In addition, a third peripheral catecholaminergic system may use dopamine (DA) as an autocrine-paracrine substance.

INTEGRATION AND ADAPTATION An understanding of sympathetic nervous and adrenomedullary hormonal system activation may be facilitated by considering integrative actions in homeostasis and in adaptive responses during stress.1 For instance, almost all the components of the shock syndrome—pallor, sweating, cutaneous vasoconstriction, piloerection, agitation, pupillary dilation, tachycardia, hyperventilation, increased total peripheral resistance, decreased gastrointestinal motility, hyperglycemia, and renal sodium retention—can be accounted for by increased activity of catecholaminergic systems in the ultimate attempt to preserve life. When these compensatory mechanisms are overwhelmed or give way, the organism dies. Many principles of endocrinology and metabolism dealing with hormones, central and peripheral neurotransmission, receptors, and second messengers originated from classic investigations of the function of catecholamines. In addition, many effective therapies for clinical disorders, such as depression, myocardial ischemia, hypertension, and circulatory shock, have resulted directly from applications of these principles.

EMBRYOLOGY The chromaffin tissue of the adrenal medulla and that of the ganglia of the sympathetic nervous system share an embryonic origin from neural crest tissue. The term chromaffin refers to the histochemical identification of intracellular granules that turn brown when treated with oxidizing agents. In adult humans, chromaffin cells exist mainly in the adrenal medulla because sympathetic nerve terminals do not occur in a form concentrated enough to produce a detectable chromaffin reaction. Precursors of sympathetic cells (sympathogonia) develop early in ontogeny in the area of the neural crest and tube. They migrate ventrally and differentiate into paravertebral and preaortic neuroblasts, the precursors of ganglion cells, and pheochromoblasts, the precursors of chromaffin cells. Thus, extraadrenal pheochromocytomas characteristically occur in paravertebral or preaortic regions. Whereas adrenocortical tissue arises from coelomic epithelium, adrenomedullary tissue arises from pheochromoblasts that migrate from sympathetic ganglia and nest within adrenocortical tissue. The adrenomedullary cells differentiate into modified neuronal (gland) cells, not neurons. The separate origins of adrenocortical and adrenomedullary tissue are evident in amphibians and reptiles, in which the two types of glandular tissue are only loosely associated, and in fish, which have anatomically distinct glands. In humans, extraadrenal chromaffin tissue degenerates after birth, and adrenomedullary chromaffin tissue matures.

FUNCTIONAL ANATOMY Sympathetic preganglionic axons originate from cell bodies in the intermediolateral cell columns of the thoracolumbar spinal cord2 (Fig. 85-1). A central neural site projecting to the intermediolateral cell columns has been identified in the rostral ventrolateral medulla,3 and catecholaminergic neurons in this region appear to participate in regulation of blood pressure by more rostral structures and by baroreflexes; however, whether the rostral ventrolateral medulla cells that project to the sympathetic preganglionic neurons actually use catecholamines as the responsible transmitters now seems unlikely. After exiting the spinal cord, the preganglionic sympathetic axons synapse in the paravertebral chain of sympathetic ganglia and preaortic ganglia. Preganglionic fibers also pass through lower thoracic and lumbar ganglia to form the splanchnic neural innervation of the adrenal medulla. Adrenal nerve activity, therefore, at least partly reflects preganglionic outflow, whereas renal nerve sympathetic activity reflects postganglionic outflow. At the ganglionic synapse and at adrenomedullary cells, the neurotransmitter is ACh.

FIGURE 85-1. Diagram of efferent autonomic pathways. The heavy lines indicate parasympathetic pathways, and the thin lines represent sympathetic pathways. The solid lines represent preganglionic pathways, and the dashed lines postganglionic pathways. Note the emanation of sympathetic fibers predominantly from the thoracolumbar spinal cord, whereas the parasympathetic fibers originate predominantly from the brainstem and the sacral spinal cord. (From Youmans WB. Fundamentals of human physiology, 2nd ed. Chicago: Year Book Medical Publishers, 1962.)

The postganglionic neurons end in nerve terminals throughout the body, especially in lattice-like networks in the adventitial and adventitial-medial layers of arterioles, where the terminals release NE into the synaptic clefts. Skeletal muscle beds, however, appear to possess both sympathetic noradrenergic and cholinergic innervation,

and eccrine sweat glands, involved in thermoregulation, possess sympathetic cholinergic innervation. The adrenal medulla secretes directly into the bloodstream. The cytoplasm of adrenomedullary cells contains large amounts of phenylethanolamine N-methyltransferase (PNMT), which catalyzes the conversion of NE to EPI. EPI constitutes ~85% of the catecholamine content of the adrenal medulla in humans and is the main catecholamine secreted by the gland. Sympathetic nerve endings are not selectively coupled with single effector cells. Instead, they form plexuses, in which one axon innervates several effector cells and each effector cell receives innervation from several axons. Fibers from the superior cervical ganglion innervate the head. Cardiac sympathetic innervation emanates from the three cervical sympathetic ganglia (the most caudal being the stellate ganglia) and from the upper thoracic ganglia, converging in the cardiac nerves. Gut organs receive innervation from preaortic plexuses, whereas the pelvic organs derive their innervation from the sacral spinal nerves and pelvic plexuses. NE is not distributed uniformly in sympathetic nerve endings, but is localized in discrete areas called varicosities, ~25,000 of which may occur on a single axon.4 Each varicosity contains ~1000 membrane-bound vesicles.5 The electron-dense core found in a proportion of the vesicles is thought to represent ~15,000 molecules of stored NE. The vesicles also contain DA b-hydroxylase (DBH), adenosine triphosphate, and various peptides, including enkephalin, neuropeptide Y, and chromogranin A. The blood supply to the adrenal medulla is derived from the inferior phrenic artery, the aorta, and the renal artery, and from a corticomedullary portal system originating in the zona reticularis. The last supply provides the anatomic substrate for the influence of adrenocortical steroids on catecholamine biosynthesis. The parasympathetic nervous system, the other limb of the autonomic nervous system, is derived mainly from cranial (vagal) and sacral cholinergic efferents. The ganglionic synapses usually are near or embedded in the parenchyma of the innervated organ.

CATECHOLAMINE SYNTHESIS The catecholamines contain a catechol (3,4-dihydroxyphenyl) nucleus (Fig. 85-2). Intracellular events in NE synthesis, turnover, and metabolism are depicted in Figure 85-3. The precursor of the catecholamines—the amino acid tyrosine—is a dietary constituent as well as a product of hepatic metabolism of dietary phenylalanine. Tyrosine is taken up by sympathetic neural tissue and is converted to dihydroxyphenylalanine (DOPA) by the enzyme tyrosine hydroxylase (TH).6,7 The specific enzymatic hydroxylation of tyrosine to DOPA is the rate-limiting step in catecholamine biosynthesis and is end product–inhibited by DOPA, DA, and NE. Although it is known that adrenomedullary chromaffin cells and tissues that possess sympathetic innervation contain TH, it is unknown whether, outside the brain, only sympathetic nerves and adrenomedullary cells contain TH. Current evidence suggests a nonneuronal synthesis of catecholamines.8

FIGURE 85-2. Structures of the endogenous catecholamines and major metabolites are illustrated. Substances in the main pathway of catecholamine biosynthesis appear in boldface. (MAO, monoamine oxidase; COMT, catechol-O-methyltransferase.)

FIGURE 85-3. Mechanisms of norepinephrine (NE) synthesis, release, and metabolism. Tyrosine (TYR) is taken up into sympathetic nerves and converted to dihydroxyphenylalanine (DOPA) by tyrosine hydroxylase (TH). A small proportion of axoplasmic DOPA enters the bloodstream. DOPA is converted to dopamine (DA) by L-aromatic amino-acid decarboxylase (LAAAD), and DA is taken up into vesicles containing dopamine b -hydroxylase (DBH) and converted to NE, or is deaminated by monoamine oxidase (MAO) to form dihydroxyphenylacetic acid (DOPAC), the main neuronal metabolite of DA. Vesicular NE can leak into the axoplasm, undergoing deamination to form dihydroxy-phenylglycol (DHPG), or it can be released by exocytosis. Most of the endogenously released NE is taken back up into the axoplasm by uptake-1 A small proportion of released NE enters the bloodstream or is taken up by extraneuronal uptake (uptake-2). NE in extraneuronal cells undergoes O-methylation by catechol-O-methyltransferase (COMT) to form normetanephrine (NMN), which enters the bloodstream or is deaminated to form methoxyhydroxyphenylglycol (MHPG) or (in the liver) to vanillylmandelic acid. DHPG and DOPAC in extraneuronal cells also undergo O-methylation to form MHPG and homovanillic acid (HVA), the latter being the main end product of DA metabolism.

DOPA is converted to DA by DOPA decarboxylase (aromatic L-amino acid decarboxylase), which is found in many tissues. DA is taken up into vesicles containing DBH and converted to NE. DBH occurs only in the vesicles of NE- or EPI-synthesizing cells. In adrenomedullary tissue, NE is methylated to form EPI by the cytosolic, nonspecific enzyme PNMT.9 Glucocorticoids, which occur at high concentrations in the adrenal medulla because of the corticomedullary portal system, induce PNMT activity.10 The catecholamine content of an organ changes little during sympathetic stimulation because of the small percent of the storage pool released, concurrent stimulation of catecholamine synthesis by induction of TH activity, and efficient recycling of the released NE. TH activity increases in response to loss of feedback inhibition during sympathetically mediated release of neurotransmitter and in response to cellular depolarization.

CATECHOLAMINE RELEASE ACh depolarizes chromaffin cells by increasing membrane permeability to sodium. Transmembrane influx of calcium directly or indirectly results from this increase in intracellular sodium; thus, cytoplasmic calcium levels increase. The latter process is thought to be part of an incompletely defined biomechanical final common pathway in cellular activation leading to exocytotic release of vesicular contents. Indirect evidence suggests that neurosecretion occurs by actual exocytosis of the contents of the storage vesicles. Other soluble vesicular contents—adenosine triphosphate, enkephalins, chromogranins, neuropeptide Y, and DBH—are released during cellular activation, but cytoplasmic macromolecules are not.11,12 and 13

Electron micrographs occasionally have demonstrated the so-called omega sign, as if a defect in the cell membrane were produced at the site of vesicle-plasmalemma fusion. Cytoplasmic NE does not appear to be released by nerve impulses. Sympathomimetic amines release NE by a nonexocytotic process because DBH also is not released and because the release is not calcium-dependent.

CATECHOLAMINE REMOVAL The most prominent removal mechanism for released NE is reuptake into the presynaptic terminal by a nonstereospecific, energy-requiring process called uptake-1.14,15 Cocaine, tricyclic antidepressants (e.g., desipramine hydrochloride), and ouabain inhibit uptake-1. The hypotensive effect of guanethidine and the hypertensive effect of tyramine depend on uptake of these drugs into presynaptic axons. Therefore, the administration of uptake-1 inhibitors attenuates these effects. Molecular genetic studies have confirmed the existence of a family of neurotransmitter transporters, including distinct transporters for DA and NE. Uptake-1 refers to the transporter for NE in sympathetic nerves. NE in the axonal cytoplasm can be taken up into vesicles by a stereoselective process, or it can be oxidized by mitochondrial monamine oxidase. Monoamine oxidase is present in both neural and nonneural tissue, and the deaminated metabolites are rapidly converted to dihydroxyphenylglycol (DHPG)16 or dihydroxymandelic acid in the cell. Vesicular uptake is not specific for NE, and a-methyldopamine can be taken up into the vesicle, converted to a-methylnorepinephrine, and released during sympathetic stimulation. NE also can be removed by extraneuronal uptake (uptake-2), a nonstereoselective, energy-requiring process that is inhibited by metanephrine and steroids. Extraneuronal tissues contain catechol-O-methyltransferase (COMT), which converts NE to normetanephrine and EPI to metanephrine. COMT exists primarily in extraneuronal tissues, especially the liver and kidney. Adrenomedullary cells also contain COMT, and O-methylation of catecholamines in the adrenal gland constitutes the main determinant of circulating concentrations of metanephrines, both in healthy individuals and in patients harboring catecholaminesecreting adrenal tumors.17 Products of combined actions of monoamine oxidase and COMT include vanillylmandelic acid and methoxyhydroxyphenylglycol, the former produced mainly in the liver and the latter produced in sympathetically innervated organs.

PRESYNAPTIC MODULATION NE can regulate its own release, by stimulating presynaptic, inhibitory a2-adrenergic receptors.18,19 Other inhibitors of NE release that appear to act presynaptically include ACh,20 DA, prostaglandins of the E series,21 histamine, and purines. Stimulatory modulators include angiotensin II22 and EPI, the latter of which appears to stimulate presynaptic b2-adrenergic receptors. Substantial clinical evidence supports modulation of NE release by a2-adrenoceptors.23 The roles of the other endogenous compounds remain less well established.

ADRENERGIC RECEPTORS Specific receptors for catecholamines exist in several tissues, including muscle (skeletal, myocardial, vascular, gastrointestinal, ciliary, and bronchial) and glands (salivary, sweat, adrenocortical, and pancreatic).24 These receptors mediate the effects of endogenously released as well as exogenously administered catecholamines (Table 85-1).

TABLE 85-1. Responses of Effector Organs to Autonomic Nerve Impulses

Ahlquist25 introduced the concept of a- and b-adrenergic receptors to explain contrasting cardiovascular responses to exogenous catecholamines. NE administration causes diffuse vasoconstriction and reflexive bradycardia, whereas EPI causes vasodilation of skeletal muscle beds, tachycardia, and increased myocardial contractility. According to this framework, the vasoconstrictor actions of NE and EPI result from stimulation of a-adrenergic receptors, whereas the cardiotonic, tachycardiac, and vasodilator properties of EPI result from stimulation of b-adrenergic receptors. Further subclassification to a1-, a2-, b1-, and b2-adrenergic receptors has been based on ligand-binding studies, on responses to synthetic agonists and antagonists, and on molecular genetic experiments.26,27 and 28 NE is an agonist at b1-, a1-, and a2-adrenergic receptors, whereas EPI is an agonist at all four subtypes. Stimulation of a1-receptors causes vasoconstriction. Phenylephrine is an example of a relatively specific a1-agonist, and prazosin is an example of an a1-antagonist. a2-Selective agonists, such as clonidine hydrochloride,29 a-methyldopa, and guanabenz,30 exert more complex effects, because stimulation of a2-receptors on vascular smooth muscle cells elicits vasoconstriction and increased blood pressure, whereas stimulation of presynaptic a2-receptors inhibits NE release from sympathetic nerve endings, and stimulation of a2-receptors in the central nervous system causes decreases in sympathetic outflow and increases in vagal outflow. The latter effects predominate, and a2-agonists are used as antihypertensive medications. a2-Antagonists (e.g., yohimbine) cause increases in plasma NE, as well as a psychiatric pattern of heightened arousal and tremulousness, in contrast with the sedation induced by a2-agonists. Pharmacologic and molecular genetic studies have demonstrated the existence of several subtypes of a2-adrenoceptors. The structure of the subtype responsible for modulation of NE release from sympathetic nerves remains unknown. b1-Agonists increase heart rate, cardiac conduction velocity, and myocardial contractility, and produce lipolysis. In contrast, b2-agonists (e.g., terbutaline sulfate and albuterol) predominantly cause skeletal muscle vasodilation, bronchial and gastrointestinal smooth muscle relaxation, increased renin secretion from the kidney, glycogenolysis, and enhanced release of NE from sympathetic nerve terminals. Because of the latter effect, b2-agonists actually may elicit delayed constriction in some vascular beds. Relatively selective b1-antagonists are being used with increasing frequency for cardiovascular disorders, such as angina pectoris, arrhythmias, and hypertension, because nonselective b-antagonists can aggravate bronchospastic responsiveness. Examples of b1-adrenergic receptor-selective medications are atenolol and metoprolol. A third type of b-adrenoceptor has been described, b3, which probably participates in catecholamine-induced lipolysis. The intracellular mechanisms activated by a-adrenergic receptor stimulation appear to be subtype-selective. Stimulation of a1-receptors releases ionized calcium into the cytoplasm from intracellular stores, by activation of the inositol triphosphate pathway31 (see Chap. 4). Stimulation of a2-receptors appears to inhibit the adenylate cyclase–cyclic adenosine monophosphate (cAMP) system. Both b1- and b2-adrenergicreceptor agonists stimulate the adenylate cyclase–cAMP system. The exact mechanism of inhibition of NE release by a2-adrenoceptor stimulation remains unknown. The calculated number of adrenoceptors on cell surfaces can increase (up-regulate) or decrease (down-regulate) in responseto changes in agonist concentrations at aor b-adrenergic receptors. Altered numbers of adrenergic receptors, altered affinity for catecholamines, or altered intracellular actions distal to receptor occupation provide explanations for tachyphylaxis (e.g., decreasing responsiveness to b-agonists after repeated administration to patients with asthma) and sensitization (e.g., increasing pressor responsiveness to administered NE in patients with idiopathic orthostatic hypotension).32,33,34 and 35 Thyroid hormones and glucocorticoids can potentiate b-adrenergic receptor–mediated responses.36 It has been proposed that decreased b-adrenergic receptor numbers in congestive heart failure may result from chronically increased release of endogenous NE.35 Although alterations in receptor numbers or in intracellular biomechanical events can affect responses to agonists, the main determinant of physiologic responses to sympathoadrenal activation is the amount of endogenous catecholamines released from the sympathetic nerve endings and the adrenal medulla, which reach

adrenoceptors on effector cells.

EFFECTS OF CATECHOLAMINES CIRCULATION Cardiovascular function, both at rest and during stress, unquestionably depends on sympathetic neural activity. Interference with sympathetic outflow or inhibition of NE release from sympathetic nerve endings causes the blood pressure to decrease even in resting, healthy adults, and b-adrenergic receptor blockade causes a decrease in heart rate at rest, as well as attenuated heart rate responses to stress. Circulatory responses during orthostasis, altered regional and total body metabolic demands imposed by exercise, changes in environmental temperature, hemorrhage, hypoglycemia, hypoxia, and emotion occur predominantly by patterned sympathoneural and adrenomedullary activation. Sympathetic stimulation causes tachycardia, increased cardiac inotropism, increasingly rapid cardiac conduction, acceleration of myocyte spontaneous depolarization, increased resistance to blood flow in most vascular beds, and venousconstriction. Because of relatively scanty sympathetic innervation of the coronary and cerebral vessels, diffuse sympathetic stimulation generally preserves blood flow to these vital organs. Total peripheral resistance may remain unchanged or may decrease, depending on the extent of EPI-induced skeletal muscle vasodilation. More indirect or longer-term effects of sympathoadrenal activation on circulatory function include (a) antidiuresis and stimulation of renal sodium retention by direct tubular effects, as well as by enhanced reninangiotensin-aldosterone system activity, reduced intrarenal hydrostatic pressure, and increased vasopressin secretion, possibly shunting corticomedullary blood flow37,38,39,40,41,42,43,44,45 and 46; (b) stimulation of cell growth and division in cardiac and vascular smooth muscle tissue42,43; (c) a2-adrenergic receptor–mediated promotion of platelet aggregation; (d) shunting of blood toward the cardiopulmonary region, which causes stimulation of low-pressure cardiac baroreceptors; and (e) adrenoceptor down-regulation. METABOLISM Metabolic effects of pharmacologic doses of EPI include non-shivering thermogenesis, hyperglycemia, hyperlipidemia, hypokalemia, and increased oxygen consumption.47,48,49,50,51,52 and 53 Most of these actions arise from b-receptor–mediated stimulation of adenylate cyclase in cell membranes, which catalyzes the conversion of adenosine triphosphate to cAMP in the cell. Mechanisms for the hyperglycemic response to EPI include stimulation of hepatic glycogen phosphorylase, the rate-limiting catalytic step in glycogenolysis; inhibition of glycogen synthase; stimulation of gluconeogenesis from amino acids and lactate; a-adrenergic receptor–mediated inhibition of insulin secretion; and stimulation of pancreatic secretion of glucagon. VISCERAL EFFECTS Stimulation of a1-adrenergic receptor causes smooth muscle contraction, and b2-receptor stimulation of smooth muscle causes relaxation. Catecholamines also may produce indirect visceral effects by modifying ACh release at parasympathetic nerve terminals (see Table 85-1). Catecholamines diffusely inhibit gut motility. They stimulate myoepithelial cells in the breast and ovary, affecting lactation and ovulation. Because DA inhibits prolactin secretion, the DA receptor agonist, bromocriptine, had been used to stop lactation in women after childbirth, but this indication was rescinded by the U.S. Food and Drug Administration. Axillary sweating is stimulated by sympathetic noradrenergic innervation of apocrine glands. Eccrine glands, which participate importantly in thermoregulation, receive sympathetic cholinergic innervation. In the eye, catecholamines stimulate secretion into the aqueous humor, and b-blockade is an established mode of therapy for some forms of glaucoma.54 Stimulation of the sympathetic innervation of the iris from the superior cervical ganglia produces mydriasis, whereas interruption of this innervation produces Horner syndrome, characterized by ptosis, myosis, and facial anhidrosis. Catecholamines affect the function of virtually all endocrine systems.54 Generally, b-adrenergic receptor stimulation accentuates release of preformed hormone, whereas a-adrenergic receptor agonism attenuates the action of the physiologic stimulus for secretion of the hormone. In the kidneys, NE and EPI promote antinatriuresis by decreasing renal blood flow and increasing secretion of renin, whereas DA promotes natriuresis by increasing renal blood flow and inhibiting Na+/K+ adenosine triphosphatase in proximal tubular cells.

CIRCULATING CATECHOLAMINES AND ACTIVITIES OF THE SYMPATHONEURAL AND ADRENOMEDULLARY SYSTEMS Much of the catecholamine content of human plasma is sulfo-conjugated, especially DA, which is ~98% conjugated in peripheral blood. Most of DA sulfate production in humans occurs in the gastrointestinal tract.55 Some of the free (i.e., non-conjugated) catecholamines circulate bound loosely to protein. The significance of this protein binding also is unclear. NE and DA are not hemocrine hormones. That is, the cardiovascular and metabolic effects of these catecholamines require supraphysiologic plasma concentrations. In contrast, even within a physiologic range, circulating EPI can exert circulatory and metabolic effects. Therefore, EPI may be considered to be a hemocrine hormone. Free and conjugated catecholamines and catecholamine metabolites are excreted in the urine. The renal mechanisms involved are poorly understood. By far the predominant free catecholamine in human urine is DA, and the main catechol is dihydroxyphenylacetic acid.56 The prominence of DA and its metabolites in human urine, in contrast with NE, EPI, and their metabolites, remains unexplained. The measurement of total urinary catechol or catecholamine excretion, therefore, is not as useful a screening test for pheochromocytoma as is the measurement of the excretion of metabolites of NE and EPI. The urinary excretion of DA in humans exceeds its renal clearance; a proportion of urinary DA must come from DA released by the kidney itself. In humans, most or all of urinary DA is derived from plasma levodopa (L -dopa).56 In the antecubital venous blood of resting, supine, healthy persons, plasma levels of NE average approximately 100 to 400 pg/mL, and those of EPI average ~5 to 50 pg/mL. Plasma levels of EPI can be less than 5 pg/mL, whereas plasma NE is detectable even in patients with sympathetic failure or those who are undergoing ganglion blockade. Plasma NE increases as a function of age among healthy persons,57 but plasma EPI does not. Catecholamine concentrations in antecubital venous blood are commonly used to indicate overall sympathetic “tone.”58,59 The validity of this application requires consideration of a few important principles. First, owing to the several removal mechanisms for NE, a large cleft-to-plasma concentration gradient exists, determined primarily by uptake-1.60 Medications or pathologic states associated with inhibition of uptake-1 or overall NE removal may result in overestimation of sympathetic outflow. Second, the vasculature of the arm possesses extensive sympathetic innervation and contributes importantly to NE in the venous drainage. Third, any of several modulators can act presynaptically to enhance or attenuate NE release for a given amount of sympathetic nerve traffic. Last, sympathoneural responses are not always homogeneous among the several vascular beds, so that measuring antecubital venous plasma NE may not detect regional abnormalities in sympathetic outflow. An effective blood–brain barrier exists for catecholamines. Because adrenalectomized patients have normal plasma NE levels, plasma NE in humans is thought to be derived mainly from peripheral sympathetic nerve endings. The complex origin of plasma NE contrasts with the simple, single origin of plasma EPI from the adrenal medulla. The view that the level of plasma EPI provides the best measure of overall sympathetic outflow ignores differential changes in plasma NE and EPI during different forms of stress. The level of plasma EPI does provide a fairly direct assessment of adrenomedullary secretion. The amount of regional release of NE into the venous drainage (“spillover”) can be measured from the regional removal of NE, the arteriovenous increment in NE across the bed, and regional blood flow. The regional removal of NE can be assessed by determining the concentrations of radioactive and total NE in arterial and regional venous blood during the infusion of tracer-labeled NE.61,62 and 63 Combined measurements of catecholamines and their metabolites provide information about rates of uptake, turnover, and metabolism of NE. Plasma DHPG reflects

intraneuronal metabolism of axoplasmic NE and, thereby, the turnover of vesicular NE, and plasma DOPA and dihydroxyphenylacetic acid partly reflect catecholamine biosynthesis. Because catecholamine synthesis normally balances turnover in sympathetic nerves, plasma DHPG levels correlate positively with plasma DOPA levels in humans.64 Many factors that are both common and difficult to control influence plasma levels of catecholamines. These include psychological stress, altered environmental temperature, posture (plasma NE doubles within 5 minutes of standing), exercise, hospitalization, medications, and sodium intake (salt restriction increases plasma NE65). Thus, the conditions of blood sampling must be monitored carefully. One clinical method for sampling blood for catecholamine determinations includes testing at a consistent time of day at least 12 hours after the subject has discontinued smoking, alcohol ingestion, and ingestion of caffeine-containing food-stuffs, and after several days of a diet containing a normal amount of salt, as indicated by 24-hour urinary sodium excretion. Because even decaffeinated coffee contains caffeic acid, a catechol converted to dihydrocaffeic acid by gut bacteria, and because dihydrocaffeic acid can cause spuriously high levels of EPI or DA, even decaffeinated coffee should not be ingested for ~12 hours before testing.66 Acetaminophen interferes with liquid chromatographic-electrochemical assays for normeta-nephrine.67 For testing, the subject lies supine with an indwelling intravenous catheter or needle in place for at least 20 minutes before phlebotomy is performed without a tourniquet. The blood is drawn into a chilled, evacuated, heparinized glass tube; additives are unnecessary. The plasma is stored at –70°C or colder until the time of assay. Because virtually any medication can affect plasma catecholamine levels, it is advisable, if possible, to discontinue the use of all medications (particularly antihypertensive medications) for at least 2 weeks before testing. Plasma catecholamines can be measured accurately by radioenzymatic assay68 or liquid chromatography with electro-chemical detection (LCED).69 LCED has largely supplanted older radioenzymatic assay procedures, because it is less expensive, does not involve radionuclides, and can be used for simultaneous measurements of the catechols DOPA, dihydroxy-phenylacetic acid, NE, EPI, DA, DHPG (the main intraneuronal metabolite of NE), and synthetic catecholamines used clinically as pressors.

DIRECT MEASUREMENT OF SYMPATHETIC NEURAL ACTIVITY Microneurography can measure sympathetic neural activity from the rate of pulse-synchronous bursts of directly recorded activity in a peripheral nerve, such as the peroneal nerve, supplying skeletal muscle.70,71 This approach avoids many interpretational difficulties associated with catecholamine measurements, but the procedure requires extensive technical training and cannot measure sympathetic outflow in visceral vascular beds. When skeletal muscle sympathetic activity has been compared with antecubital venous plasma NE, the agreement between these values has been surprisingly strong.72,73 The agreement probably reflects concurrent changes in sympathetic outflows to several beds. OTHER METHODS Power spectra of heart rate variability include a low-frequency component thought to reflect cardiac sympathetic tone. Parasympathetic activity also influences this component, however, and no studies have demonstrated the validity of analyses of heart rate variability in specifically indicating cardiac sympathetic activity in humans. Nuclear scanning after the injection of radioactive sympathomimetic amines or catecholamines can visualize cardiac sympathetic innervation and may enable assessment of cardiac sympathetic function.74,75

PATTERNS OF CATECHOLAMINERGIC ACTIVATION Sympathoneural and adrenomedullary activation in response to physiologic or pharmacologic manipulations can be generalized or patterned. Consideration of the appropriateness of adjustments to a particular challenge can help to explain the observed patterns of responses. Sympathetic neural outflow to most vascular beds increases in response to orthostasis, isometric exercise, exposure to cold, and extracellular volume depletion; thus, plasma NE increases.76 Pharmacologic agents such as amphetamine, tyramine, clonidine, and isoproterenol also affect plasma NE levels preferentially. The pharmacologic findings imply that mechanisms of modulation of catecholamine release may differ between sympathetic nerve endings and adrenomedullary cells. The physiologic stimuli have in common a relatively minor influence on compensatory mechanisms for fuel production or utilization and seem mainly to elicit sympathetic neural responses leading to appropriate shifts in blood flow distribution or glandular secretion.Stimulation of arterial “high-pressure” baroreceptors appears to inhibit sympathetic outflow diffusely, although with quantitative differences among regional beds.77 Adrenomedullary activity increases markedly in response to hypoglycemia, hemorrhage, asphyxiation, circulatory collapse, and emotional distress; therefore, plasma EPI increases to a greater extent than does plasma NE in these situations.77,78 and 79 Because of the potential threat to the organism's existence, even small decreases in levels of glucose trigger circulatory, metabolic, and visceral responses to maximize delivery of this vital fuel; adrenomedullary activation contributes importantly to these responses. Many stimuli, when intense, elicit combined adrenomedullary and sympathetic neural discharge. The resulting syndrome— tachycardia, vasoconstriction, pallor, sweating, pupillary dilation, hyperglycemia, renal retention of sodium, ileus, hyperventilation, piloerection (“gooseflesh”), and agitation—defines the appearance of a patient in shock and has awed clinicians since ancient times. The pattern of catecholaminergic responses to psychological stress in humans appears to be analogous to that associated with the defense reaction in animals preparing for fight-or-flight responses. Renal, gut, and cutaneous vasoconstriction occur, along with skeletal muscle vasodilation, tachycardia, and predominantly systolic hypertension. One theory about the pathogenesis of essential hypertension states that susceptible persons undergo many defense reactions daily in response to conditional stimuli, without the actual physical fight-or-flight behavior occurring. These chronically repeated pressor episodes eventually lead to more long-lasting structural adaptation of the arterioles. Because, in these situations, skeletal muscle sympathetic outflow increases little, or decreases, and because the arm contributes importantly to NE levels in its venous drainage, plasma NE levels in antecubital venous blood may change little.80 Vasodepressor responses involve skeletal vasodilation; cutaneous, renal, and gut vasoconstriction; EPI release; and vagally mediated bradycardia. Directly recorded skeletal sympathetic activity decreases markedly during these responses, and plasma NE fails to increase.81,82 Patterning of catecholaminergic responses to different stressors argues against the doctrine of nonspecificity,83 a founding principle of the stress theory of Hans Selye, who introduced and popularized the notion of stress as a medical scientific idea. Alternatively, stress responses may contain a degree of primitive specificity, reflecting the evolution of adaptively advantageous patterns of responses involving several effector systems, including catecholaminergic systems. Abel first deduced the chemical structure of a hormone; EPI. Elliott's hypothesis about the neuronal release of an “EPI-like substance” foreshadowed the theory of chemical neurotransmission. The discovery of cAMP, the first identified intracellular messenger, and of cellular activation by phosphorylation, depended on a knowledge of the hormonal effects of EPI. Ahlquist's explanation for the different effects of catecholamines, and Iversen's and Axelrod's findings about the disposition of catecholamines, led to the identification of the molecular structures of adrenoceptors and catecholamine transporters and to the discovery of pharmacologic agents now widely used in cardiology, neurology, and psychiatry. Thus, for a century, the study of catecholaminergic systems has proved to be a most fruitful endeavor. If the overall goal of physiologic research is to identify the “algorithms” that the body uses to maintain the internal environment, medical history predicts that catecholamine research will spearhead this quest. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

Goldstein DS. Stress, catecholamines, and cardiovascular disease. New York: Oxford University Press, 1995. Gabella G. Structure of the autonomic nervous system. London: Chapman & Hall, 1976. Reis DJ, Granata AR, Joh TH, et al. Brain stem catecholamine mechanisms in tonic and reflex control of blood pressure. Hypertension 1984; 6:117. Dahlström A, Haggendal J. Some quantitative studies on the noradrenaline content in the cell bodies and terminals of a sympathetic adrenergic neuron system. Acta Physiol Scand 1966; 67:271. Geffen LB, Livett BG. Synaptic vesicles in sympathetic neurons. Physiol Rev 1971; 9:119. Blaschko H. Catecholamine biosynthesis. Br Med Bull 1973; 29:105. Udenfriend S. Tyrosine hydroxylase. Pharmacol Rev 1966; 18:43. Kawamura M, Schwartz JP, Nomura T, et al. Differential effects of chemical sympathectomy on expression and activity of tyrosine hydroxylase and levels of catecholamines and DOPA in peripheral tissues of rats. Neurochem Res 1999; 24:25.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83.

Axelrod J. Methylation reactions in the formation and metabolism of catecholamines and other biogenic amines. Pharmacol Rev 1966; 18:95. Wurtman RJ, Axelrod J. Control of enzymatic synthesis of adrenaline in the adrenal medulla by adrenal cortical steroids. J Biol Chem 1966; 241:2301. Slotkin TA, Kirshner N. All-or-none secretion of adrenal medullary storage vesicle contents in the rat. Biochem Pharmacol 1973; 22:205. Smith AD. Mechanisms involved in the release of noradrenaline from sympathetic nerves. Br Med Bull 1973; 29:123. Weinshilboum RM, Thoa NB, Johnson DG, et al. Proportional release of norepinephrine and dopamine-beta-hydroxylase from sympathetic nerves. Science 1971; 174:1349. Iverson LL. Catecholamine uptake process. Br Med Bull 1973; 29:130. Iverson LL. The uptake and storage of noradrenaline in sympathetic nerves. Cambridge: Cambridge University Press, 1967. Eisenhofer G, Keiser HR, Friberg P, et al. Plasma metanephrines are markers of pheochromocytoma produced by catechol-O-methyltransferase within tumors. J Clin Endocrinol Metab 1998; 83:2175. Goldstein DS, Eisenhofer G, Stull R, et al. Plasma dihydroxyphenylglycol and the intraneuronal disposition of norepinephrine in humans. J Clin Invest 1988; 81:213. Goldstein DS, Golczynska A, Stuhlmuller J, et al. A test of the “epinephrine hypothesis” in humans. Hypertension 1999; 33:36. Langer SZ. Presynaptic regulation of catecholamine release. Biochem Pharmacol 1974; 23:1793. Levy MN, Blattberg B. Effect of vagal stimulation on the overflow of norepinephrine into the coronary sinus during cardiac sympathetic nerve stimulation in the dog. Circ Res 1976; 38:81. Hedqvist P. Modulating effect of prostaglandin E-2 on noradrenaline release from the isolated cat spleen. Acta Physiol Scand 1969; 75:511. Zimmerman BG. Adrenergic facilitation by angiotensin: does it serve a physiological function? Clin Sci 1976; 39:566. Grossman E, Chang PC, Hoffman A, et al. Evidence for functional a2-adrenoceptors on vascular sympathetic nerve endings in the human forearm. Circ Res 1991; 69:887. O'Dowd BF, Lefkowitz RJ, Caron MG. Structure of the adrenergic and related receptors. Annu Rev Neurosci 1989; 12:67. Ahlquist RP. A study of the adrenotropic receptors. Am J Physiol 1948; 153:586. Lefkowitz RJ. Direct binding studies of adrenergic receptors: biochemical, physiologic and clinical implications. Ann Intern Med 1979; 91:450. Hoffman BB, Lefkowitz RJ. Alpha-adrenergic receptor subtypes. N Engl J Med 1980; 302:1390. Motulsky HJ, Insel PA. Adrenergic receptors in man. Direct identification, physiologic regulation, and clinical alterations. N Engl J Med 1982; 307:18. Jarrott B, Conway EL, Maccarrone C, Lewis SJ. Clonidine: understanding its disposition, sites and mechanism of action. Clin Exp Pharmacol Physiol 1987; 14:471. Gehr M, MacCarthy EP, Goldberg M. Guanabenz: a centrally acting natriuretic antihypertensive drug. Kidney Int 1986; 29:1203. Minneman KP. a1-Adrenergic receptor subtypes, inositol phosphates, and sources of cell Ca 2+. Pharmacol Rev 1988; 40:87. Hollister AS, FitzGerald GA, Nadeau JHJ. Acute reduction in human platelet alpha-2-adrenoceptor affinity for agonist by endogenous and exogenous catecholamines. J Clin Invest 1983; 72:1498. Fraser J, Nadeau J, Robertson D, Wood AJ. Regulation of human leukocyte beta receptors by endogenous catecholamines: relationship of leukocyte beta receptor density to the cardiac sensitivity to isoproterenol. J Clin Invest 1981; 67:1777. Krall JF, Connelly M, Tuck ML. Acute regulation of beta adrenergic catecholamine sensitivity in human lymphocytes. J Pharmacol Exp Ther 1980; 214:554. Bristow MR, Ginsburg R, Minobe W, et al. Decreased catecholamine sensitivity and beta-adrenergic-receptor density in failing human hearts. N Engl J Med 1982; 307:205. Williams LT, Lefkowitz RJ, Watanabe AM, et al. Thyroid hormone regulation of beta-adrenergic receptor number. J Biol Chem 1977; 252:2787. Gottshall RW, Itskovitz HD. Redistribution of renal cortical blood flow by renal nerve stimulation and norepinephrine infusion. Proc Soc Exp Biol Med 1977; 164:60. Gottschalk CW. Renal nerves and sodium excretion. Annu Rev Physiol 1979; 41:229. Prosnitz EH, DiBona GF. Effect of decreased renal sympathetic nerve activity on renal tubular sodium reabsorption. Am J Physiol 1978; 235:F557. Bello-Reuss E, Trevino DL, Gottschalk CW. Effect of renal sympathetic nerve stimulation on proximal water and sodium reabsorption. J Clin Invest 1976; 57:1104. Wagermark J, Ungerstedt U, Ljungqvist A. Sympathetic innervation of the juxtaglomerular cells of the kidney. Circ Res 1968; 22:149. Bevan RS, Tsuru H. Functional and structural changes in the rabbit ear artery after sympathetic denervation. Circ Res 1981; 49:478. Simpson P, McGrath A. Norepinephrine-stimulated hypertrophy of cultured rat myocardial cells is an alpha-1 adrenergic response. J Clin Invest 1983; 72:732. Levi J, Coburn J, Kleeman CR. Mechanism of the antidiuretic effect of beta-adrenergic stimulation in man. Arch Intern Med 1976; 136:25. Shimamoto K, Miyahara ML. Effect of norepinephrine infusion on plasma vasopressin levels in normal human subjects. J Clin Endocrinol Metab 1976; 43:201. Berl T, Cadnapaphornchai P, Harbottle JA, Schrier RW. Mechanism of stimulation of vasopressin release during beta-adrenergic stimulation with isoproterenol. J Clin Invest 1974; 53:857. Himms-Hagen J. Effects of catecholamines on metabolism. In: Blaschko H, Muscholl E, eds. Catecholamines: handbook of experimental pharmacology, vol Berlin: Springer-Verlag, 1972:363. Landsberg L, Saville ME, Young JB. The sympathoadrenal system and regulation of thermogenesis. Am J Physiol 1984; 247:E181. Rosen SG, Clutter WE, Shah SD, Cryer PE. Direct alpha-adrenergic stimulation of hepatic glucose production in human subjects. Am J Physiol 1983; 245:E616. Galster AD, Clutter WE, Cryer PE, et al. Epinephrine plasma thresholds for lipolytic effects in man: measurements of fatty acid transport with [1-13C] palmitic acid. J Clin Invest 1981; 67:1729. Deibert DC, DeFronzo RA. Epinephrine-induced insulin resistance in man. J Clin Invest 1980; 65:717. Smith U. Adrenergic control of lipid metabolism. Acta Med Scand 1983; 672(Suppl):41. Brown MJ, Brown DC, Murphy MB. Hypokalemia from beta-2-receptor stimulation by circulating epinephrine. N Engl J Med 1983; 309:1414. Potter DE. Adrenergic pharmacology of aqueous humor dynamics. Pharmacol Rev 1981; 33:133. Goldstein DS, Swoboda KJ, Miles JM, et al. Sources and physiological significance of plasma dopamine sulfate. J Clin Endocrinol Metab 1999; 84:2523. Wolfovitz E, Grossman E, Folio CJ, et al. Derivation of urinary dopamine from dihydroxyphenylalanine (DOPA) in humans. Clin Sci 1993; 84:549. Ziegler MG, Lake CR, Kopin IJ. Plasma noradrenaline increases with age. Nature 1976; 261:333. Lake CR, Ziegler MG, Kopin IJ. Use of plasma norepinephrine for evaluation of sympathetic neuronal function in man. Life Sci 1976; 18:1315. Goldstein DS, McCarty R, Polinsky RJ, Kopin IJ. Relationship between plasma norepinephrine and sympathetic neural activity. Hypertension 1983; 5:552. Kopin IJ, Zukowska-Grojec Z, Bayorh M, Goldstein DS. Estimation of intrasynaptic norepinephrine concentrations at vascular neuroeffector junctions in vivo. Naunyn Schmiedebergs Arch Pharmacol 1984; 325:298. Esler M, Jennings G, Korner P, et al. Total, end organ–specific, noradrenaline plasma kinetics in essential hypertension. Clin Exp Hypertens 1984; 6:507. Goldstein DS, Zimlichman R, Stull R, et al. Measurement of regional neuronal removal of norepinephrine in man. J Clin Invest 1985; 76:15. Goldstein DS, Eisenhofer G, Sax FL, et al. Plasma norepinephrine pharmacokinetics during mental challenge. Psychosom Med 1987; 49:591. Goldstein DS, Polinsky RJ, Garty M, et al. Patterns of plasma levels of catechols in neurogenic orthostatic hypotension. Ann Neurol 1989; 26:558. Linares OA, Zech LA, Jacquez JA, et al. Effect of sodium-restricted diet and posture on norepinephrine kinetics in humans. Am J Physiol 1988; 254:E 222. Goldstein DS, Stull R, Markey SP, et al. Dihydrocaffeic acid: a common contaminant in the liquid chromatographic-electrochemical measurement of plasma catecholamines in man. J Chromatogr 1984; 311:148. Lenders JWM, Eisenhofer G, Armando I, et al. Determination of metanephrines in plasma by liquid chromatography with electrochemical detection. Clin Chem 1993; 39:97. Peuler JD, Johnson GA. Simultaneous single isotope radioenzymatic assay of plasma norepinephrine, epinephrine and dopamine. Life Sci 1977; 2:625. Goldstein DS, Stull RW, Zimlichman R, et al. Simultaneous measurement of DOPA, DOPAC, and catecholamines in plasma by liquid chromatography with electrochemical detection. Clin Chem 1984; 30:815. Sundlof G, Wallin BG. Human muscle nerve sympathetic activity at rest. Relationship to blood pressure and age. J Physiol (Lond) 1978; 274:621. Wallin BG, Nerhed C. Relationship between spontaneous variations of muscle sympathetic activity and succeeding changes of blood pressure in man. J Auton Nerv Syst 1982; 6:293. Morlin C, Wallin BG, Eriksson BM. Muscle sympathetic activity and plasma noradrenaline in normotensive and hypertensive man. Acta Physiol Scand 1983; 119:117. Wallin BG, Sundlof G, Eriksson B-M, et al. Plasma noradrenaline correlates to sympathetic muscle nerve activity in normotensive man. Acta Physiol Scand 1981; 111:69. Schwaiger M, Kalff V, Rosenspire K, et al. Noninvasive evaluation of sympathetic nervous system in human heart by positron emission tomography. Circulation 1990b; 82:457. Goldstein DS, Holmes CS, Stuhlmuller JE, et al. 6-[18F]Fluorodopamine PET scanning in the assessment of cardiac sympathoneural function. Clin Auton Res 1997; 7:17. Robertson D, Johnson GA, Robertson RM, et al. Comparative assessment of stimuli that release neuronal and adrenomedullary catecholamines in man. Circulation 1979; 59:637. Ninomiya I, Nisimaru N, Irisawa H. Sympathetic activity to the spleen, kidney, and heart in response to baroreceptor input. Am J Physiol 1971; 221:1346. Chien S. Role of the sympathetic nervous system in hemorrhage. Physiol Rev 1967; 47:214. Johnson TS, Young JB, Landsberg L. Sympathoadrenal responses to acute and chronic hypoxia in the rat. J Clin Invest 1983; 71:1263. Goldstein DS, Dionne R, Sweet J, et al. Circulatory, plasma catecholamine, cortisol, lipid, and psychological responses to real-life stress (wisdom tooth extractions): effects of diazepam sedation and of inclusion of epinephrine with the local anesthetic. Psychosom Med 1982; 44:259. Robertson RM, Medina E, Shah N, et al. Neurally mediated syncope: patho-physiology and implications for treatment. Am J Med Sci 1999; 317:102. Wallin BG, Sundlof G. Sympathetic outflow to muscles during vasovagal syncope. J Auton Nerv Syst 1982; 6:287. Pacak K, Palkovits M, Yadid G, et al. Heterogeneous neuroendocrine responses to various stressors: a test of Selye's doctrine of non-specificity. Am J Physiol 1998; 275:R1247.

CHAPTER 86 PHEOCHROMOCYTOMA AND OTHER DISEASES OF THE SYMPATHETIC NERVOUS SYSTEM Principles and Practice of Endocrinology and Metabolism

CHAPTER 86 PHEOCHROMOCYTOMA AND OTHER DISEASES OF THE SYMPATHETIC NERVOUS SYSTEM HARRY R. KEISER Pheochromocytoma Clinical Manifestations Diagnosis Urine Tests Blood Tests Pharmacologic Tests Localization Treatment Pheochromocytoma in Pregnancy Malignant Pheochromocytoma Adrenal Medullary Hyperplasia Other Tumors that Produce Catecholamines Orthostatic Hypotension Familial Dysautonomia (Riley-Day Syndrome) Chapter References

PHEOCHROMOCYTOMA The most important disease of the adrenal medulla is pheochro-mocytoma.1,2,3,4,5,6 and 7 Eighty-five percent to 90% of these tumors occur in the adrenal glands. They arise from cells of neural crest origin and can also be found in some extraadrenal locations (i.e., carotid body, aortic chemoreceptors, sympathetic ganglia, organ of Zuckerkandl). These extraadrenal tumors have been called paragangliomas, chemodectomas, and the like, depending on their anatomic location. Any tumor that makes, stores, and secretes catecholamines and produces symptoms of excessive catecholamine release should be considered a pheochromocytoma and treated as such, regardless of its location. Pheochromocytomas are rare except in certain familial settings. They occur equally in both sexes and at any age, although they are most frequent between the ages of 30 and 50 years. They occur in all races, but less frequently in blacks. When they occur sporadically, 10% are bilateral. When they occur in families, ~50% are bilateral and they are usually intraadrenal. Pheochromocytomas may occur as a solitary entity or they may be familial, associated with other diseases, such as multiple endocrine neoplasia type 2A or type 2B, von Hippel-Lindau disease type 2, or neurofibromatosis type 1. In most familial disorders inheritance is autosomal dominant8 (Table 86-1) (see Chap. 188). In one study, 23% of unselected patients with pheochromocytoma were found to be carriers of familial disorders—19% had von Hippel-Lindau disease and 4% had multiple endocrine neoplasia type 2.9 In addition, 38% of carriers of von Hippel-Lindau disease and 24% of carriers of multiple endocrine neoplasia type 2 had pheochromocytoma as the only manifestation of their syndrome. Thus, all patients in families with multiple endocrine neoplasia type 2 or von Hippel-Lindau disease should be genetically screened to determine if they are carriers and require screening for specific phenotypic traits.9a Patients who are hypertensive or are scheduled to undergo surgery for another aspect of the disease should be screened for pheochromocytoma, even if they are asymptomatic. Although pheochromocytoma is not a normal part of multiple endocrine neoplasia type 1, it has been associated with pancreatic islet cell tumors in some families.10 In all situations in which several different tumors have been discovered contemporaneously, the pheochromocytoma should be removed before any other surgery is performed. An adrenal gland should be removed only if it contains tumor. A normal-appearing adrenal gland should be left intact and carefully monitored. A pheochromocytoma may not develop in the remaining gland; in one study it appeared in 52% of patients after a mean of 11.87 years.11

TABLE 86-1. Pathologic Conditions Associated with Pheochromocytoma

Although ~5% of patients with pheochromocytoma have neurofibromatosis, only 1% of patients with neurofibromatosis have pheochromocytoma.1 In children, 50% of pheochromocytomas are solitary and intraadrenal, 25% involve the adrenals bilaterally, and 25% are extraadrenal (see Chap. 87). CLINICAL MANIFESTATIONS The hallmark of a pheochromocytoma is hypertension, either paroxysmal or sustained (both forms occur with equal frequency). Usually, the hypertension is labile. A typical paroxysm is characterized by a sudden major increase in blood pressure, a severe, throbbing headache, profuse sweating over most of the body (especially the trunk), palpitations with or without tachycardia, anxiety, a sense of doom, skin pallor, nausea with or without emesis, and abdominal pain.1,2 A study of 2585 hypertensive patients, including 11 with pheochromocytoma, noted that headache, sweating, and palpitations were reported by 71%, 65%, and 65%, respectively, of patients with pheochromocytoma.12 When these three manifestations were present and associated with hypertension, they indicated the diagnosis of pheochromocytoma with a specificity of 93.8% and a sensitivity of 90.9%. In the absence of hypertension and this triad of symptoms, the diagnosis of pheochromocytoma could be excluded with a certainty of 99.9%. The hypertensive episode may appear to occur spontaneously, or it may follow acute physical (but not psychological) stress, an increase in abdominal pressure (induced, for example, by palpation of the tumor or defecation), a hypotensive episode, or any activation of the sympathetic nervous system. Hypertensive episodes are usually not caused by anxiety or psychological stress. They may last a few minutes to several hours, and they leave the patient feeling exhausted. These hypertensive episodes may occur in patients who are otherwise normotensive, or they may affect patients with sustained hypertension. Hypertensive episodes may be precipitated by drugs that lower blood pressure acutely (i.e., saralasin13), as well as by opiates, adrenocorticotropic hormone (ACTH), dopamine antagonists (i.e., metoclopramide14 and droperidol), radiographic contrast media, indirectacting amines (such as tyramine and amphetamine, which act by displacing the normal neurotransmitter), proprietary cold preparations, and drugs that block neuronal catecholamine reuptake (i.e., tricyclic antidepressants and cocaine). Other symptoms include dizziness and constipation. The patient may experience postural hypotension, hyperglycemia, hypermetabolism, weight loss, and even psychic changes. The hyperglycemia is generally mild, occurs with the hypertensive attacks, and usually does not require treatment. A few patients may say that they feel “flushed,” but observation of an actual flush should bring into question the diagnosis of a pheochromocytoma.3 The hypertensive crisis associated with pheochromocytoma may precipitate a myocardial infarction, even in the absence of coronary artery disease, or a cerebral hemorrhage. A prolonged excess of catecholamines may produce a cardiomyopathy that is manifested by congestive heart failure. Unexplained shock may also be the presenting feature of pheochromocytoma.15 A pheochromocytoma with all the classic signs and symptoms is easy to diagnose. However, the tumor may present with only part of the syndrome. Early in the course of a pheochro-mocytoma, symptoms may be very mild and infrequent, becoming progressively more severe and more frequent later. A patient may have symptoms for years before the diagnosis is made. The extensive differential diagnosis of a pheochromocytoma includes anxiety and panic attacks, thyrotoxicosis, abrupt withdrawal of clonidine therapy, amphetamine use, cocaine use, ingestion of tyramine-containing foods or proprietary cold preparations (e.g., phenylephrine, pseudoephedrine, phenylpropanolamine) while taking monoamine oxidase inhibitors, hypoglycemia, insulin reaction, angina pectoris or myocardial infarction, brain tumor, subarachnoid hemorrhage, cardiovascular deconditioning, menopausal syndrome, neuroblastoma (in children), and toxemia of pregnancy, among others. A pseudopheochromocytoma has been reported in a

patient with parkinsonism treated with selegiline, a monoamine oxidase type B inhibitor, and fluoxetine, an antidepressant that inhibits serotonin uptake.16 Clozapine, a tricyclic dibenzodiazepine, has been reported to mimic pheochromocytoma with hypertension and increased urinary catecholamine levels, both of which revert to normal when the drug is stopped.17 To make these diagnostic distinctions, the physician must obtain a detailed history and, if possible, observe and monitor a hypertensive episode. If this is done, the self-administration of various drugs, the abrupt withdrawal of clonidine (which causes a rebound in central sympathetic outflow resulting in the release of large amounts of epinephrine and norepinephrine), and other episodic disorders are evident. Two of the most difficult differential diagnoses are panic attacks and cardiovascular deconditioning, the latter of which indicates that a patient has been so sedentary that minimal physical activity causes marked tachycardia and sympathetic nervous system activity. These conditions present diagnostic difficulties because neither is characterized by specific symptoms, physical findings, or diagnostic tests. Thus, a pheochromocytoma must be excluded before either of these diagnoses may be established. Pheochro-mocytoma should be suspected in patients with hypertension that is refractory to therapy and in those who have a history of hypertensive episodes occurring during previous general anesthesia or abdominal surgery. Pheochromocytomas may produce calcitonin, opioid peptides, somatostatin, ACTH, and vasoactive intestinal peptide.4 The high serum calcitonin level may mistakenly suggest a medullary carcinoma of the thyroid until the pheochromocytoma is removed.18 In some patients with pheochromocytoma, the ACTH production has been known to cause Cushing syndrome,19 and production of vasoactive intestinal peptide has resulted in watery diarrhea.6 Pheochromocytomas also produce and secrete into the blood neuron-specific enolase, chromogranin A (CGA), neuropeptide Y, and adrenomedullin. DIAGNOSIS URINE TESTS The diagnosis of a pheochromocytoma must rest on biochemical determinations (i.e., the demonstration of elevated levels of catecholamines or their metabolites in blood or in urine). The traditional and easiest method is the measurement of epinephrine and norepinephrine, of metanephrines (methoxycatecholamines), or of vanillylmandelic acid in a 24-hour collection of urine (Table 86-2).1,2,3 and 4 The urine must be collected in strong acid (i.e., 20 mL of 6N HCl or 25 mL of 50% acetic acid) in an easily sealed, leakproof container; it need not be refrigerated. The measurement of metanephrines in a single random urine sample, or of norepinephrine in a single overnight sample of urine, has been used successfully to diagnose pheochromocytoma.20,21 However, such random specimens show greater variability and, therefore, have less diagnostic sensitivity than do 24-hour samples. Creatinine should be measured in all collections of urine to ensure adequacy of the collection and to serve as a denominator for timed or random samples. Simultaneous analysis of norepinephrine and its metabolite dihydroxyphenolglycol may improve specificity.22 The amount of urinary total metanephrines and vanillylmandelic acid, when expressed in relation to creatinine, decreases with age in normal children.23 Although an amine-free or “vanillylmandelic acid” diet is useful in metabolic studies, it is usually unnecessary in the diagnosis of a pheochromocytoma now that more specific analyses are used. The patient should cease taking all medications, if possible (see Table 86-2). If antihypertensive therapy must be continued, diuretics, vasodilators (hydralazine or minoxidil), calcium channel blockers, and angiotensin-converti ng enzyme inhibitors cause minimal interference. Although any one of the three urinary tests usually indicates the diagnosis, the determination of norepinephrine and epinephrine, or of normetanephrine and metanephrine, is preferred.24,25 Specific assays for urinary epinephrine and norepinephrine are preferable to measurement of total catecholamines, because increased epinephrine may be the only abnormal finding in patients with multiple endocrine neoplasia.26 A markedly elevated value for total catecholamines (i.e., 2000 µg per 24 hours) with normal values for metanephrines and vanillylmandelic acid suggests that the patient is taking methyldopa (metabolized to a-methylnorepinephrine and determined as a catecholamine) and does not have a pheochro-mocytoma. In most patients with a pheochromocytoma, the urine test results are abnormally elevated at all times. For the few patients with only infrequent episodes of hypertension, urine is best collected when the patient is hypertensive. This is accomplished most easily by giving the patient 5 mL of 6N HCl in a 1-L container with instructions to be followed when hypertensive symptoms occur. When symptoms are noted, the patient should void, discard the urine, and note the time. Then, 2 to 3 hours later, the patient should void again, now saving the urine in the container and again noting the time. This urine specimen should be assayed for epinephrine, norepinephrine, and creatinine. A greater than twofold increase in either cate-cholamine over levels obtained from testing a similar control collection is diagnostic.

TABLE 86-2. Effects of Various Drugs on Results of Biochemical Urine Tests for Pheochromocytoma*

BLOOD TESTS The blood sample to test for catecholamines27 (see Chap. 85) must be drawn under controlled circumstances; the patient should be fasting and supine, and the needle should be in place in the vein for at least 20 minutes before the sample is drawn. The blood sample should immediately be placed on ice and centrifuged and the plasma drawn off and frozen within 1 hour. Normal values are generally 300 pg/mL for epinephrine are diagnostic of a pheochromocytoma. Intermediate values may not represent a tumor but rather an exaggerated physiologic response. In this situation, a suppression test (see the following discussion) is indicated. The presence of dihydrocaffeic acid, a catecholamine metabolite of the caffeic acid that is present in all forms of coffee (decaffeinated, as well as regular) and many cola beverages may produce false-positive results.28 Therefore, patients are best studied in the morning after an overnight fast. A report suggests that the plasma ratio of norepinephrine to dihydroxyphenolglycol (mainly from neuronally released norepinephrine) is >2.0 in patients with a pheochromocytoma and is 35 mm Hg systolic and 25 mm Hg diastolic after administration of 1 to 5 mg of the drug as an intravenous bolus is considered to be diagnostic. This reduction begins 2 minutes after the injection and lasts ~10 minutes. Because patients with a pheochromocytoma may become hypotensive after a large dose of phentolamine, the prudent course is to inject only 1.0 mg initially and to observe the response. If blood pressure is unaffected, the remaining 4.0 mg in the vial can then safely be administered. However, the test should be performed only when the patient is hypertensive and preferably when he or she is not receiving any other antihypertensive agents. The blood pressure and heart rate must be monitored every 30 to 60 seconds, and a patent intravenous line and vasopressors must be at hand. False-positive tests are common.

Clonidine (Catapres), a centrally acting, orally active, a2-adrenergic agonist, can also be used as a diagnostic suppression test.36 Blood is drawn to test for plasma catecholamines both before and 3 hours after the oral administration of clonidine (0.3 mg/70 kg). Blood pressure declines in most patients with hypertension of any cause. Plasma catecholamine levels decrease if they are under physiologic control, whereas in patients with pheochromocytoma, plasma catecholamines remain the same or increase. This test is very safe but is useful only if basal cate-cholamine values are abnormally high. Similarly, pentolinium, administered by the intravenous route, has been used in a suppression test. However, it has no unique advantages, and the intravenous form is no longer available in the United States. Histamine,37 tyramine, glucagon,37,38 and metoclopramide14 each has been used as a provocative test for pheochromocytoma. Such provocation is inherently dangerous, as indicated by the several deaths that have been reported during histamine tests. These tests should be used only if repeated assays of blood and urine are equivocal and if symptoms are classic, truly episodic, but too infrequent to permit prompt diagnosis. Although many physicians focus on the blood pressure response, it is only of value if the patient has discontinued all medications. Plasma catecholamines are best measured before and during the test (generally 2 minutes after injection of the drug) and only significant elevations of plasma catecholamines should be considered as diagnostic of a pheochromocytoma. If this is done, the provocation can provide useful diagnostic information, without any risk, in a patient who is receiving a-and b-adrenergic blockers. The usual dose of glucagon is 1.0 mg, delivered as an intravenous bolus. The usual dose for metoclopramide is 1 to 10 mg, also delivered as an intravenous bolus (the 1-mg dose is recommended initially; if the result is negative, then the 10-mg dose may be tried). Close monitoring of blood pressure is required, and phentolamine must be at hand for the treatment of any episodes of severe hypertension. Histamine and tyramine are rarely used anymore. A glucagon provocative test followed by a clonidine suppression test can be easily and safely performed on outpatients, and the combination yields increased sensitivity and specificity in the diagnosis of pheochromocytoma.39 Even when glucagon is given to patients with known pheochromocytoma, only 18% have an elevation of systolic blood pressure over 200 mm Hg. The hypertension responds rapidly to bolus intravenous administration of phentolamine, 5 mg. Urinary epinephrine and norepinephrine secretion is normally suppressed by sleep. In patients with pheochromocytoma, however, the catecholamine level may remain unregulated by sleep; this effect can be further accentuated with the addition of oral clonidine before bedtime.40 The diagnostic utility and specificity of this procedure remain to be determined. LOCALIZATION A computed tomographic (CT) scan of the abdomen is the best all-around technique for locating a pheochromocytoma. However, attempts to locate the tumor should not be made until biochemical studies have confirmed its presence. The demonstration of a mass in an adrenal gland does not prove that the mass is a pheochromocytoma, just as failure to find a mass in either adrenal gland does not prove that the patient does not have a pheochromocytoma. However, in certain familial settings (e.g., von Hippel-Lindau disease or multiple endocrine neoplasia type 2), a computed tomogram of the abdomen or a scan with radiolabeled metaiodobenzylguanidine (MIBG) may point to the need for further workup when the basal laboratory data have been equivocal and the patient is scheduled for surgery for another component of the syndrome. 9 A careful history and physical examination may yield important clues as to the location of a pheochromocytoma, as in the case of postmicturition hypertension secondary to a pheochromocytoma of the urinary bladder.41 If epinephrine constitutes >15% of the total urinary catecholamines, then the tumor is almost always intraadrenal. Overall sensitivity of the studies in the localization of pheochromocytoma was 89% for CT, 98% for magnetic resonance imaging (MRI), and 81% for MIBG scan.42 MRI is a useful diagnostic tool because it provides sufficient histologic specificity to distinguish pheochromocytomas from other adrenal masses via T2 and spin echo sequences.43,44 The iodine-131 (131I)–labeled MIBG scan is accurate in revealing 80% to 95% of pheochromocytomas, but it may also visualize neuroblastoma and some other APUD (amine precursor uptake and decarboxylation) tumors (i.e., carcinoid and medullary carcinoma of the thyroid).45,45a Labetalol reduces tumor uptake of MIBG, and the drug should be stopped up to 1 week before an MIBG scan.46 Other drugs known to reduce or suspected of reducing MIBG uptake include reserpine, cocaine, tricyclic antidepressants, sympathomimetics (including non-prescription compounds containing phenylpropanolamine), and calcium channel blockers.46 MIBG scanning is most useful in detecting multiple tumors, tumor metastases, and tumors in unusual locations (e.g., cardiac tumors).47 Both 123I-MIBG and single photon emission CT have been used in place of planar imaging to provide improved resolution.48 Use of a radiolabeled somatostatin analog, such as octreotide, is not as helpful as MIBG in locating a pheochromocytoma. In a comparative study of patients with known pheochromocytoma metastases, detection rates were 80% with MIBG and 44% with octreotide. However, octreotide imaging detected six metastases that were not detected by MIBG scanning. Thus, octreotide scanning may be used as a complementary method in detecting metastases when results from MIBG imaging are negative.49 The major role of ultrasonography has been to determine whether a mass seen on radiographic examination is attached to the kidney and whether it is solid or cystic. However, transesophageal echocardiography has proven to be valuable in the visualization of intrapericardial tumors.50 With the ready availability of CT and MRI, arteriography for the diagnosis of a pheochromocytoma is rarely indicated. An arteriogram imposes significant risk of a serious hypertensive crisis and offers little additional information. In the rare event that one is needed to clarify complicated vascular anatomy before surgery, the patient should be prepared and monitored as if he or she were undergoing surgery. A midstream injection into the abdominal aorta may yield all needed information and is much less likely to trigger a hypertensive crisis than is selective arterial injection. The measurement of catecholamine levels in selected veins is a procedure that is also rarely needed, except when an extraadrenal tumor that escaped removal during previous surgery must be located. This procedure generally does not help to locate an intraadrenal tumor, because catecholamine levels in the adrenal veins normally vary tremendously over time, during stress, and from one side to the other. An excess of norepi-nephrine in the venous effluent of one adrenal would suggest a pheochromocytoma, because epinephrine normally predominates. Also, dilution effects are important, because the right adrenal vein is short and can be entered directly from the inferior vena cava, whereas the left adrenal vein is accessible only through the left renal vein. Simultaneous measurement of cortisol or aldosterone can provide the denominator necessary to control for the distance of the catheter from the tumor (see also Chap. 88). TREATMENT Surgical removal of a pheochromocytoma is clearly the treatment of choice, because an unresected tumor represents a buried time bomb waiting to explode with a potentially fatal hypertensive crisis.1,2,3 and 4 Safe surgical removal requires an integrated, cooperative, team effort by an internist, an anesthesiologist, and a surgeon, preferably one with prior experience in removal of pheochromocytoma (see Chap. 89). Preparation should start at least 7 days before surgery with the administration of the nonspecific a-adrenergic receptor blocker phenoxybenzamine (Dibenzyline). The initial dosage of 10 mg twice daily should be increased to an average dosage of 0.5 to 1.0 mg/kg daily, delivered in two to three divided doses. Prazosin (Minipress),51 a specific a1-antagonist (2–5 mg two to three times daily) or labetalol (Normodyne, Trandate), a drug with both a-and b-antagonist activity (200–600 mg orally, twice daily) may also be used. Large doses of these drugs may be necessary to control blood pressure and to restore the patient's diminished plasma volume. As normal blood volume is restored, the degree of postural hypotension decreases. However, if too much a-blocker is administered, the degree of postural hypotension increases and may cause symptoms, just as it does in normal subjects. A b-adrenergic blocker seldom is needed unless a significant tachycardia or catecholamine-induced arrhythmia occurs. Clearly, a b-blocker should never be used in the absence of an a-blocker, because the former would exacerbate epinephrine-induced vasoconstriction by blocking its vasodilator component. The author has found that metyrosine (Demser, a-methyl-L -tyrosine) is very useful in treating patients with pheochromocytoma. The drug competitively inhibits tyrosine hydroxylation,52 the rate-limiting step in catecholamine synthesis. It facilitates blood pressure control both before and during surgery, especially during the induction of anesthesia and manipulation of the tumor.53 Treatment is customarily begun at a dosage of 250 mg every 6 hours; thereafter, the dosage is increased by 250 to 500 mg per day, as necessary, up to a total dose of 4 g per day. Because metyrosine is a substituted amino acid, it readily enters the brain. There, it inhibits catecholamine synthesis and may produce sedation (frequently) and extrapyramidal signs (rarely, parkinsonism in older patients). These symptoms disappear rapidly when the drug is discontinued. Adverse central nervous system symptoms are best handled by reducing the dosage if the patient reports unusually vivid or frightening dreams or if significant behavioral changes occur. Although the author prefers the more stable blood pressure produced by the combination of a-antagonists and metyrosine, physicians in some centers, such as the Cleveland Clinic and Mayo Clinic, report equally good results without metyrosine. The calcium channel blocker nifedipine (10 mg per administration, as necessary), which is absorbed quickly from the stomach, is useful in controlling infrequent spikes in blood pressure.54,55 Nicardipine, another calcium channel blocker, has been used alone to control blood pressure both before and during surgical removal of pheochromocytoma.56 Calcium channel blockers are potent vasodilators, and they also block the entry of calcium, which is needed for release of catecholamines from the tumor cell. During surgery, phentolamine, nicardipine, or nitroprusside should be administered to control hypertensive episodes. The infusion of dilutions of the drug is preferable to bolus injections. Small amounts of b-blockers, such as propranolol, in intravenously administered doses of 0.1 to 1.0 mg can be used, if needed, to treat tachycardia or catecholamine-induced arrhythmias. If hypotension occurs, either during surgery (especially after the tumor is removed) or after surgery, then volume replacement is the treatment of choice.1,4 The use of

pressor agents is ill-advised, especially if long-lasting a-blockers have been used, because of the difficulty with which patients are weaned from these agents. Control of postoperative hypotension is even more difficult if both metyrosine and phenoxybenzamine (Dibenzyline) are used preoperatively. This is because the former inhibits catecholamine synthesis by both the tumor and the sympathetic nervous system, whereas the latter blocks the action of any catecholamines that are synthesized. Thus, the vascular system is effectively paralyzed in a dilated state. The best way to control blood pressure is by careful attention to volume replacement. The volume of fluid required is frequently large (i.e., one-half to one and one-half times the patient's calculated blood volume) during the first 24 to 48 hours after removal of the tumor, because this long is required for the effects of both drugs to disappear and for the sympathetic nervous system to resume autoregulation. This return to autoregulation becomes apparent when the renal output suddenly begins to increase and the blood pressure and heart rate remain stable. At this time, replacement volumes can be decreased to standard amounts (i.e., 125 mL per hour). Occurrence of postoperative hypertension may mean that some tumor tissue was not resected. However, during the first 24 hours after surgery, hypertension is most likely attributable to pain, volume overload, or autonomic instability.52 Because major surgery causes large increases in both plasma and urinary catecholamine levels, a 5- to 7-day interval should precede any attempt to collect samples for biochemical evidence of residual tumor. A suppression test with phentolamine or clonidine may be very useful in clarifying the cause of persistent hypertension after the removal of a pheochromocytoma. In 25% of such cases, the hypertension is attributable to preexisting hypertension, usually the essential form. PHEOCHROMOCYTOMA IN PREGNANCY A pheochromocytoma in a pregnant woman is a very dangerous condition.1,5,57 The very high overall maternal and fetal mortality rates of 48% and 55%, respectively, as had been reported in 1980, can be reduced to 0% and 15%, respectively, only if the diagnosis of pheochromocytoma is made antenatally.58,59 As soon as the diagnosis is made, a-adrenergic blockers should be administered; phenoxybenzamine is the drug of choice. Localization is best accomplished by ultrasonography or MRI. If a pheochromocytoma is discovered during the first or second trimester, the tumor should be removed as soon as the patient is prepared adequately with a-adrenergic blockers. This has been done laparoscopically.60 The fetus can remain undisturbed, but spontaneous abortion is very likely. If a pheochromocytoma is discovered during the third trimester, the patient may be treated with a-adrenergic blockers and carefully monitored until the fetus matures to the point of viability. Then, cesarean section and removal of the tumor can be accomplished in one operation. If uncontrolled hypertension, hemorrhage, or other emergency occurs, the tumor should be removed at once. MALIGNANT PHEOCHROMOCYTOMA A pheochromocytoma is usually surrounded by a firm capsule or pseudocapsule. The cut surface shows many areas of hemorrhagic necrosis. Microscopic examination reveals that the tumor is composed of large, pleomorphic, chromaffin cells (Fig. 86-1) that contain many dense core granules on electron microscopy. For a pathologist to determine if a pheochromocytoma is benign or malignant is very difficult, if not impossible. The major criteria for differentiation are invasion of capsular blood vessels by tumor and metastatic invasion of other tissues. Without these findings, a pheochromocytoma may only appear to be benign. Thus, patients should have their blood pressure checked annually, and blood or urine catecholamine levels should be monitored annually for 5 years to detect any recurrence. A documented recurrence does not necessarily mean that the original tumor was malignant, but it should prompt a fresh evaluation of the patient.61 Assessing the serum level of various neuroendocrine tumor markers, such as CGA or neuron-specific enolase,62 may provide additional information that allows differentiation between benign and malignant tumors.63 Flow cytometric DNA analyses of pheochromocytoma have indicated that nondiploid tumors are more prone to aggressive behavior than diploid tumors; therefore, the former should be monitored more carefully.64

FIGURE 86-1. A, Cross-section of a well-demarcated intraadrenal pheochromocytoma with stippled areas of congestion. The normal adrenal cortex, which had a characteristic yellow-orange color when fresh, is marked with arrows. B, The typical histology of an adrenal pheochromocytoma is shown with cells in nests and trabeculae separated by a thin vascular stroma (hematoxylin and eosin; ×400). C, Electron microscopic view of an adrenal pheochromocytoma with many large, dense-core catecholamine-containing granules, some marked with arrows. These granules, with a thin uniform halo between the core and the investing membrane, are the epinephrine-containing type. Granules with a prominent eccentric lucent space between the core and the limiting membrane are the norepinephrine-containing type. ×13,500 (B, Courtesy of Dr. Ilona Linnoila.)

Approximately 10% to 15% of pheochromocytomas are malignant. The larger number applies to extraadrenal tumors. Malignant pheochromocytomas grow larger, invade adjacent structures, and metastasize to bone, liver, lung, and lymph nodes. Aggressive surgical removal of all accessible tumor tissue is appropriate. Symptoms should be controlled with phenoxybenzamine or prazosin, metyrosine, and b-adrenergic blockers, as needed. Radiation therapy to bony metastases relieves pain and frequently kills tumor at that site. Originally, 131I-labeled MIBG therapy held promise as a mode of treatment for malignant pheochromocytoma. However, current reports indicate that the treatment is only palliative.65 The rate of spread of malignant pheochromocytomas is highly variable; median survival is 5 years. A rapidly growing pheochromocytoma metastasis that produces impending compression of the spinal cord is a medical emergency that necessitates immediate high doses of corticosteroids and radiation therapy. A chemotherapeutic regimen used for the treatment of the closely related neuroblastoma consists of intravenous cyclophosphamide, 750 mg/m2, on day 1; vincristine, 1.4 mg/m2, on day 1; and dacarbazine, 600 mg/m2, on days 1 and 2. This regimen is then repeated every 21 days with adjustments being made in the dose of cyclophosphamide as hematologic parameters change and in the dose of vincristine if neurologic toxicity is noted. Fourteen patients with malignant pheochromocytoma were treated and followed for 6 to 35 months (median duration of 25 months). Two patients had complete remissions, 6 patients had >50% regression of all measurable tumor sites, and 3 patients had >25% regression (median duration of response was >17 months). In two patients, the disease stabilized after one course of therapy; one patient had progressive disease despite therapy.66 All responding patients experienced improvement in performance status and hypertension, and a reduction in or elimination of antihypertensive therapy became possible. The excretion of catecholamines or their metabolites correlated with the tumor response in all cases. However, all treated patients eventually died of their disease. Therefore, this regimen of combination chemotherapy can produce a partial remission of significant duration in many patients with malignant pheochromocytoma; however, it cannot cure these patients, and it is effective for only a finite period. This regimen is recommended only after all other therapeutic modalities have been exhausted and the patient is documented as having progressive and measurable disease that is negatively affecting his or her quality of life. ADRENAL MEDULLARY HYPERPLASIA Adrenal medullary hyperplasia has been reported in a few cases, most of which have been in the familial setting of multiple endocrine neoplasia type 2A or 2B.67 The diagnosis is determined by analysis of an adrenal gland removed from a patient who appears to have a pheochromocytoma on the basis of signs, symptoms, and laboratory tests. Yet, on histologic examination, the adrenal gland does not contain a tumor but has diffuse, nonnodular, adrenal medullary hyperplasia, characterized by increased mitotic activity, decreased corticomedullary ratio, increased weight, and total catecholamine content. These cases suggest, but do not prove, that adrenal medullary hyperplasia may be a precursor of pheochromocytoma. In humans, no demonstrable physiologic consequences of the alternative state (i.e., adrenal medullary hypoplasia or aplasia) are seen. OTHER TUMORS THAT PRODUCE CATECHOLAMINES Neuroblastoma, ganglioneuroblastoma, and ganglioneuroma are all solid tumors of neural crest origin that occur in sympathetic ganglia or the adrenal medulla. They all can synthesize and excrete the catecholamines and metabolites that are found in pheochromocytoma. Moreover, they also excrete increased amounts of the

catecholamine precursors dihydroxyphenylalanine (DOPA) and dopamine, and of the metabolite homovanillic acid.68 Only one in five patients with neuroblastoma is hypertensive, probably owing to the secretion of large amounts of inactive catecholamine precursors and only small amounts of active catecholamines.69 The tumors form a wide spectrum of disease activity. Neuroblastomas occur mainly in children and are highly malignant. However, spontaneous regressions have been reported.70 Ganglioneuromas occur in both adults and children and are benign, well-differentiated tumors. Ganglio-neuroblastomas are intermediate in both their malignant nature and cellular differentiation. These tumors should be diagnosed and treated as if they were pheochromocytomas.

ORTHOSTATIC HYPOTENSION Orthostatic hypotension is defined as a critical and symptomatic reduction in blood pressure on the assumption of upright posture. It is caused by the failure of normal blood pressure homeo-stasis, and it reflects the loss of vasoconstrictor reflexes in both resistance and capacitance vessels. This loss is attributable to defects in either peripheral (postganglionic) neurons or central (preganglionic) neurons.71 The first category, which is the largest, includes many diseases with associated peripheral neuropathies, such as diabetes mellitus, tabes dorsalis, alcoholism, amyloido-sis, and the Holmes-Adie syndrome (unilateral, enlarged pupil, an absent or delayed response to light, a slow reaction on convergence, and a good response to miotics or mydriatics). In these diseases, the orthostatic hypotension is usually accompanied by dysfunctions in potency, sweating, and sphincter control. A consensus statement has established the following categories of primary autonomic insufficiencies or dysautonomias: (a) pure autonomic failure that occurs without other neurologic features, (b) Parkinson's disease with autonomic failure, and (c) multiple system atrophy, a sporadic, progressive disorder of adults characterized by autonomic dysfunction, parkinsonism, and ataxia in any combination. The latter category is divided into three types: (a) striatonigral degeneration when parkinsonism predominates, (b) olivopontocerebellar atrophy when cerebellar features predominate, and (c) Shy-Drager syndrome when autonomic failure predominates.71 Further differentiation of these disorders has been made on the basis of neurochemical analyses and positron emission tomography using 6-[18F]fluorodopamine.72 A classification of dysautonomias has been proposed72 in which sympathetic neurocirculatory failure results from peripheral sympathetic denervation or decreased or absent sympathoneural traffic, with or without signs of central neural degeneration, and in which both parkinsonism with sympathetic neurocirculatory failure and multiple system atrophy without sympathetic neurocirculatory failure differ from the Shy-Drager syndrome. Patients with the Shy-Drager syndrome have poor or no response to therapy with levodopa-carbidopa, intact cardiac sympathetic innervation, and a poor prognosis, whereas patients with parkinsonism, postural hypotension, and cardiac sympathetic denervation have a better prognosis.72 Treatment for orthostatic hypotension is mainly symptomatic once all treatable causes have been addressed. Certain non-pharmacologic measures should be implemented at once, including getting the patients on their feet for even short periods and having patients squat, bend over, and cross the legs. Support hose, fitted elastic stockings, and pressure suits have been used, but they correct only mild disorders. Fludrocortisone therapy plus a high-salt diet is the best treatment approach. Initial doses of 0.1 mg of fludrocortisone daily may be increased by 0.1 mg each week, as necessary, up to a maximum of 1 mg per day. This therapy should be combined with an increase in dietary salt to >150 mEq per day, either by consumption out of the salt shaker or by use of salt tablets. Serious complications occur that frequently limit this form of therapy. These include recumbent hypertension, congestive heart failure, edema, and hypokalemia. Recumbent hypertension is best treated by raising the head of the patient's bed.73,74 and 75 Various drugs that have been effective in the treatment of orthostatic hypotension include phenylpropanolamine, mido-drine, yohimbine, and indomethacin. Ibuprofen, caffeine, and methylphenidate have been ineffective.76 Generally, dosages have been difficult to control and must be individualized; many patients become hypertensive when recumbent and hypotensive when erect. One approach, which has been effective in some difficult cases, is to provide long-term administration of an oral monoamine oxidase inhibitor and to have the patient use phenylephrine nose drops before sitting or standing.77 Such a regimen requires an intelligent, cooperative patient and some experimentation. Rapid atrial pacing on demand has been used successfully in one carefully selected patient with bradycardia and low cardiac output.78 One research group has experimented with the use of a portable, servo-controlled, automated syringe, controlled by a microcomputer, to infuse a vasopressor for maintenance of the patient's blood pressure at a predetermined level.79

FAMILIAL DYSAUTONOMIA (RILEY-DAY SYNDROME) Familial dysautonomia, also known as Riley-Day syndrome, is characterized by autonomic instability. In neonates, the disease is manifested by dysphagia, difficulty in feeding, absence of fungiform papillae of the tongue, hyporeflexia, and slow movement.80 Later, abnormalities in sweating, temperature control, and blood pressure regulation become evident, along with impairment in sensations of pain, temperature, and taste as well as lack of lacrimation. Hypersensitivity to both adrenergic and cholinergic agonists and loss of skin flare after the injection of histamine are noted. A loss of small-caliber neurons in peripheral nerves and in dorsal root and sympathetic ganglia also occurs. The disease is transmitted as an autosomal recessive trait and affects mainly Ashkenazi Jews. The gene responsible appears to reside at 9q31-33.81 Only a few patients survive to adulthood. Usually, the diagnosis is self-evident. However, the labile blood pressure may lead to confusion with pheochromocytoma or neuroblastoma. Symptomatic postural hypotension has been treated with mido-drine (see earlier). Basal supine plasma levels of norepinephrine are normal, but no increase occurs on assumption of upright posture. Urinary levels of dopamine and homovanillic acid are increased, whereas levels of vanillylmandelic acid and methoxy-hydroxyphenylglycol are decreased. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

Manger WM, Gifford RW Jr. Clinical and experimental pheochromocytoma, 2nd ed. Cambridge, MA: Blackwell Science, 1996. Gifford RW Jr, Manger WM, Bravo EL. Pheochromocytoma. Endocrinol Metab Clin North Am 1994; 23:387. Ram CV, Engelman K. Pheochromocytoma—recognition and management. Curr Probl Cardiol 1979; 4:1. Kuchel O. Adrenal medulla: pheochromocytoma. In: Genest J, Kuchel O, Hamet P, Cantin M, eds. Hypertension, 2nd ed. New York: McGraw-Hill, 1983:947. Young JB, Landsberg L. Catecholamines and the adrenal medulla: pheochromocytoma. In: Wilson JD, Foster DW, eds. William's textbook of endocrinology, 9th ed. Philadelphia: WB Saunders, 1998:chap 13, 665. Hull CJ. Phaeochromocytoma: diagnosis, preoperative preparation and anaesthetic management. Br J Anaesthiol 1986; 58:1453. Samaan NA, Hickey RC. Pheochromocytoma. Semin Oncol 1987; 14:297. Sarosi G, Doe RP. Familial occurrence of parathyroid adenomas, pheochromocytoma and medullary carcinoma of the thyroid with amyloid stroma (Sipples syndrome). Ann Intern Med 1968; 68:1305. Neumann HPH, Berger DP, Sigmund G, et al. Pheochromocytomas, multiple endocrine neoplasia type 2, and von Hippel-Lindau disease. N Engl J Med 1993; 329:1531.

9a. Sgambati MT, Stolle C, Choyke PL, et al. Mosaicism in von Hippel-Lindau disease: lessons from kindreds with germline mutations identified in offspring with mosaic parents. Am J Hum Genet 2000; 66:84. 10. Carney JA, Go VLW, Gordon H, et al. Familial pheochromocytoma and islet cell tumor of the pancreas. Am J Med 1980; 68:515. 11. Lairmore TC, Ball DW, Baylin SB, Wells SA Jr. Management of pheochromocytomas in patients with multiple endocrine neoplasia type 2 syndromes. Ann Surg 1993; 217:595. 12. Plouin PF, Degoulet P, Tugaye A, et al. Le dépistage du phéochromocytome: chez quels hypertendus?: Étude sémiologique chez 2585 hypertendus dont 11 ayant un phéochromocytome. Nouv Press Med 1981; 10:869. 13. Dunn FG, DeCarvalho JGR, Kem DC, et al. Pheochromocytoma crisis induced by saralasin: relation of angiotensin analogue to catecholamine release. N Engl J Med 1976; 295:605. 14. Plouin PF, Ménard J, Corvol P. Hypertensive crisis in patient with pheochromocytoma given metoclopramide. Lancet 1976; 2:1357. 15. Bergland BE. Pheochromocytoma presenting as shock. Am J Emerg Med 1989; 7:44. 16. Montastruc JL, Chamontin B, Senard JM, et al. Pseudopheochromocytoma in parkinsonian patient treated with fluoxetine plus selegiline. Lancet 1993; 341:555. 17. Li JKY, Yeung VTF, Leung CM, et al. Clozapine: a mimicry of phaeochromocytoma. Aust N Z J Psychiatry 1997; 31:889. 18. Heath H III, Edis AJ. Pheochromocytoma associated with hypercalcemia and ectopic secretion of calcitonin. Ann Intern Med 1979; 91:208. 19. Forman BH, Marban E, Kayne RD, et al. Ectopic ACTH syndrome due to pheochromocytoma: case report and review of the literature. Yale J Biol Med 1979; 52:181. 20. Kaplan NM, Kramer NJ, Holland OB, et al. Single-voided urine metanephrine assays in screening for pheochromocytoma. Arch Intern Med 1977; 137:190. 21. Ganguly A, Henry DP, Yune HY, et al. Diagnosis and localization of pheochromocytoma: detection by measurement of urinary norepinephrine excretion during sleep, plasma norepinephrine concentration and computerized axial tomography (CT scan). Am J Med 1979; 67:21. 22. Duncan MW, Compton P, Lazarus L, Smythe GA. Measurement of norepinephrine and 3,4-dihydroxyphenylglycol in urine and plasma for the diagnosis of pheochromocytoma. N Engl J Med 1988; 319:136. 23. Gitlow SE, Mendlowitz M, Wilk EK, et al. Excretion of catecholamine metabolites by normal children. J Lab Clin Med 1968; 72:612. 24. Stein PP, Black HR. A simplified diagnostic approach to pheochromocytoma: a review of the literature and report of one institution's experience. Medicine (Baltimore) 1991; 70:46. 25. Graham PE, Smythe GA, Edwards GA, Lazarus L. Laboratory diagnosis of pheochromocytoma: which analytes should we measure? Ann Clin Biochem 1993; 30:129. 26. Hamilton BP, Landsberg L, Levine RJ. Measurement of urinary epinephrine in screening for pheochromocytoma in multiple endocrine neoplasia type II. Am J Med 1978; 65:1027. 27. Shoup RE, Kissinger PT, Goldstein DS. Rapid liquid chromatographic methods for assay of norepinephrine, epinephrine, and dopamine in biological fluids and tissues. In: Ziegler MG, Lake CR, eds. Frontiers of clinical neuro-science, vol 2. Norepinephrine. Baltimore: Williams & Wilkins, 1984:38. 28. Goldstein DS, Stull R, Markey SP, et al. Dihydrocaffeic acid: a common contaminant in the liquid chromatographic electrochemical measurement of plasma catecholamines in man. J Chromatogr 1984; 311:148. 29. Brown MJ. Simultaneous assay of norepinephrine and its deaminated metabolite, dihydroxyphenylglycol, in plasma: a simplified approach to the exclusion of pheochromocytoma in patients with borderline elevation of plasma noradrenaline concentration. Eur J Clin Invest 1984; 14:67. 30. Lenders JWM, Willemsen JJ, Beissel T, et al. Value of the plasma norepi-nephrine/3,4-dihydroxyphenylglycol ratio for the diagnosis of pheochromocytoma. Am J Med 1992; 92:147. 31. Lenders JWM, Keiser HR, Goldstein DS, et al. Plasma metanephrines in the diagnosis of pheochromocytoma. Ann Intern Med 1995; 123:101. 31a. Eisenhofer G, Lenders JW, Linehan WM, et al. Plasma normetanephrine and metanephrine for detecting pheochromocytoma in von Hippel-Lindau disease and multiple endocrine neoplasia type 2. N Engl J Med 1999; 340:1872.

32. Eisenhofer G, Keiser H, Friberg P, et al. Plasma metanephrines are markers of pheochromocytoma produced by catechol-O-methyltransferase within tumors. J Clin Endocrinol Metab 1998; 83:2175. 33. O'Connor DT, Bernstein KN. Radioimmunoassay of chromogranin A in plasma as a measure of exocytotic sympathoadrenal activity in normal subjects and patients with pheochromocytoma. N Engl J Med 1984; 311:764. 34. Canale MP, Bravo EL. Diagnostic specificity of serum chromogranin-A for pheochromocytoma in patients with renal dysfunction. J Clin Endocrinol Metab 1994; 78:1139. 35. Baudin E, Gigliotti A, Ducreux M, et al. Neuron-specific enolase and chromogranin A as markers of neuroendocrine tumours. Br J Cancer 1998; 78:1102. 36. Bravo EL, Tarazi RC, Fouad FM, et al. Clonidine suppression test: a useful aid in the diagnosis of pheochromocytoma. N Engl J Med 1981; 305:623. 37. Sheps SG, Maher FT. Histamine and glucagon tests in diagnosis of pheochromocytoma. JAMA 1968; 205:895. 38. Levinson PD, Hamilton BP, Mersey JH, Kowarski AA. Plasma norepinephrine and epinephrine responses to glucagon in patients with suspected pheochromocytomas. Metabolism 1983; 32:998. 39. Grossman E, Goldstein DS, Hoffman A, Keiser HR. Glucagon and clonidine testing in the diagnosis of pheochromocytoma. Hypertension 1991; 17:733. 40. Macdougall IC, Isles CG, Stewart H, et al. Overnight clonidine suppression test in the diagnosis and exclusion of pheochromocytoma. Am J Med 1988; 84:993. 41. Raper AJ, Jessee EF, Texter JH Jr, et al. Pheochromocytoma of the urinary bladder: a broad clinical spectrum. Am J Cardiol 1977; 40:820. 42. Jalil ND, Pattou FN, Combemale F, et al. Effectiveness and limits of preoperative imaging studies for the localization of pheochromocytomas and paragangliomas: a review of 282 cases. French Association of Surgery (AFC), and The French Association of Endocrine Surgeons (AFCE). Eur J Surg 1998; 164:23. 43. Fink IJ, Reinig JW, Dwyer AJ, et al. MR imaging of pheochromocytomas. J Comput Assist Tomogr 1985; 9:454. 44. Doppman JL, Reinig JW, Dwyer AJ, et al. Differentiation of adrenal masses by magnetic resonance imaging. Surgery 1987; 102:1018. 45. Sisson JC, Frager MS, Valk TW, et al. Scintigraphic localization of pheochromocytoma. N Engl J Med 1981; 305:12. 45a. Rainis T, Ben-Haim S, Dickstein G. False positive metaiodobenzylguanidine scan in a patient with a huge adrenocortical carcinoma. J Clin Endocrinol Metab 2000; 85:5. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81.

Khafagi FA, Shapiro B, Fig LM, et al. Labetolol reduces iodine-131 MIBG uptake by pheochromocytoma and normal tissues. J Nucl Med 1989; 30:481. Jonsson A, Hallengren B, Manhem P, et al. Cardiac pheochromocytoma. J Intern Med 1994; 236:93. Sinclair AJ, Bomanji J, Harris P, et al. Pre- and post-treatment distribution pattern of 123 I-MIBG in patients with phaeochromocytomas and paragangliomas. Nucl Med Commun 1989; 10:567. Kopf D, Bockisch A, Steinert H, et al. Octreotide scintigraphy and catecholamine response to an octreotide challenge in malignant pheochromocytoma. Clin Endocrinol (Oxf) 1997; 46:39. Rosamond TL, Hamburg MS, Vacek JL, Borkon AM. Intrapericardial pheochromocytoma. Am J Cardiol 1992; 70:700. Cubeddu L, Zarate NA, Rosales CB, Zschaeck DW. Prazosin and propranolol in preoperative management of pheochromocytoma. Clin Pharmacol Ther 1982; 32:156. Engelman K. Pheochromocytoma. Clin Endocrinol Metab 1977; 6:769. Perry RR, Keiser HR, Norton JA, et al. Surgical management of pheochromocytoma with the use of metyrosine. Ann Surg 1990; 212:621. Lenders JW, Sluiter HE, Thien T, Willemsen J. Treatment of a pheochromocytoma of the urinary bladder with nifedipine. BMJ 1985; 290:1624. Chimori K, Miyazaki S, Nakajima T, Miura K. Preoperative management of pheochromocytoma with the calcium antagonist nifedipine. Clin Ther 1985; 7:372. Proye C, Thevenin D, Cecat P, et al. Exclusive use of calcium channel blockers in preoperative and intraoperative control of pheochromocytomas: hemodynamics and free catecholamine assays in ten consecutive patients. Surgery 1989; 106:1149. Schenker JG, Chowers I. Pheochromocytoma and pregnancy. Obstet Gynecol Surg 1971; 26:739. Fudge TL, McKinnin WMP, Geary WL. Current surgical management of pheochromocytoma during pregnancy. Arch Surg 1980; 115:1224. Harper MA, Murnaghan GA, Kennedy L, et al. Phaeochromocytoma in pregnancy: five cases and a review of the literature. Br J Obstet Gynaecol 1989; 96:594. Demeure MJ, Carlsen B, Traul D, et al. Laparoscopic removal of a right adrenal pheochromocytoma in a pregnant woman. J Laparoendosc Adv Surg Tech A 1998; 8:315. Brennan MF, Keiser HR. Persistent and recurrent pheochromocytoma: the role of surgery. World J Surg 1982; 6:397. Oishi S, Sato T. Elevated serum neuron-specific enolase in patients with malignant pheochromocytoma. Cancer 1988; 61:1167. Linnoila RI, Lack EE, Steinberg SM, Keiser HR. Decreased expression of neuropeptides in malignant paragangliomas. Hum Pathol 1988; 19:41. Pang LC, Tsao KC. Flow cytometric DNA analysis for the determination of malignant potential in adrenal and extraadrenal pheochromocytomas or paragangliomas. Arch Pathol Lab Med 1993; 117:1142. Loh KC, Fitzgerald PA, Matthay KK, et al. The treatment of malignant pheochromocytoma with iodine-131 metaiodobenzylguanidine (131 I-MIBG): a comprehensive review of 116 reported patients. J Endocrinol Invest 1997; 20:648. Averbuch SD, Steakley CS, Young RC, et al. Malignant pheochromocytoma: effective treatment with a combination of cyclophosphamide, vincristine, and dacarbazine. Ann Intern Med 1988; 109:267. DeLellis RA, Wolfe HJ, Gagel RF, et al. Adrenal medullary hyperplasia: a morphometric analysis in patients with familial medullary thyroid carcinoma. Am J Pathol 1976; 83:177. Gitlow SE, Bertani LM, Rausen A, et al. Diagnosis of neuroblastoma by qualitative and quantitative determination of catecholamine metabolites in urine. Cancer 1970; 25:1377. Weinblatt ME, Heisel MA, Siegal SE. Hypertension in children with neurogenic tumors. Pediatrics 1983; 71:947. Conference on the biology of neuroblastoma. J Pediatr Surg 1968; 3:103. Consensus statement on the definition of orthostatic hypotension, pure autonomic failure, and multiple system atrophy: the Consensus Committee of the American Autonomic Society and the American Academy of Neurology. Neurology 1996; 46:1470. Goldstein DS, Holmes C, Cannon RO, et al. Sympathetic cardioneuropathy in dysautonomias. N Engl J Med 1997; 336:696. Wieling W, van Lieshout JJ. Investigation and treatment of autonomic circulatory failure. Curr Opin Neurol Neurosurg 1993; 6:537. Freeman R, Miyawaki E. The treatment of autonomic dysfunction. J Clin Neurophysiol 1993; 10:61. Chobanian AV, Volicer L, Tifft CP, et al. Mineralocorticoid-induced hypertension in patients with orthostatic hypotension. N Engl J Med 1979; 301:68. Jordan J, Shannon JR, Biaggioni I, et al. Contrasting actions of pressor agents in severe autonomic failure. Am J Med 1998; 105:166. Itskovitz HD, Wartenburg A. Combined phenylephrine and tranylcypromine for postural hypotension. Am Heart J 1983; 106:598. Moss AJ, Glaser W, Topol E. Atrial tachypacing in the treatment of a patient with primary orthostatic hypotension. N Engl J Med 1980; 302:1456. Polinsky RJ, Samaras GM, Kopin IJ. Sympathetic neural prosthesis for managing orthostatic hypotension. Lancet 1983; 1:901. Brant PW, McKusick VA. Familial dysautonomia: a report of genetic and clinical studies with a review of the literature. Medicine 1970; 49:343. Eng CM, Slaugenhaupt SA, Blumenfeld A, et al. Prenatal diagnosis of familial dysautonomia by analysis of linked CA-repeat polymorphisms on chromosome 9q31-q33. Am J Med Genet 1995; 59:349.

CHAPTER 87 ADRENOMEDULLARY DISORDERS OF INFANCY AND CHILDHOOD Principles and Practice of Endocrinology and Metabolism

CHAPTER 87 ADRENOMEDULLARY DISORDERS OF INFANCY AND CHILDHOOD WELLINGTON HUNG Catecholamines in the Infant and Child Pheochromocytoma Clinical Features Laboratory Findings Localization Familial Pheochromocytoma Differential Diagnosis Treatment Postoperative Care Adrenal Medullary Hyperplasia Chapter References

CATECHOLAMINES IN THE INFANT AND CHILD The physiology of the adrenal medulla has been discussed in Chapter 85. The normal values in pediatric patients for excretion of urinary catecholamines and their metabolites are shown in Table 87-1 and Table 87-2. Values have been published according to age, body weight, and surface area, and in relation to milligrams of urinary creatinine.1,2,3,4,5 and 6 In children, the daily urinary excretion of catecholamines and metabolites increases with age and is independent of the size of individuals.1 No sex difference is found. The dietary content does not significantly alter the quantity of catecholamines, vanillylmandelic acid, homovanillic acid, or metanephrine excreted in the urine.7

TABLE 87-1. Urinary Catecholamine Excretion in Full-Term and Premature Neonates on Day 1 of Age (Mean ± Standard Deviation)

TABLE 87-2. 24-Hour Urinary Excretion of Catecholamines and Metabolites in Normal Children by Age (Mean ± Standard Deviation)

Plasma epinephrine and metanephrine have been studied in normal infants and children using radioenzymatic assay. The concentrations fall rapidly during the first few minutes after birth, remain at this low level for the subsequent 3 hours, and then decline further by 12 to 48 hours of life.8 At 48 hours of life and later, the plasma epinephrine level (mean, 26 pg/mL) and the norepinephrine level (mean, 283 pg/mL) are comparable to values in the resting adult.

PHEOCHROMOCYTOMA Less than 5% of all pheochromocytoma cases occur in childhood. Pheochromocytoma is approximately twice as common in boys as in girls. In children, most tumors occur in the adrenal medulla, but they may also be found in aberrant tissue along the sympathetic chain, the thorax, the paraaortic area, the aortic bifurcation, the retroperitoneum, and the bladder. In children with pheochromocytoma, the incidence of malignant adrenal tumors has been reported to be as high as 25%.10 CLINICAL FEATURES The symptoms and signs of pheochromocytoma in children are presented in Table 87-3. Pheochromocytomas in pediatric patients are extremely variable in clinical presentation, which can lead to delay in diagnosis.11 In a large review of 100 children, 140 tumors were found.12 Sixty-eight patients had single tumors, 19 of which were extraadrenal. Among the 32 patients with two or more tumors, 20 had bilateral adrenal tumors, 8 had both intraadrenal and extraadrenal tumors, and 4 had multiple extraadrenal tumors.

TABLE 87-3. Symptoms and Signs of Pheochromocytoma in Children

The association of pheochromocytoma with neurocutaneous syndromes is well known. These syndromes include mucosal neuromas (see Chap. 188) and neurofibromatosis. Bilateral pheochromocytoma may occur in patients with von Hippel-Lindau disease (a syndrome characterized by dominantly inherited angiomatosis of the retina, cerebellar angioma, and angiomas of other organs).13 Extraadrenal pheochromocytoma may be found in patients with the triad of Carney (gastric epithelioid leiomyosarcoma, pulmonary chondroma, and functioning extraadrenal paraganglioma).14 LABORATORY FINDINGS As in adults, in children the definitive diagnosis of pheochromocytoma requires the detection of elevated urine or blood levels of catecholamines and their metabolites (Fig. 87-1). Generally, discontinuing all medications at least 2 weeks before obtaining urine collections is wise.

FIGURE 87-1. Serum and urinary constituents used for the diagnosis of pheochromocytoma. Norepinephrine and epinephrine are converted to normetanephrine and metanephrine, respectively, by the enzyme catechol-O-methyltransferase. Normetanephrine and metanephrine are then converted to vanillylmandelic acid by the enzyme monamine oxidase. (See Chap. 85 for details.)

The most commonly used diagnostic tests in children are 24-hour urine determinations of free catecholamines, vanillylmandelic acid, and metanephrines.15 Assays of urinary catecholamines and vanillylmandelic acid have been associated with an approximate 25% incidence of false-negative findings, whereas such results occur in only 4% of metanephrine determinations. Measurement of plasma catecholamines may be a useful adjunct to 24-hour urine studies.15a Patients must remain supine while blood samples are obtained; nevertheless, plasma determinations offer the major advantage of obviating 24-hour urine collections, which can be difficult in young children. Routine laboratory findings may include an elevated hematocrit, which may be attributable to decreased plasma volume. Hyperglycemia may be present, resulting from decreased insulin release and increased gluconeogenesis. Elevated plasma renin activity (PRA) and aldosterone levels may also occur.16 Formerly, pharmacologic tests to aid in the diagnosis were used widely; however, they are rarely indicated at the present time.17 Although the provocative tests are unnecessary and can be dangerous, the clonidine suppression test is a safe and accurate method for confirming the presence of a pheochromocytoma.18 LOCALIZATION Once the diagnosis of pheochromocytoma has been established, anatomic localization is essential. In children, the use of ultrasonography, computed tomography, and magnetic resonance imaging has essentially eliminated the need for preoperative localization of the tumor with arteriographic studies19 (see Chap. 88). Selective caval sampling for catecholamine levels can be performed when an extraadrenal tumor is suspected but cannot be demonstrated by other techniques. Metaiodobenzylguanidine (MIBG) labeled with iodine-131 has been used routinely as a diagnostic imaging agent in children to localize pheochromocytomas.20 Imaging cannot distinguish between benign and malignant tumors.21 An occasional pheochromocytoma may not be revealed with MIBG imaging, and [18F]fluoro-deoxyglucose positron emission tomography has been used successfully in detecting these tumors.22 FAMILIAL PHEOCHROMOCYTOMA Approximately 10% of pheochromocytomas in pediatric patients are familial, a frequency four times that in adults.23 Testing all members of a family for pheochromocytomas is, therefore, important. Pheochromocytomas have also been associated with medullary thyroid carcinoma and multiple endocrine neoplasia types 2A and 2B. Therefore, all patients with pheochromocytomas should undergo careful palpation of the thyroid gland, and serum calcitonin levels should be determined. However, preoperatively, the serum calcitonin concentration can be elevated due to production of calcitonin by the pheochromocytoma (see Chap. 53, Chap. 86 and Chap. 188). DIFFERENTIAL DIAGNOSIS Particularly in children, pheochromocytoma is a difficult disease to detect by history and physical examination alone. Hyperthyroidism, cardiac disease, diabetes mellitus, and anxiety reaction may be considered initially in the differential diagnosis, but appropriate laboratory results exclude them. TREATMENT Surgical removal of the pheochromocytoma should be undertaken after adequate preoperative preparation, which should include the administration of a-adrenergic and, sometimes, b-adrenergic blocking agents.24 The type and dosage of blockade vary from patient to patient and depend on the type of catecholamines produced by the tumor and on the response to the therapeutic agents. For a-adrenergic blockade, phenoxybenzamine hydrochloride (Dibenzyline) or phentolamine mesylate (Regitine) can be used. Phenoxybenzamine is superior because it offers a longer duration of action and smoother control and has fewer side effects than phentolamine. In children, the usual dosage is 20 to 50 mg twice a day for 10 to 14 days preoperatively (or 1–2 mg/kg per day divided, every 6–8 hours). The dosage of these and other a-adrenergic blocking drugs that is necessary to achieve the desired degree of blockade must be determined by careful titration while the patient is hospitalized. Propranolol may be used for the control of b-adrenergic effects, but only after prior a-blockade. The appropriate oral dosage is 10 to 30 mg three to four times daily. Drugs such as a-methyltyrosine that inhibit the synthesis of catecholamines have also been used in the treatment of pheochromocytoma. These agents block the conversion of tyrosine to dihydroxyphenylalanine (DOPA), thereby preventing the symptoms caused by an excess production of catecholamines. Children with pheochromocytoma may have an increased red cell mass and chronic hypovolemia. Prolonged adrenergic activity and vasoconstriction are associated with the hypovolemia. The hypovolemia need not be manifested by a decreased hematocrit. Preoperative treatment with phenoxy-benzamine for several days gradually causes expansion of the intravascular volume. Preoperative blood volume studies can be obtained, and blood can be administered in amounts calculated to expand the intravascular volume to 10% above normal.25 Laparoscopic adrenalectomy has been used in adults but infrequently in pediatric patients.25a With greater experience this technique will probably be used more often in children. In children, as in adults, the preoperative, operative, and postoperative management is of extreme importance.26 If bilateral pheochromocytomas are diagnosed preoperatively, the child should receive intramuscular glucocorticoid therapy for 3 days before surgery in anticipation of bilateral adrenalectomy. POSTOPERATIVE CARE Postoperative management should include monitoring the patient for the persistence of hypertension, which suggests residual pheochromocytoma or renovascular

damage. The clinician should be alert for a contracted vascular volume, which may be present despite preoperative a-adrenergic blockade. Hypoglycemia may occur after removal of a pheochromocytoma; therefore, the plasma glucose concentration should also be monitored. Urinary catecholamines and their metabolites should be measured during the first postoperative week. These should normalize within that time unless remaining tumor is present. Postoperatively, all patients diagnosed as having pheochromocytoma should undergo follow-up evaluations. At 6-month intervals, urinary or plasma catecholamine and metabolite studies are performed to detect any recurrence or, if the lesion was malignant, any active metastases. Rarely, a seemingly benign lesion may reveal its true nature by the subsequent finding of metastases years later; therefore, long-term follow-up is essential.27

ADRENAL MEDULLARY HYPERPLASIA Adrenal medullary hyperplasia is a rare disorder.28 It has been described in children who were diagnosed clinically as having pheochromocytoma but in whom no tumor could be found on surgical exploration or at autopsy. The criteria that have been proposed for the diagnosis of adrenal medullary hyperplasia include (a) a clinical history of episodic hypertension with other symptoms and signs suggesting pheochromocytoma, generally associated with increased urinary catecholamine levels during attacks; (b) diffuse expansion of the adrenal medulla into the tail of the adrenal gland; (c) a medulla composed of enlarged cells with or without pleomorphism; and (d) an increased medulla/cortex ratio, together with an increased medullary weight. The suggestion has been made that diffuse adrenal medullary hyperplasia may be the initial pathologic change that subsequently leads to the development of nodular hyperplasia and a pheochromocytoma.29 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

De Schaepdryver AF, Hooft C, Delbeke MJ, Van Den Noortgaete M. Urinary catecholamines and metabolites in children. J Pediatr 1978; 93:266. Voorhess ML. Urinary catecholamine excretion by healthy children. Pediatrics 1967; 39:252. Gitlow SE, Mendlowitz M, Wilk EK, et al. Excretion of catecholamine catabolites by normal children. J Lab Clin Med 1968; 72:612. Nakai T, Yamanda R. Urinary catecholamine excretion by various age groups with special reference to clinical value of measuring catecholamines in newborns. Pediatr Res 1983; 17:456. Maxwell GM, Crompton S, Davies A. Urinary catecholamine levels in the newborn infant. Eur J Pediatr 1985; 143:171. Tuchman M, Morris CL, Ramnaraine ML, et al. Value of random urinary homovanillic acid and vanillylmandelic acid levels in the diagnosis and management of patients with neuroblastoma: comparison with 234-hour urine collections. Pediatrics 1985; 75:324. Weetman RM, Rider PS, Oei TO, et al. Effect of diet on urinary excretion of VMA, HVA, metanephrine and total free catecholamine in normal preschool children. J Pediatr 1976; 88:46. Elito RJ, Lam R, Leake RD, et al. Plasma catecholamine concentration in infants at birth and during the first 48 hours of life. J Pediatr 1980; 96:311. Hume DM. Pheochromocytoma in the adult and in the child. Am J Surg 1960; 99:458. Perel Y, Schlumberger M, Marguerite G, et al. Pheochromocytoma and paraganglioma in children: a report of 24 cases of the French Society of Pediatric Oncology. Pediatr Hematol Oncol 1997; 14:413. Januszewicz P, Wieteska-Klimczak, Wysznka T. Pheochromocytoma in children: difficulties in diagnosis and localization. Clin Exp Hypertens 1990; 4:571. Stackpole RH, Melicow MM, Uson AC. Pheochromocytoma in children: report of nine cases and review of the first 100 published cases with follow-up studies. J Pediatr 1963; 63:315. Ritter MM, Frilling A, Crossey PA, et al. Isolated familial pheochromocytoma as a variant of von Hippel-Lindau disease. J Clin Endocrinol Metab 1996; 81:1035. Carney JA. The triad of gastric epithelioid leiomyosarcoma, pulmonary chondroma, and functioning extra-adrenal paraganglioma. Medicine 1983; 62:159. Fonkalsrud EW. Pheochromocytoma in childhood. Prog Pediatr Surg 1991; 26:103.

15a.Raber W, Raffesberg W, Kmen E, et al. Pheochromocytoma with normal urinary and plasma catecholamines but elevated plasma free metanephrines in a patient with adrenal incidentaloma. The Endocrinologist 2000; 10:65. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.

Hung W, August GP. Hyperreninemia and secondary hyperaldosteronism in pheochromocytoma. J Pediatr 1979; 94:215. Young WF Jr. Pheochromocytoma and primary aldosteronism: diagnostic approaches. Endocrinol Metab Clinics North Am 1997; 26:801. Bravo EL, Tarazi RC, Fouad FM, et al. Clonidine-suppression test: a useful aid in the diagnosis of pheochromocytoma. N Engl J Med 1981; 305:623. Farrelly CA, Daneman A, Martin DJ, Chan HSL. Pheochromocytoma in childhood: the important role of computed tomography in tumor localization. Pediatr Radiol 1984; 13:210. Gelfand MJ. Metaiodobenzylguanidine in children. Semin Nucl Med 1993; 23:231. Abramson SJ. Adrenal neoplasms in children. Radiol Clin North Am 1997; 35:1415. Arnold DR, Villemagne VL, Civelek AC, et al. FDG-PET: a sensitive tool for the localization of MIBG-negative pelvic pheochromocytomas. Endocrinologist 1998; 8:295. Levine C, Skimming J, Levine E. Familial pheochromocytomas with unusual associations. J Pediatr Surg 1992; 27:447. Brunjes A, Johns VJ, Crane MG. Pheochromocytoma. N Engl J Med 1960; 262:393. Schwartz DL, Gann DS, Haller JA. Endocrine surgery in children. Surg Clin North Am 1974: 54:363.

25a.Clements RH, Goldstein RE, Holcomb III GW. Laparoscopic left adrenalectomy for pheochromocytoma in a child. J Pediatr Surg 1999; 34:1408. 26. 27. 28. 29.

Turner MC, Lieberman E, DeQuattro V. The perioperative management of pheochromocytomas in children. Clin Pediatr 1992; 31:583. Em SH, Shandling B, Wesson D, Filler RM. Recurrent pheochromocytomas in children. J Pediatr Surg 1990; 10:1063. Kurihara K, Mizuseki K, Kondo T, et al. Adrenal medullary hyperplasia. Hyperplasia-pheochromocytoma sequence. Acta Pathol Jpn 1990; 40:683. Qupty G, Ishay A, Peretz H, et al. Pheochromocytoma due to unilateral adrenal medullary hyperplasia. Clin Endocrinol 1997; 47:613.

CHAPTER 88 DIAGNOSTIC IMAGING OF THE ADRENAL GLANDS Principles and Practice of Endocrinology and Metabolism

CHAPTER 88 DIAGNOSTIC IMAGING OF THE ADRENAL GLANDS DONALD L. MILLER Adrenal Adenomas Adrenal Lesions with Endocrine Function Pheochromocytoma Hyperaldosteronism Cushing Syndrome Adrenal Hypofunction Chapter References

Current imaging techniques permit the adrenal gland to be visualized with superb clarity and spatial resolution. Except in rare circumstances, other methods of adrenal imaging have been supplanted by computed tomography (CT) and magnetic resonance imaging (MRI). CT should be the initial study for adrenal imaging in virtually all patients. It is capable of demonstrating the adrenal glands in virtually 100% of normal individuals. It provides greater spatial resolution than MRI and is less expensive. Ultrasonography costs less than CT, but it is operator dependent, has a high false-negative rate, and is often unable to permit imaging of the left adrenal gland. CT is more accurate than ultra-sonography and can demonstrate both normal adrenal glands in virtually all patients (Fig. 88-1). MRI is more helpful for differential diagnosis of a known adrenal mass, but CT is a more appropriate screening technique. Oral and intravenous contrast materials are not normally required but can be helpful in some very thin patients.

FIGURE 88-1. Normal adrenal glands in a patient with abundant retro-peritoneal fat. The right gland lies directly behind the inferior vena cava (C) and consists of medial and lateral limbs. On the right, a small nodule (arrow) is seen on the tip of the medial limb, not an uncommon finding in middle-aged patients.

The limbs of a normal adrenal gland vary considerably in length and width from individual to individual. The usual length is from 4 to 6 cm, and the usual width is from 2 to 3 mm. A good rule of thumb, when interpreting CT or MRI examinations, is that each limb should be no wider than the ipsilateral crus of the diaphragm.1

ADRENAL ADENOMAS Incidentally discovered adrenal adenomas (incidentalomas) are found in 1% to 4% of the population on CT scans.1 Considerable effort has been devoted to the development of CT and MRI techniques for differentiating benign adrenal tumors from other adrenal masses. Adrenal adenomas are round or oval, homogeneous masses with a smooth, well-defined margin, and are usually 3 cm in diameter or that have irregular margins have a high probability of malignancy.3 Biopsy should be performed.

FIGURE 88-2. Adrenal adenoma. A homogeneous, oval, 2.5-cm mass with smooth margins is seen in the right adrenal gland (arrow). On this noncontrast CT scan, its attenuation is –3 Hounsfield units. It was an incidental finding in this 40-year-old woman.

In patients who do not have a known malignancy, an incidentally discovered adrenal mass >5 cm may be an adrenal cancer. This size criterion had a sensitivity of 93% and a specificity of 64% for differentiating adenomas and carcinomas in a series of 210 patients with incidentally discovered adrenal masses.4 Other features suggestive of adrenal carcinoma are central areas of decreased attenuation, calcification, and evidence of hepatic, venous, or nodal spread (Fig. 88-3).5 Note, however, that large, degenerated adrenal adenomas often contain areas of calcification, hemorrhage, or necrosis.5a No universally reliable imaging criteria, other than size, are available for differentiating adrenal adenoma from adrenal carcinoma on CT scans.

FIGURE 88-3. Adrenal carcinoma. A, The 9-cm mass in the left adrenal gland (arrows) has low signal intensity on this coronal T2-weighted magnetic resonance image and shows some central inhomogeneity. Whatever its imaging characteristics, the tumor's size alone suggests that adrenal carcinoma should be strongly considered.

B, A more anterior coronal image demonstrates paraaortic adenopathy, which supports the diagnosis of adrenal carcinoma.

Adrenal adenomas, regardless of whether they demonstrate clinical endocrine function, usually contain an abundance of intracytoplasmic lipid; metastases do not.6 Most CT and MRI maneuvers used to diagnose adrenal adenomas are designed to demonstrate or quantify this lipid content. The simplest method is measurement of adrenal attenuation on a CT scan obtained without intravenous contrast material. CT attenuation is measured in Hounsfield units (HU), with water arbitrarily assigned a value of 0 HU. The greater the amount of lipid in an adrenal mass, the lower its attenuation. Some investigators have reported a mean attenuation of 2.5 HU for adrenal adenomas versus 32 HU for other adrenal lesions (see Fig. 88-2),7 and others have found a mean attenuation of 4 HU for adenomas and 37 HU for other adrenal masses.8 Unfortunately, substantial variability exists in the amount of lipid present from adenoma to adenoma; therefore, substantial variability is seen in attenuation as well. Threshold values from 20 HU to 2 HU have been used in the radiologic literature to distinguish between adenomas and other adrenal lesions. With a 20-HU threshold, sensitivity is 88%, but specificity is 84%. If a 2-HU threshold is used, specificity increases to 100%, but sensitivity is only 47%.9 Many centers use a 10-HU cutoff value.10 These threshold values cannot be used if the patient has been given intravenous contrast material for the CT scan, because this causes adrenal enhancement. Instead, attenuation may be measured at a specific time (³3 min) after contrast material administration. The attenuation threshold depends on the delay between contrast material administration and scanning. A threshold of 39 HU at 30 minutes has been recommended.8 An alternative CT method for evaluation of adrenal lesions is the use of washout curves. After intravenous administration of contrast material, adrenal adenomas demonstrate contrast material washout (loss of enhancement) earlier and more rapidly than other adrenal masses.11,12 The histopathologic and pathophysiologic correlates of this behavior are unclear. A similar phenomenon has been observed with MRI. Initial reports suggest that the sensitivity and specificity of this technique exceed 96%.11,12 MRI with chemical-shift imaging is another way to demonstrate adrenal lipid. This is generally done using opposed-phase images. T1-weighted gradient-echo MR images can be acquired so that water and lipid spins are either in phase or opposed. If lipid is present, comparison of in-phase and opposed-phase images will demonstrate that the signal intensity of the tissue is lower on the opposed-phase image (Fig. 88-4). Quantitative analysis may be performed by calculating the chemical-shift ratio—equal to the lesion-to-spleen intensity ratio in in-phase images divided by the lesion-to-spleen ratio in opposed-phase images (see Fig. 88-4). Eighty percent sensitivity and nearly 100% specificity can be achieved with this technique, regardless of whether qualitative or quantitative methods are used.5,13 In a study of 134 adrenal masses, the combination of MR chemical shift imaging and MR gadolinium washout techniques yielded a sensitivity of 91%, a specificity of 94%, and an accuracy of 93% for differentiating benign and malignant adrenal masses.13a

FIGURE 88-4. Magnetic resonance chemical-shift imaging in a patient with a left adrenal adenoma. A, An axial in-phase image demonstrates a left adrenal mass (arrow) of signal intensity approximately equal to that of the spleen (S). B, An opposed-phase image demonstrates markedly lower signal intensity in the same adrenal mass (arrow) as compared to the spleen, which indicates that the mass has a significant lipid content.

MR chemical–shift imaging and CT-attenuation techniques correlate well; a lesion that is considered indeterminate by one technique is likely to be indeterminate with the other technique.14 This is not always true, however. A cost-effective algorithm uses CT attenuation as the first step, followed, if necessary, by MRI with chemical-shift imaging. If both of these studies are equivocal, biopsy should be performed.15 Clinically silent adrenal adenomas cannot be differentiated from functioning adrenal adenomas on the basis of any cross-sectional imaging method (i.e., CT, MRI, ultrasonography). This is an endocrine diagnosis, not a radiologic one, because imaging studies demonstrate morphology, not function. However, MRI signal intensity characteristics can usually aid in differentiating benign adrenal adenomas from otherwise morphologically identical pheochromocytomas, and often from small adrenal carcinomas.16 Adrenal scintigraphy with the cholesterol analog NP-59 (131I-6b-iodomethylnorcholesterol), an investigational drug, has been used to differentiate adrenal adenomas from adrenal metastases. Advances in CT and MRI have essentially eliminated the need for this agent, which accumulates in, and irradiates, the adrenal glands, the gonads, and the thyroid.1,17 111In-pentetreotide, an octreotide analog, may be helpful in problem cases because it is concentrated in most malignant adrenal lesions, but not in most benign lesions.18 Uncommon nonfunctioning adrenal lesions identified on imaging studies include cystic lesions (hydatid cyst, endothelial cyst), solid lesions (hemangioma, ganglio-neuroma, angiosarcoma, primary malignant melanoma), and solid fatty lesions (myelolipoma, collision tumor). Most of these lesions do not have specific imaging features.18a

ADRENAL LESIONS WITH ENDOCRINE FUNCTION The cardinal rule of endocrine radiology may be expressed in the phrase diagnosis first, localization second. When dealing with a patient with a suspected endocrine abnormality, the clinician must establish the diagnosis first, using endocrinologic methods. Only after the diagnosis is determined should radiologic methods be used in an attempt to locate an anatomic abnormality. Failure to heed this rule often results in a costly series of tests that produce only false-positive findings. PHEOCHROMOCYTOMA Adrenal pheochromocytomas are almost always >2 cm in diameter and are readily identified by CT. Central necrosis and calcification may occur (Fig. 88-5).2 Pheochromocytomas demonstrate very high signal intensities on T2-dependent spin-echo MRI images and are readily visible.5 In a series of 282 patients with known pheochromocytomas, MRI had a sensitivity of 98%, CT had a sensitivity of 89%, and scanning with 131I-metaiodoben-zylguanidine (MIBG) had a sensitivity of 81%.19 For this reason, MRI is the procedure of choice for the initial localization of pheochromocytomas.

FIGURE 88-5. A, This contrast-enhanced computed tomographic scan shows a 3-cm pheochromocytoma in the left adrenal gland that contains focal areas of necrosis

(arrows). B, A T2-weighted magnetic resonance imaging (MRI) scan shows very high signal intensity (arrowheads), typical for a pheochromocytoma. This characteristic high signal intensity on T2-weighted MRI scans makes MRI the preferred localization technique for ectopic pheochromocytomas.

Patients with elevated catecholamine levels and a unilateral adrenal mass on CT or MRI require no further localization studies. Bilateral tumors or adrenal medullary hyperplasia are common in multiple endocrine neoplasia syndrome (MEN) type 2A (Sipple syndrome) and type 2B. When pheochromocytoma is suspected, MRI of the adrenal glands should be performed first, because 90% of these tumors are intraadrenal. If the adrenal glands are normal, MRI or CT of the entire abdomen and pelvis is appropriate, because 98% of all pheochromocytomas and 85% of all extraadrenal pheochromocytomas occur below the diaphragm.20 Extraadrenal pheochromocytomas may occur in paravertebral locations, in the organ of Zuckerkandl, and anywhere from the base of the skull (e.g., glomus jugulare tumors) to the neck of the urinary bladder. MIBG imaging has proven useful for the detection of ectopic pheochromocytomas. MIBG has a molecular structure that resembles norepinephrine. It is actively concentrated in chromaffin tissue and the adrenergic nervous system. Although MIBG imaging is less sensitive overall than MRI or CT, it is more specific than either.21 For extraadrenal lesions, MIBG scanning has a sensitivity of 67% to 100%, and a specificity of 96%.20 It is particularly helpful for the detection of metastases in patients with malignant pheochromocytomas (Fig. 88-6). MIBG scans alone may not provide sufficient anatomic detail to guide the surgeon, but they help to direct CT and MRI studies toward the ectopically located lesion.

FIGURE 88-6. A, Iodine-131–labeled metaiodobenzylguanidine (MIBG) concentrates in pheochromocytomas, as demonstrated in the scan of this patient with bilateral adrenal tumors (arrows). This image was obtained 48 hours after injection of the radiopharmaceutical. The delay is a drawback of this method. B, In patients with malignant pheochromocytoma, a whole body scan may demonstrate metastases. A prominent lesion is seen in this patient's skull (arrow) and other metastases are noted elsewhere. Radioactivity is also evident in normal liver (open arrow) and the urinary bladder (curved arrow) on this anterior scan obtained 24 hours after administration of MIBG.

In the rare patient in whom CT, MRI, and MIBG studies yield negative or equivocal results, arteriography or venous sampling may sometimes be helpful.22 These procedures are safe in patients receiving adequate doses of blocking agents, such as phenoxybenzamine. Nonionic contrast agents do not appear to affect circulating catecholamine levels.23 HYPERALDOSTERONISM Hyperaldosteronism (Conn syndrome) may be caused by a discrete, functioning aldosterone-producing adenoma (APA) or by bilateral hyperplasia. The decision between medical and surgical therapy hinges on the diagnosis of unilateral adrenal involvement (adenoma) or bilateral hyperplasia.24 Because MRI is more expensive than CT and has a lower spatial resolution, CT is the preferred modality for initial evaluation of the adrenal glands in patients with hyperaldosteronism.17 Aldosterone-producing adenomas often appear on CT as relatively lucent, focal areas of decreased attenuation compared with the surrounding adrenal gland (Fig. 88-7).2 This presumably reflects the high concentration of corticosteroids in the lesion. APAs are small: 50% are 5 mm, but smaller APAs still may occasionally be missed. CT scans that demonstrate normal-appearing adrenal glands bilaterally are not diagnostic of hyperplasia, because a small, undetected adenoma may be present.25 One should also remember that on rare occasions the aldosterone source may be a renal or ovarian tumor.26

FIGURE 88-7. Aldosterone-producing adenomas are usually small and often contain relatively lucent focal areas of decreased attenuation (arrow).

A more common problem in differentiating adenoma from hyperplasia on CT and MRI is the presence of bilateral adrenal nodules. This finding is not diagnostic of idiopathic hyperplasia. In one series, 6 of 7 patients with bilateral adrenal nodules on CT had a unilateral aldosteronoma confirmed at surgery (Fig. 88-8).27 The same difficulty arises when multiple nodules are identified in one gland, or when CT evidence of bilateral adrenal enlargement and a unilateral nodule is present. With current high-definition CT scanners, multiple small nodules or limb thickening are often observed in the adrenal glands of older adults.28 Chemical-shift MRI can be used to define which, if any, of the nodules has a high lipid content and is therefore an APA.29 In one series, chemical shift imaging was positive in 6 of 7 patients (86%) with APA and 8 of 9 patients (89%) with hyperplasia.27

FIGURE 88-8. The bilateral adrenal masses (arrows) in this patient with hyperaldosteronism suggest the diagnosis of idiopathic hyperaldosteronism. Bilateral adrenal venous sampling revealed a right-sided aldosteronoma, confirmed by surgery. Aldosterone levels became normal after right adrenalectomy.

Patients with primary hyperaldosteronism who have imaging findings of an obvious unilateral adrenal nodule and a normal contralateral gland on CT scans, and clinical findings that support a diagnosis of APA, may proceed directly to surgery.1 Patients with CT evidence of bilateral adrenal nodules or bilateral normal-appearing adrenal glands should undergo bilateral adrenal vein sampling.17,24,29 Scintigraphy with NP-59 has been used in patients with hyperaldosteronism, but it is relatively insensitive for APAs 1.5 times the A:C ratio of the peripheral sample, and the A:C ratios of blood from the normal (suppressed) adrenal and the peripheral vein are essentially the same. With adrenal hyperplasia, samples from both glands exhibit A:C ratios of >1.5 times the A:C ratio from the peripheral vein. Because catheterization of the right adrenal vein may be difficult, some patients have been evaluated using only left adrenal vein and peripheral samples. This method is accurate if an APA is present in the right adrenal gland, but bilateral adrenal hyperplasia and a left adrenal APA cannot be differentiated without a right adrenal vein sample. CUSHING SYNDROME Bilateral adrenal hyperplasia secondary to ACTH-dependent Cushing disease can produce slightly enlarged glands; however, the adrenals may have a normal appearance. Paraneoplastic or ectopic production of ACTH may result in a greater degree of adrenal enlargement. Occasionally, unilateral macronodular hyperplasia may simulate the appearance of an adenoma (Fig. 88-9).28 This may lead the unwary physician to recommend inappropriate unilateral adrenalectomy.

FIGURE 88-9. A 3-cm nodule (black arrow) in the right adrenal gland of a patient with Cushing syndrome. The presence of hyperplasia in the left adrenal gland (white arrows) identifies this as adrenocorticotropic hormone (ACTH)–dependent hypercortisolism with a unilateral nodule (i.e., asymmetric ACTH-dependent macronodular hyperplasia).

Two forms of bilateral adrenal involvement in ACTH-independent Cushing syndrome are found. ACTH-independent macronodular adrenal hyperplasia, also called massive macronodular hyperplasia, is a rare, distinct cause of Cushing syndrome. 33 Approximately 40 cases had been reported in a recent review.33 Multiple adrenal nodules are present bilaterally, ranging in size from microscopic to 4 cm.28 They obscure the normal adrenal contour, and the adrenal gland is recognizable as such only by its location. These patients require bilateral adrenalectomy, which is curative.33 Primary pigmented nodular adrenocortical disease may be a component of Carney complex, an autosomal dominant MEN syndrome that includes myxomas, spotty skin pigmentation, and tumors of the adrenal cortex, pituitary, thyroid, or gonads.34,35 The pigmented nodules range from microscopic to 8 mm in diameter. Older patients may have larger nodules, up to 3 cm in diameter.36 In individuals aged 14 years or older, CT scans obtained with 5-mm sections demonstrate tiny nodules in the adrenal limbs, with intervening areas of atrophy, which give rise to a characteristic “string-of-beads” appearance.28,36 The lipofuscin pigment in the nodules does not affect their CT or MR appearance. ADRENAL HYPOFUNCTION In idiopathic Addison disease, the size of the adrenal glands may be small or normal. Adrenal calcification or high attenuation of the adrenal glands can be seen in hypoadrenalism owing to certain specific causes, such as chronic tuberculosis, histoplasmosis, and hemochromatosis (Fig. 88-10). In hypoadrenalism of acute onset, the identification of enlarged glands with a normal contour should suggest acute granulomatous adrenalitis secondary to tuberculosis or histoplasmosis. Prompt treatment may restore adrenal function. Rarely, hypoadrenalism may be due to bilateral adrenal metastases (Fig. 88-11).

FIGURE 88-10. A, In chronic adrenal insufficiency due to granulomatous infection, bilateral calcification is often present (arrow and arrowhead). B, Acute adrenal insufficiency due to histoplasmosis presents as enlarged inhomogeneous adrenal glands that retain their adreniform shape.

FIGURE 88-11. Bilateral adrenal lymphoma causing hypoadrenalism. In this 69-year-old man, both adrenal glands are enlarged, and paraaor-tic adenopathy (arrows) is also evident.

CHAPTER REFERENCES 1. 2. 3. 4. 5.

Cook DM. Adrenal mass. Endocrinol Metab Clin North Am 1997; 26:829. Kawashima A, Sandler CM, Fishman EK, et al. Spectrum of CT findings in nonmalignant disease of the adrenal gland. Radiographics 1998; 18:393. Candel AG, Gattuso P, Reyes CV, et al. Fine-needle aspiration biopsy of adrenal masses in patients with extraadrenal malignancy. Surgery 1993; 114:1132. Terzolo M, Ali A, Osella G, Mazza E. Prevalence of adrenal carcinoma among incidentally discovered adrenal masses: a retrospective study from 1989 to 1994. Arch Surg 1997; 132:914. Krebs TL, Wagner BJ. MR imaging of the adrenal gland: radiologic-pathologic correlation. Radiographics 1998; 18:1425.

5a. Newhouse JH, Heffess CS, Wagner BJ, et al. Large degenerated adrenal adenomas: radiologic-pathologic correlation. Radiology 1999; 210:385. 6. 7. 8. 9. 10. 11. 12. 13.

Korobkin M, Giordano TJ, Brodeur FJ, et al. Adrenal adenomas: relationship between histologic lipid and CT and MR findings. Radiology 1996; 200:743. Korobkin M, Brodeur FJ, Yutzy GG, et al. Differentiation of adrenal adenomas from nonadenomas using CT attenuation values. AJR Am J Roentgenol 1996; 166:531. Szolar DH, Kammerhuber F. Quantitative CT evaluation of adrenal gland masses: a step forward in the differentiation between adenomas and non-adenomas? Radiology 1997; 202:517. Boland GWL, Lee MJ, Gazelle GS, et al. Characterization of adrenal masses using unenhanced CT: an analysis of the CT literature. AJR Am J Roentgenol 1998; 171:201. Peppercorn PD, Grossman AB, Reznek RH. Imaging of incidentally discovered adrenal masses. Clin Endocrinol 1998; 48:379. Szolar DH, Kammerhuber FH. Adrenal adenomas and nonadenomas: assessment of washout at delayed contrast-enhanced CT. Radiology 1998; 207:369. Korobkin M, Brodeur FJ, Francis IR, et al. CT time-attenuation washout curves of adrenal adenomas and nonadenomas. AJR Am J Roentgenol 1998; 170:747. Outwater EK, Blasbalg R, Siegelman ES, Vala M. Detection of lipid in abdominal tissues with opposed-phase gradient-echo images at 1.5 T: techniques and diagnostic importance. Radiographics 1998; 18:1465.

13a. Heinz-Peer G, Hönigschnabl S, Schneider B, et al. Characterization of adrenal masses using MR imaging with histopathologic correlation. AJR Am J Roentgenol 1999; 173:15. 14. Outwater EK, Siegelman ES, Huang AB, Birnbaum BA. Adrenal masses: correlation between CT attenuation value and chemical shift ratio at MR imaging with in-phase and opposed-phase sequences. Radiology 1996; 200:749. 15. McNicholas MMJ, Lee MJ, Mayo-Smith WM, et al. An imaging algorithm for the differential diagnosis of adrenal adenomas and metastases. AJR Am J Roentgenol 1995; 165:1453. 16. Lee MJ, Mayo-Smith WW, Hahn PF, et al. State-of-the-art MR imaging of the adrenal gland. Radiographics 1994; 14:1015. 17. Young WF Jr. Pheochromocytoma and primary aldosteronism: diagnostic approaches. Endocrinol Metab Clin North Am 1997; 26:801. 18. Maurea S, Lastoria S, Salvatore M, et al. The role of radiolabeled somato-statin analogs in adrenal imaging. Nucl Med Biol 1996; 23:677. 18a. Otal P, Escourrou G, Mazerolles C, et al. Imaging features of uncommon adrenal masses with histopathologic correlation. Radiographics 1999; 19:589. 19. Jalil ND, Pattou FN, Combemale F, et al. Effectiveness and limits of preoperative imaging studies for the localisation of pheochromocytomas and paragangliomas: a review of 282 cases. Eur J Surg 1998; 164:23. 20. Whalen RK, Althausen AF, Daniels GH. Extra-adrenal pheochromocytoma. J Urol 1992; 147:1. 21. Maurea S, Cuocolo A, Reynolds JC, et al. Diagnostic imaging in patients with paragangliomas: computed tomography, magnetic resonance and MIBG scintigraphy comparison. Q J Nucl Med 1996; 40:365. 22. Walker IABL. Selective venous catheterization and plasma catecholamine analysis in the diagnosis of phaeochromocytoma. J R Soc Med 1996; 89:216P. 23. Mukherjee JJ, Peppercorn PD, Reznek RH, et al. Pheochromocytoma: effect of nonionic contrast medium in CT on circulating catecholamine levels. Radiology 1997; 202:227. 24. Ganguly A. Primary aldosteronism. N Engl J Med 1998; 339:1828. 25. Young WF Jr, Stanson AW, Grant CS, et al. Primary aldosteronism: adrenal venous sampling. Surgery 1996; 120:913. 26. Abdelhamid S, Müller-Lobeck H, Pahl S, et al. Prevalence of adrenal and extra-adrenal Conn syndrome in hypertensive patients. Arch Intern Med 1996; 156:1190. 27. Doppman JL, Gill JR Jr, Miller DL, et al. Distinction between hyperaldosteronism due to bilateral hyperplasia and unilateral aldosteronoma: reliability of CT. Radiology 1992; 184:677. 28. Doppman JL. The dilemma of bilateral adrenocortical nodularity in Conn's and Cushing's syndromes. Radiol Clin North Am 1993; 31:1039. 29. Doppman JL. Problems in endocrinologic imaging. Endocrinol Metab Clin North Am 1998; 26:973. 29a. Sohaib SA, Peppercorn PD, Allan C, et al. Primary hyperaldosteronism (Conn syndrome): MR imaging findings. Radiology 2000; 214:527. 30. Gleason PE, Weinberger MH, Pratt JH, et al. Evaluation of diagnostic tests in the differential diagnosis of primary aldosteronism: unilateral adenoma versus bilateral micronodular hyperplasia. J Urol 1993; 150:1365. 31. Doppman JL, Gill JR Jr. Hyperaldosteronism: sampling the adrenal veins. Radiology 1996; 198:309. 32. Tokunaga K, Nakamura H, Marukawa T, et al. Adrenal venous sampling analysis of primary aldosteronism: value of ACTH stimulation in the differentiation of adenoma and hyperplasia. Eur Radiol 1992; 2:223. 33. Swain JM, Grant CS, Schlinkert RT, et al. Corticotropin-independent macronodular adrenal hyperplasia: a clinicopathologic correlation. Arch Surg 1998; 133:541. 34. Stratakis CA, Kirschner LS. Clinical and genetic analysis of primary bilateral adrenal diseases (micro- and macronodular disease) leading to Cushing syndrome. Horm Metab Res 1998; 30:456. 35. Carney JA. The Carney complex (myxomas, spotty pigmentation, endocrine overactivity, and Schwannomas). Dermatol Clin 1995; 13:19. 36. Doppman JL, Travis WD, Nieman L, et al. Cushing syndrome due to primary pigmented nodular adrenocortical disease: findings at CT and MRI. Radiology 1989; 172:415.

CHAPTER 89 SURGERY OF THE ADRENAL GLANDS Principles and Practice of Endocrinology and Metabolism

CHAPTER 89 SURGERY OF THE ADRENAL GLANDS GARY R. PEPLINSKI AND JEFFREY A. NORTON Surgical Anatomy and Embryology Indications for Adrenalectomy Adrenal Incidentaloma Preoperative Patient Preparation Adrenalectomy Postoperative Care Operative Sequelae Postoperative Follow-Up Chapter References

Adrenalectomy is the most effective means available to eradicate localized cancers that arise within the adrenal glands, to eliminate adrenal sources of hormone overproduction, to relieve symptoms caused by mass effect, and to promptly establish definitive diagnoses for potentially malignant adrenal masses. Minimally invasive laparoscopic approaches for adrenalectomy have evolved dramatically and are now commonly practiced in high-volume endocrine surgery centers. Laparoscopic adrenalectomy can be safely performed by experienced surgeons, is effective treatment for selected adrenal conditions, and is consistently associated with prompt resumption of full patient activity. Open adrenalectomy remains the standard operation for resection of large adrenal tumors and adrenal cancers. In addition to surgical advances, success in treating patients with adrenal disorders is also due to improved understanding of endocrine pathophysiology, more sensitive and specific radiologic imaging and localization studies, and advances in anesthesia in perioperative patient care. This chapter presents the surgical perspective to assessing and treating the varied disorders of the adrenal glands.

SURGICAL ANATOMY AND EMBRYOLOGY Each adrenal gland lies high within the retroperitoneum in a central location within the body, near the midline and at the junction of the chest and abdomen. Careful dissection must be carried out near the inferior vena cava, kidneys, diaphragm, liver, aorta, splenic vessels, spleen, stomach, and pancreas. The normal gland is identified resting on the superior aspect of the kidney and is usually 3 to 5 cm long, 2 to 3 cm wide, 0.5 cm thick, and 3 to 6 g in weight. The right adrenal is triangular in shape and abuts the posterolateral surface of the inferior vena cava. The left adrenal gland lies close to the aorta and is more crescentic in shape. Each gland is surrounded by a fibrous capsule and embedded within areolar perirenal fat. The normal adrenal cortex is bright yellow, and the medulla appears reddish brown. On palpation of the suprarenal area, the adrenal gland can be distinctly recognized by its firm consistency. The adrenal glands are highly vascular. The arterial blood supply for each gland is variable and is derived from numerous sources. Several small arteries penetrate the perimeter of each gland along its superior, medial, and inferior aspects, arising from the phrenic artery superiorly, directly from the aorta medially, and from the renal arteries inferiorly. Venous drainage of the adrenals is more constant; a single vein drains each gland. Knowledge of the specific location and course of each vein is crucial for successful surgery. Venous bleeding may result from damage to the fragile veins and may be difficult to localize and control. The right adrenal vein is wide and short, ~5 mm in length. This vein exits the right adrenal at its medial surface and drains directly into the inferior vena cava at its posterolateral aspect. Thus, this vessel may be difficult to control from an anterior exposure of the right adrenal, and the vein is easily torn when large right adrenal masses are retracted, which is a potentially fatal event. The left adrenal vein exits the left adrenal from its anterior surface and has a longer course medially to drain into the left renal vein, although aberrant venous drainage sometimes occurs. Because intraoperative exploration is the definitive evaluation for the presence of disease, the common ectopic locations where adrenal tumors occur must be known. This requires an understanding of adrenal embryology. The adrenal cortex develops from coelomic mesoderm near the urogenital ridge. Accessory adrenal cortical tissue may be identified as yellow masses in tissues surrounding the adrenal glands, in the kidneys, ovaries, in broad ligaments, or in the testes. The incidence of functioning extraadrenal cortical tissue is very low; these tissues need not be excised. The adrenal medullae arise from ectodermal neural crest cells in the thoracic region that migrate ventrolateral to the aorta and along adrenal vessels until contact is made with the primitive adrenocortical cells. Approximately 10% of sporadic pheochromocytomas arise in extraadrenal locations, most commonly in the organ of Zuckerkandl located to the left of the aortic bifurcation near the origin of the inferior mesenteric artery. Pheochromocytomas may arise anywhere in the periaortic sympathetic ganglia, however, as well as in the bladder, mediastinum, neck, anus, and vagina. Familial cases of pheochromocytomas have ~50% incidence of bilateral adrenal involvement. In contrast to ectopic adrenocortical tissue, extraadrenal pheochromocytomas frequently cause symptoms, and all gross tumor must be resected.

INDICATIONS FOR ADRENALECTOMY Surgical resection is the primary treatment for adrenal tumors that are functionally active or potentially malignant (Table 89-1). In general, adrenalectomy is indicated for functional cortical adenomas causing Cushing syndrome or primary aldosteronism, virilizing or feminizing tumors, adrenocortical carcinoma, pheochromocytoma, and incidental solid adrenal masses that are >6 cm in diameter on computed tomographic (CT) scan. Laparoscopic adrenalectomy has become the procedure of choice for solitary benign functional adenomas or pheochromocytomas that are small. Open resection is required for large masses (>5–6 cm in diameter) and cancers.

TABLE 89-1. Adrenalectomy Procedures Indicated by Preoperative Diagnosis

Bilateral adrenalectomy is indicated for conditions of primary bilateral adrenal gland hypersecretion, such as primary pigmented micronodular adrenal hyperplasia and massive macronodular adrenocortical hyperplasia (see Table 89-1).1 Bilateral adrenalectomy is also an effective treatment for severe, debilitating Cushing syndrome caused by an ectopic adrenocorticotropic hormone (ACTH)–secreting tumor that is either occult and not identified on imaging studies or metastatic and unresectable. Bilateral adrenalectomy may also be indicated for those patients with Cushing disease in whom transsphenoidal surgery and medical management have not been successful.2 Laparoscopy is the preferred surgical approach for these conditions because typical adrenal gland sizes are small enough to permit a minimally invasive procedure, and severely debilitated patients tolerate laparoscopic adrenalectomy better than open surgery. One exception is massive macronodular adrenocortical hyperplasia, which may require open resection because the adrenal glands may be quite large and tend to adhere to adjacent tissues. Patients with multiple endocrine neoplasia (MEN) type 2A or type 2B have a high incidence of bilateral adrenal pheochromocytomas. Therefore, some surgeons recommend bilateral adrenalectomy at the time of detection of the initial pheochromocytoma, even if the contralateral gland appears free of disease. This is commonly performed laparoscopically. Others argue that the risk of acute adrenal insufficiency after bilateral adrenalectomy is unacceptably high and can be lethal if untreated. In addition, 12 years may elapse before a pheochromocytoma arises in the uninvolved contralateral adrenal gland.3 Consequently, if the contralateral adrenal gland appears normal by preoperative studies and intraoperative examination, some recommend unilateral adrenalectomy to resect the identified pheochromocytoma and long-term follow-up with serial urinary screening tests for pheochromocytoma to detect the development of disease in the contralateral adrenal gland. Patient

compliance with follow-up must be assured.

ADRENAL INCIDENTALOMA The more widespread availability and use of high-resolution CT scanners have resulted in the detection of asymptomatic adrenal masses, called adrenal “incidentalomas,” which may have previously gone unrecognized. The dilemma for the surgeon is to identify and resect the minority of such tumors that are malignant or functional, and avoid risks of unnecessary operations for benign, nonfunctional adenomas, which occur commonly and do not require any treatment. Initial evaluation of a solid incidentaloma consists of a careful history-taking and physical examination to detect signs or symptoms suggestive of hypercortisolism, primary aldosteronism, pheochromocytoma, primary adrenocortical cancer, or adrenal metastasis from an undiagnosed malignancy arising in another tissue (Fig. 89-1). Laboratory studies are obtained based on clinical suspicion of the presence of one of these disorders.4 In addition, a 24-hour urinary free cortisol level should be obtained in each patient with an incidentaloma, because addisonian crisis may occur postoperatively in a few patients with clinically occult Cushing syndrome if stress doses of corticosteroids are not administered.4a Blood pressures and serum potassium level are also measured in each patient to help exclude an aldosteronoma.

FIGURE 89-1. Management of an incidentally discovered adrenal mass. An asymptomatic solid adrenal mass may be identified on abdominal computed tomographic (CT) scans obtained to evaluate other intraab-dominal processes. Functioning or potentially malignant adrenal lesions must be resected, whereas nonfunctioning, benign adrenocortical adenomas require no therapy. *, Clinical assessment consists of a careful history and physical examination for evidence of adrenal hyperfunction. Biochemical assessment includes measurement of the plasma potassium level and 24-hour urinary levels of catecholamines, vanillylmandelic acid, metanephrines, and free cortisol. **, If cancer metastatic to the adrenal gland is suspected, then CT-guided fine-needle aspiration may be attempted at this point for a confirmatory diagnosis.

Importantly, death may result from even a minor procedure that induces a surge of catecholamines from a pheochromocytoma in a patient who is either undiagnosed or inadequately prepared with b-blockade.5 Therefore, the presence of a pheochromocytoma must be definitively excluded in every patient with a solid adrenal incidentaloma before any procedures are undertaken. Clinical evaluation alone is inadequate, because these tumors may produce only episodic symptoms and may be clinically occult. A finding of normal 24-hour urine levels of catecholamines, vanillylmandelic acid, and metanephrines excludes the presence of a pheochromocytoma. In addition, on T2-weighted magnetic resonance images (MRIs), a pheochromocytoma characteristically appears extremely bright (signal intensity is three times that of the liver). In contrast, primary adrenocortical cancer and cancer that is metastatic to the adrenal gland appear only slightly brighter than the liver on MRI T2-weighted images, and cortical adenomas appear darker than the liver.6 If an adrenal incidentaloma is a solid mass that is not functional based on the previously described evaluation, and the presence of a pheochromocytoma can be excluded, then the final differentiation is a malignancy from a benign adrenal adenoma. Malignancy is most conclusively established or excluded by resecting each incidentaloma. However, this approach is not justified because primary adrenal carcinoma is rare, whereas benign adrenal adenomas occur frequently. If an adrenal mass is likely to be a metastasis from a malignancy arising in another tissue, then percutaneous biopsy for cytologic analysis, after a pheochromocytoma has been excluded, may be useful if a diagnosis of metastatic cancer can be established, and adrenalectomy can be avoided. However, if metastatic cancer is not likely, then percutaneous fine-needle biopsy of an adrenal incidentaloma is not indicated, because cytologic analysis cannot accurately differentiate primary adrenal cancers from benign adenomas. Laparoscopic adrenalectomy may be reasonable in healthy, young patients with moderate-sized nonfunctional adrenal incidentalomas (4–6 cm in diameter). The procedure must be performed safely with low risk by an experienced surgeon. In such cases, laparoscopic adrenalectomy provides a prompt, definitive diagnosis and peace of mind for the patient. Recovery to full preoperative activity occurs promptly, and repeated evaluations and testing are unnecessary if the mass is benign. The best radiologic factor to discriminate nonfunctioning benign and malignant adrenal tumors is the diameter of the mass on CT. Most adrenocortical adenomas are 6 cm is >35%. Therefore, every solid adrenal mass that is ³6 cm in diameter should be resected (see Fig. 89-1). An adrenocortical carcinoma may be detected at an early stage, at 99% of all 45,X embryos can be assumed to be aborted. The high rate at which 45,X cell lines are spuriously detected in chorionic villi, however, raises the possibility that the frequency of 45,X in embryos is overestimated. Monosomy X may be relatively restricted to villi or membranes, those tissues most often available for cytogenetic studies, and the aborted embryo may have a different chromosomal complement. In any case, intrauterine growth restriction is characteristic of the rare surviving 45,X neonate.66 GONADAL DEVELOPMENT AND X-CHROMOSOME INACTIVATION. The absence of oocytes in monosomy X is the result of increased oocyte atresia, not failure of germ-cell formation. The 45,X embryos and 45,X neonates have germ cells.29,30 Inasmuch as germ cells are present in 45,X embryos, the fact that 3% to 5% of 45,X

individuals menstruate spontaneously is not too surprising. Several fertile 45,X individuals have been reported.67 Because (according to the Lyon hypothesis) X chromosomes in excess of one are inactivated, why 45,X individuals should manifest developmental abnormalities is not so obvious. Relatively normal ovarian development occurs in 39,X mice and in most other monosomy X mammalian organisms. Data support several related explanations for these findings. First, some loci on the human heterochromatic (inactive) X are not inactivated. For example, the locus for steroid sulfatase, which is located on Xp, is not inactivated, and it would not be surprising if ovarian determinants likewise escaped X inactivation. Second, X inactivation never occurs in human oocytes, as evidenced by the fact that females who are heterozygous for the enzyme glucose-6-phosphate dehydrogenase synthesize both alleles in oocytes.68 The parental origin of 45,X is of interest. In humans, 70% of live-born 45,X individuals have lost a paternal sex chromosome.69 This helps explain why mean maternal age is not increased for 45,X abortuses or live borns. 70 Murine monosomy (39,X) also results from the loss of a paternal sex chromosome at the time of fertilization.71 Most commonly, individuals with gonadal dysgenesis have a 45,X karyotype and present with the Turner stigmata.72 Secondary sexual development usually does not occur in 45,X individuals (Fig. 90-15). Pubic and axillary hair fail to develop in normal quantity. Although well differentiated, the external genitalia, the vagina, and the müllerian derivatives (e.g., uterus) remain small. As is true for virtually all individuals with gonadal dysgenesis, estrogen and androgen levels are decreased; levels of follicle-stimulating hormone (FSH) and luteinizing hormone (LH) are increased.

FIGURE 90-15. A, Appearance of one of original seven cases of Turner syndrome reported in 1938. B, Same patient, photographed in 1972, almost 35 years after publication of the original case report. The karyotype was documented to be 45,X. The patient had received little estrogen in intervening years and experienced severe osteoporosis. (Courtesy of Dr. R. Rebar and Dr. S. S. C. Yen, University of California, San Diego.)

NONGONADAL FEATURES OF TURNER SYNDROME. The common anomalies of Turner syndrome include epicanthal folds, high arched palate, low nuchal hairline, webbed neck, shield like chest, coarctation of the aorta, ventricular septal defect, renal anomalies, pigmented nevi, lymphedema, hypoplastic nails, and cubitus valgus (Table 90-3). Inverted nipples and double eyelashes may be present as well. No feature is pathognomonic, but in aggregate, they form a spectrum of anomalies more likely to exist in 45,X individuals than in normal 46,XX individuals. These anomalies are the Turner stigmata, the presence of which suggests the coexistence of gonadal dysgenesis.

TABLE 90-3. Somatic Features Associated with 45,X Karyotype

Individuals with a 45,X karyotype have low birth weights (adjusted mean, 2851.1 ± 65.1 g).66 Total body length at birth is sometimes less than normal, but often it is normal. Height velocity before puberty generally lies in the 10th to the 15th percentile, and the mean heights of 45,X adults (16 years or older) range from 141 to 146 cm (55.5–57.5 in).72,73,74 and 75 In untreated patients, the epiphyses remain open; additional growth occurs when sex steroids are administered. Despite the diminished final height of such patients, their adult stature tends to correlate with parental height.75 (Also see ref. 75a.) That not all patients with gonadal dysgenesis are short indicates that sex steroid deficiency is not the cause. For example, normal stature is characteristic of individuals with 46,XX gonadal dysgenesis. Growth hormone (GH) levels have long been considered essentially normal in individuals with gonadal dysgenesis.76,77 Cellular resistance to GH, however, has been suggested. This relative resistance may be overcome by treatment with exogenous GH at higher doses than those used for classic GH deficiency.78 Anti-GH antibodies have been detected, and GH reserve may be decreased.79,80 More evidence suggests that abnormalities in GH or insulin-like growth factor-I (IGF-I), which fall below the normal range after 8 years of age, are secondary to the lack of gonadal activation and estrogen secretion.81 Short stature may be common in gonadal dysgenesis because the epiphyses are structurally abnormal. This hypothesis is compatible with observations that decreased growth occurs in the long bones, teeth, and skull.82,83 One aspect of Turner syndrome may be a skeletal dysplasia. Most 45,X patients have normal intelligence, but any given 45,X patient has a slightly higher probability of being retarded than a 46,XX individual. The frequency of overt retardation is 11% to 17%.72 Biases of ascertainment dictate that this prevalence probably represents the maximum risk. Performance IQ is definitely lower than verbal IQ, however; 45,X individuals have an unusual cognitive defect characterized by an inability to appreciate the shapes and relations of objects with respect to one another (i.e., space-form blindness).84,85 and 86 The patients usually appear socially immature, probably in part because they are short and sexually immature.87 METABOLIC ALTERATIONS. Diabetes mellitus, thyroid disease, and essential hypertension often are present in individuals with 45,X karyotypes. Abnormal oral glucose-tolerance tests may occur in as many as 40% of these individuals.88 Both autoimmune thyroiditis and Graves disease are observed with increased frequency in patients with Turner stigmata. Approximately one-third of adult 45,X patients have essential hypertension, which also may occur in young 45,X girls. Therapy for any metabolic alterations is not unique, although reduction or careful monitoring of exogenous estrogen therapy may be necessary. Abnormalities of the X Chromosome. Several different abnormalities of the X chromosome have been associated with gonadal dysgenesis. The cytologic origin of these defects was considered earlier. DELETION OF THE SHORT ARM OF CHROMOSOME X. A terminal deletion of the X short arm [del(Xp)] may or may not cause gonadal dysgenesis, short stature, and other features of the Turner stigmata. The phenotype depends on the amount of Xp that is deficient. Spontaneous menstruation, albeit usually leading to secondary amenorrhea, has occurred in almost 40% of reported 46,X,del(X)(p11) individuals (see Fig. 90-7). Almost all 46,X,del(X)(p21) individuals menstruate, but only approximately one-half become pregnant.7,34,89,90 These data indicate that ovarian tissue persists more often in del(Xp) individuals than in 45,X individuals and that complete ovarian failure (with primary amenorrhea) occurs only if the proximal and the terminal portions of Xp are deleted. However, the mean adult heights are 140 cm (55 in) for 46,X,del(X)(p11) and 146.5 cm (57.5 in) for 46,X,del(X)(p21) persons89,91 (see Fig. 90-8). Inasmuch as 46,X,del(X)(p21) individuals are short but have normal ovarian function, determinants for ovarian maintenance and for stature must be located in different regions of Xp; statural determinants are more distal.35,36,91 No evidence yet exists that any given X-specific probe bears a special relation to X-ovarian or X-structural

determinants. ISOCHROMOSOME FOR THE X LONG ARM (1[XQ]). Almost all 46,X,i(Xq) patients have streak gonads, short stature, and features of the Turner stigmata. In addition to having a duplication of Xq,i(Xq), these individuals differ from del(X)(p11) persons because the terminal portion and almost all of Xp is deleted. The better gonadal function in del(X)(p11) than in i(Xq) individuals is consistent with the location of ovarian determinants at several different sites on Xp. A locus on Xp may be deleted in i(Xq) karyotype but be retained near the centromere in 46,X,del(X)(p11) cases. Because duplication of Xq (i.e., 46,X,i[Xq]) does not compensate for deficiency of Xp, the gonadal determinants on Xq and Xp must have different functions. Whether duplication of Xq per se produces abnormalities is unknown. DELETION OF THE LONG ARM OF CHROMOSOME X (DEL[XQ]). Most patients with a deletion of the X long arm have streak gonads and never menstruate. This is especially true of individuals with del(X)(q13). However, deletions of distal Xq are more likely to be associated with premature ovarian failure than with primary amenorrhea. The Xq appears to contain more than one region that is required for normal ovarian function. Perhaps several Xq loci can affect ovarian function in additive fashion. The only clue to the nature of any of these products is the suggestion that the diaphanous (DIA) gene, localized to Xq25, plays a role; however, other genes in distal Xq are also pivotal to ovarian development. Originally, deletions of Xq were not thought to result in short stature, but later tabulations show a definitely decreased mean height of persons with del(X)(q13).36,72,89,92 Whether the short stature reflects specific loci or vicissitudes of X-inactivation is unclear.93 Mosaicism MOSAICISM INVOLVING ONLY X CHROMOSOMES. The 45,X/46,XX individuals have fewer anomalies than 45,X individuals. In one survey, 12% of 45,X/46,XX individuals menstruated, compared with only 3% of 45,X individuals.72 In that survey, the mean adult height was greater in 45,X/46,XX persons than in 45,X individuals. More mosaic patients (25%) than nonmosaic patients (5%) reach adult heights greater than 152 cm (60 in). Somatic anomalies are less likely to occur in 45,X/46,XX than in 45,X patients. MOSAICISM WITH A Y CHROMOSOME. Individuals with a 45,X cell line and at least one line containing a Y chromosome manifest a variety of phenotypes, ranging from almost normal males with cryptorchidism or penile hypospadias to females indistinguishable from those with the 45,X Turner syndrome. The different phenotypes presumably reflect different tissue distributions of the various cell lines, although this assumption remains unproven. At any rate, 45,X/46,XY individuals may show unambiguous female external genitalia, ambiguous external genitalia, or almost normal male external genitalia. Some 45,X/46,XY individuals with female external genitalia have the Turner stigmata and are clinically indistinguishable from 45,X individuals. Others, however, are female but of normal stature and without somatic anomalies. As in other types of gonadal dysgenesis, the external genitalia, vagina, and müllerian derivatives remain unstimulated because of deficient sex steroids. Breasts fail to develop, and little pubic or axillary hair grows. In fact, breast development in a 45,X/46,XY individual should lead one to suspect an estrogen-secreting tumor, most commonly a gonadoblastoma or dysgerminoma. The streak gonads of 45,X/46,XY individuals usually are indistinguishable histologically from the streak gonads of individuals with 45,X gonadal dysgenesis. However, gonadoblastomas or dysgerminomas develop in 15% to 20% of 45,X/46,XY individuals.55,56,94 Such neoplasms may arise as early as the first two decades of life. Gonadoblastomas occur almost exclusively in 46,XY or 45,X/46,XY individuals and usually are benign. However, they may be associated with dysgerminomas or other germ-cell tumors that are malignant. The gonads of 45,X/46,XY individuals should be extirpated regardless of the patient's age. Because of the risk of neoplasia, 45,X/46,XY gonadal dysgenesis should be differentiated from forms of gonadal dysgenesis lacking a Y chromosome. When the polymerase chain reaction for SRY was used, unrecognized Y-chromosome material was found in 1 of 40 patients with Turner syndrome.95 Because the detection of SRY sequences in patients with gonadal dysgenesis is correlated with the presence of Y-chromosomal DNA and carries the risk of tumor development, the application of this technique in search of SRY may be justified in all individuals with this disorder. Individuals may show one streak gonad and one dysgenetic testis. The terms asymmetric gonadal dysgenesis or mixed gonadal dysgenesis are often applied to such individuals. They usually have ambiguous external genitalia. Many investigators believe that the phenotype of asymmetric gonadal dysgenesis is almost always associated with 45,X/46,XY mosaicism, with ostensibly nonmosaic cases reflecting merely an inability to analyze appropriate tissues. Most 45,X/46,XY individuals with ambiguous external genitalia have müllerian derivatives (e.g., a uterus). The presence of a uterus is helpful diagnostically, because a uterus is absent in most genetic forms of male pseudohermaphroditism. If an individual has ambiguous external genitalia, bilateral testes, and a uterus, one may reasonably infer the presence of 45,X/46,XY mosaicism, whether or not both lines can be demonstrated cytogenetically. Occasionally, the uterus is rudimentary, or a fallopian tube fails to develop ipsilateral to a testis. The 45,X/46,XY mosaicism has less commonly been detected in individuals with almost normal male external genitalia. In some individuals, hypospadias is present, but the sex-of-rearing is unequivocally male. The 45,X/47,XXY and 45,X/46,XY/47,XXY complements exist, albeit much less often than 45,X/46,XY. These complements are associated with the same phenotypic spectrum as 45,X/46,XY. Of particular interest is one family in which two and possibly three sibs had 45,X/46,XY/47,XYY mosaicism.96 The parents were second cousins, suggesting recessive factors. Gonadal Dysgenesis in 46,XX Individuals. The first individual with gonadal dysgenesis and an apparently normal female (46,XX) complement was reported in 1960.12 By 1971, a survey of 61 such individuals led to the conclusion that the disorder was inherited in an autosomal recessive fashion.12 A study of all XX gonadal dysgenesis cases in Finland confirmed this conclusion.97 The external genitalia and the streak gonads (see Fig. 90-14) in XX gonadal dysgenesis are indistinguishable from those in gonadal dysgenesis secondary to a sex chromosomal abnormality. Likewise, the endocrine profiles do not differ, but individuals with XX gonadal dysgenesis usually have normal stature (Fig. 90-16). Several pathogenic mechanisms can be postulated, but firm data have not been gathered. Phenocopies for XX gonadal dysgenesis also are well recognized (see Chap. 96).

FIGURE 90-16. A case of 46,XX gonadal dysgenesis: a 19-year-old woman with primary amenorrhea. Notice the normal stature and diminished breast development. Circulating gonadotropin levels were markedly elevated.

Both XX gonadal dysgenesis and neurosensory deafness have occurred in multiple sibs in several families; the occurrence of deaf but fertile male sibs confirms autosomal inheritance.98 The coexistence of gonadal and auditory anomalies probably indicates a syndrome distinct from XX gonadal dys-genesis without deafness (i.e., genetic heterogeneity). Further evidence for genetic heterogeneity can be cited. In several other families, unique patterns of somatic anomalies indicate the existence of mutant genes distinct from those already discussed. These include: XX gonadal dysgenesis and myopathy; XX gonadal dysgenesis and cerebellar ataxia; XX gonadal dysgenesis and metabolic acidosis; XX gonadal dysgenesis, microcephaly, and arachnodactyly.99,100 and 101 Follicle-Stimulating Hormone–Receptor Mutations. At least one form of XX gonadal dysgenesis is now known to be caused by a mutation of the FSH receptor (FSHR). Affected women present with primary or secondary amenorrhea and elevated serum FSH levels indicative of premature ovarian failure97,102 (see Chap. 96). The mutation originally was identified in a large number of sporadic and familial cases in Finland. Most cases were found in north central Finland, a sparsely populated

part of the country. The overall frequency of the disorder in Finland was 1 per 8300 females, a relatively high incidence attributed to a founder effect. The segregation ratio of 0.23 for female sibs was consistent with autosomal recessive inheritance, as was the consanguinity rate of 12%. Sib-pair analysis using polymorphic DNA markers was used to localize the gene to a specific region of chromosome arm 2p, a region known to contain genes for both FSHR and the LH receptor (LHR). One specific mutation (C566T:alanine to valine) in exon 7 was observed in six multiplex families.102,103 On transvaginal ultrasonography, most of these patients have demonstrable ovarian follicles, raising the possibility of residual receptor activity.103 The C566T mutation was not found in all Finnish XX gonadal dysgenesis patients and is rarely detected in 46,XX women with ovarian failure who reside outside Finland.104,105 Gonadal Dysgenesis in 46,XY Individuals. Gonadal dysgenesis also can occur in 46,XY individuals. In XY gonadal dys-genesis, affected individuals are phenotypic females who show sexual infantilism and bilateral streak gonads (Fig. 90-17). The gonads may undergo neoplastic transformation (20– 30% prevalence)56 (Fig. 90-18). At least one form of XY gonadal dysgenesis results from an X-linked recessive or male-limited autosomal dominant gene.106,107 Sporadic cases may result from deletion or point mutations within SRY on the Y short arm. 14,108,109 Further evidence for genetic heterogeneity lies in the existence of at least three syndromes having XY gonadal dysgenesis as one of their components: XY gonadal dysgenesis and long-limbed camptomelic dwarfism, XY gonadal dysgenesis and ectodermal defects, and the genitopalatocardiac syndrome.36,110,111 and 112

FIGURE 90-17. A and B, Two examples of 46,XY gonadal dysgenesis. Both 16-year-old individuals presented with primary amenorrhea and markedly elevated concentrations of circulating gonadotropins. Both patients had dysgerminoma of an ovary; the patient in B also had a large gonadoblastoma of the contralateral ovary. Breast development as in B is extremely rare and no doubt secondary to hormone production by the patient's gonadal neoplasm. (B, Photograph reproduced from Villanueva AL, Benirschke K, Campbell J, et al. Complete development of secondary sexual characteristics in a case of 46,XY gonadal dysgenesia. Obstet Gynecol 1984; 64:68S.)

FIGURE 90-18. Gross appearance of the uterus, fallopian tubes, and gonads of 46,XY patient with gonadal dysgenesis. Right gonad (cut) contained dysgerminoma.

Rudimentary Ovary Syndrome and Unilateral Streak Gonad Syndrome. The rudimentary ovary syndrome is a poorly defined entity of unknown cause said to be characterized by decreased numbers of follicles. This “syndrome” is heterogeneous, not a single entity. Many cases have been associated with sex chromosomal abnormalities, particularly 45,X/46,XX mosaicism. Similar statements also apply to individuals with the unilateral streak ovary syndrome. For example, a unilateral streak gonad and a contralateral polycystic ovary have been observed in a 46,XX/46,X,i(Xq) individual who became pregnant.113 EVALUATION AND TREATMENT OF GONADAL DYSGENESIS. When Turner stigmata are present, the diagnosis of gonadal dysgenesis usually is made early in childhood. The index of suspicion should be high for any infant with lymphedema of the hands and feet at birth, especially because the somatic anomalies are not very obvious in neonates (see Fig. 90-9). Other children present for evaluation of sexual infantilism, and still others with a male sex-of-rearing may not virilize at the expected age of puberty. Short stature is another common reason that evaluation is sought. The measurement of circulating gonadotropin concentrations and the determination of the karyotype can establish the diagnosis. Longitudinal studies have documented elevated gonadotropin levels at all ages in gonadal dysgenesis, indicating absence of appropriate feedback inhibition of the hypothalamic-pituitary unit by the dysgenetic gonads even in childhood.114,115 Chromosomal studies are indicated to eliminate the possibility of a Y chromosome. The use of the polymerase chain reaction to detect sequences of SRY may well be warranted. If the phenotype and karyotype are compatible, the only tissue needed is blood for lymphocyte culture. If the phenotype and karyotype are incompatible (e.g., tall “45,X” subjects), skin or gonadal fibroblasts also should be cultured to detect any mosaicism. If a Y chromosome is identified, surgical extirpation of the dysgenetic gonads is indicated to prevent neoplasms. Streak gonads usually can be removed by laparoscopy. In appropriate cases in which disseminated malignancies do not involve the gonads, the uterus may be left in situ for donor in vitro fertilization or embryo transfer. The evaluation of other commonly involved organ systems should include a careful physical examination, with special attention to the cardiovascular system, and should include thyroid function tests (including antibodies), fasting blood glucose level, renal function tests, and an intravenous urogram or a renal ultrasonographic scan. The treatment of individuals with short stature has received significant attention. A multicenter, prospective, randomized trial of administration of GH, alone and in combination with oxandrolone, was initiated in 1983.116,117 Data on 62 girls, who have received 3 to 6 years of treatment, have been published. Given an average height of 143 cm for untreated girls in the United States, the mean height of 151.9 cm in the 30 girls whose therapy was terminated represented a net increase of 8.1 cm (therapy was terminated because the subjects had met study criteria for cessation of treatment, including bone age >14 years and a growth velocity of 150 cm, the widely accepted lower limit of normal height for women in the United States, is now attainable by most girls with this disorder. What the exact dose of GH should be and what additional benefit oxandrolone may contribute are not known. GH administered in doses 25% above those recommended for GH deficiency is proving to be remarkably safe. To enable the patient to achieve sexual maturation, estrogen therapy should be initiated when the patient is psychologically ready, perhaps at the age of 12 to 13 years, and after GH therapy is completed. Because the aim is to mimic normal pubertal development, therapy is initiated with low-dose estrogen alone (such as oral

conjugated estrogens 0.3 mg daily) given continuously for the first several months. A progestogen (5–10 mg of oral medroxyprogesterone acetate or 200 mg of oral micronized progesterone daily for 14 days each month) is then added to prevent the development of endometrial hyperplasia, either when the patient notices vaginal bleeding or after 6 months of unopposed estrogen therapy. Thereafter, the dose of estrogen is increased slowly over 1 to 2 years to daily oral doses of 1.25 to 2.5 mg of conjugated estrogen, 1 to 2 mg of micronized 17b-estradiol, 1.25 to 2.5 mg of piperazine estrone sulfate orally, or 0.05 to 0.1 mg of estradiol transdermally, with the patch changed once or twice weekly depending on the specific preparation. Generally, continued breast tenderness is the first sign that the dose of estrogen is excessive. These patients must be observed especially closely for the development of hypertension with estrogen therapy. The patients and their parents should be informed of the emotional and physical changes that will occur with therapy. Additional principles for estrogen replacement may be found in Chapter 100 (also see ref. 119a). Individuals with mosaic gonadal dysgenesis and other chromosomal aberrations such as del(X)(p11) or del(X)(q25 to q27) may develop normally at puberty. The decision whether to initiate estrogen therapy can be made on the basis of growth rates, bone age, uterine size determined by ultrasonography, and circulating gonadotropin concentrations. FSH levels in the normal range for age imply functional gonads. Examination of individuals with sexual infantilism who have been treated with large doses of estrogen (especially conjugated estrogens) from the outset to effect maturation of secondary sexual characteristics often reveals abnormal and inappropriate breast development. The areola often becomes abnormally large, with minimal additional breast tissue. Whether breast contour can be improved in individuals subjected to such inappropriate therapy is not clear. Pregnancies can now be achieved in these individuals, with success rates of >50%, by using donor oocytes.120 Hormone replacement regimens using transdermal estradiol-17b and intramuscular progesterone have generally been associated with the highest pregnancy rates among those regimens aimed at preparing the endometrium to receive an embryo. TRIPLE-X SYNDROME (47,XXX). The existence of individuals with a 47,XXX karyotype was first reported in 1959, and an association with premature ovarian failure was noted.121 Since then, study has shown that patients with the triple-X syndrome need not have any impairment in fertility nor any shortening of their reproductive lives.122,123 Premature menopause occurs with increased frequency, however, compared with its incidence in karyotypically normal individuals.124 Moreover, reports of patients with triple-X syndrome associated with immunoglobulin deficiency, together with the finding that the control of T-cell function may be related to the X chromosome, suggest an association between immunologic abnormalities and triple-X syndrome in females with premature ovarian failure.125,126 and 127 Because affected individuals are phenotypically normal, they usually are identified only after presenting with hypergo-nadotropic amenorrhea. Treatment for the premature ovarian failure is the same as it is for menopausal women and is discussed in Chapter 100. ULLRICH-NOONAN SYNDROME. A syndrome now known as the Ullrich-Noonan or pseudo-Turner syndrome has been described in which phenotypic females present with many of the Turner stigmata and a normal 46,XX chromosomal complement.128,129,131 Moreover, females have functioning ovaries, although the onset of puberty may be delayed. Males also may present with this disorder. The common stigmata include short stature, webbed neck, ptosis, and (unlike in gonadal dysgenesis) right-sided congenital heart disease. Pulmonic stenosis (occurring in perhaps 50% of individuals) and atrial septal defect are the most common cardiac anomalies, although ventricular septal defect, ventricular hypertrophy, patent ductus arteriosus, coarctation of the aorta, and aortic stenosis may be found. Mental retardation, pectus excavatum, cubitus valgus, and lymphedema also may occur (see Chap. 92). The incidence of this syndrome has been estimated at ~1 in 8000, with >80% of cases arising from spontaneous mutations.132 Familial clusters consistent with autosomal dominant inheritance have been described. The existence of this syndrome reinforces the need to perform a karyotype study of individuals presenting with the Turner stigmata. Otherwise, individuals thought to have gonadal dysgenesis may develop normally at puberty and become fertile, much to the surprise of the physician. SEMINIFEROUS TUBULE DYSGENESIS AND ITS VARIANTS Klinefelter Syndrome. Males with at least one Y chromosome and at least two X chromosomes have Klinefelter syndrome.7 Most cases are 47,XXY, but the phenotype may also be associated with 46,XY/47,XXY; 48,XXYY; and 49,XXXXY complements.133,134 and 135 The most characteristic features are seminiferous tubule dysgenesis and androgen deficiency. Somatic anomalies sometimes coexist.133 The presence of a chromosomal abnormality and of elevated gonadotropin levels differentiates Klinefelter syndrome from hypogonadotropic hypogonadism. These conditions are discussed in Chapter 115. 46,XX Sex-Reversed Males. The 46,XX sex-reversed males are phenotypic males with bilateral testes.7 Their chromosomal complement, however, appears to be that of a female (see Chap. 115). Affected patients have small testes and signs of androgen deficiency but otherwise have a normal male appearance. Facial and body hair are decreased, and pubic hair may be distributed in the pattern characteristic of females. Approximately one-third have gynecomastia. The penis and scrotum are small but usually well differentiated, and wolffian derivatives are normal. By definition, the sex-of-rearing is not in doubt.136,137 Seminiferous tubules are decreased in number and in size, peritubular and interstitial fibrosis is present, Leydig cells are hyperplastic, and spermatogonia usually cannot be detected. Occasionally, immature spermatogonia are found, and sometimes the ejaculate contains spermatozoa. In 46,XX males, as in 46,XX true hermaphrodites, testes develop, contrary to expectations that a Y chromosome is required for testicular differentiation. Translocation of the TDF from the Y to the X chromosome has been documented by molecular studies in most cases (80%).63,138 Moreover, familial aggregates of 46,XX males alone or 46,XX males and 46,XX true hermaphrodites also have been reported.61,62,63 and 64 Like XX true hermaphrodites, XX males in these kindreds may not show the Y-X translocation.63,64 These observations support the possibility that autosomal sex-reversal genes exist or that mutation in an autosomal or X chromosomal gene that permits testicular differentiation in the absence of TDF has occurred.34,35,64,138 OTHER DISORDERS AFFECTING THE TESTES Germinal Cell Aplasia. Del Castillo and coworkers were the first to describe normally virilized yet sterile males with germinal cell aplasia (i.e., Sertoli cell–only syndrome or del Castillo syndrome).139 Seminiferous tubules lack spermatogonia, and the testes are slightly smaller than average. Leydig cell function is normal, however, and secondary sexual development is normal. In germinal cell aplasia, FSH levels are elevated, but LH levels are normal.139a Tubular hyalinization and sclerosis usually do not occur. Occasionally, a few spermatozoa are present, but affected individuals usually are sterile. Despite infertility, androgen therapy is unnecessary because secondary sexual development is normal. In five families, a male with this condition has had a sister with streak gonads (see Chap. 115). Congenital Anorchia. Males (46,XY) with anorchia (i.e., vanishing testis syndrome) have unambiguous male external genitalia, normal wolffian derivatives, no müllerian derivatives, and no detectable testes7 (see Chap. 115). Unilateral anorchia is not extraordinarily rare, but bilateral anorchia is uncommon. Somatic abnormalities are rare. Despite the absence of testes, the phallus is well differentiated. The pathogenesis presumably involves atrophy of the fetal testes after 12 to 16 weeks of gestation, by which time genital virilization has occurred. The vasa deferentia terminate blindly, often in association with the spermatic vessels. The diagnosis should be applied only if testicular tissue is not detected in the scrotum, the inguinal canal, or the entire path along which the testes descend during embryogenesis. Splenic gonadal fusion also can occur, mimicking the disorder. Heritable tendencies exist, but the occurrence of monozygotic twins discordant for anorchia suggests that genetic factors are not paramount in all cases.140,141 Perhaps a heritable tendency toward in utero torsion of the testicular artery exists, explaining occasional familial aggregates (see Chap. 115). Syndrome of Rudimentary Testes. Men have also been reported who had well-formed but small testes (6 times as frequent as in the control group. (See also ref.42.) Patients who have undergone orchiopexy should be informed about the increased risk of malignancy and the need for frequent self-examination. Generally, for postpubertal patients with unilateral cryptorchidism, orchiopexy has been recommended. Studies suggest that orchiectomy is the treatment of choice for the majority of postpubertal males presenting with unilateral cryptorchidism.43 The investigators cite the following reasons for recommending orchiectomy instead of orchiopexy: (a) the majority of cryp-torchid testes have absent to very low numbers of spermatogo-nia, (b) a significant potential for malignancy exists, and (c) the chance of torsion of the undescended testicle is increased. When anorchia is present, or after orchiectomy, a testicular prosthesis should be placed in the scrotum to minimize the psychological impact of an empty scrotum. CHAPTER REFERENCES 1. 2. 3. 4.

Feldman KW, Smith DW. Fetal phallic growth and penile standards for newborn male infants. J Pediatr 1975; 86:395. Flatau E, Josefsberg Z, Reisner SH, et al. Penile size in the newborn infant. J Pediatr 1875; 87:663. Schonfeld WA, Beebe GW. Normal growth and variation in the male genitalia from birth to maturity. J Urol 1942; 48:759. Evans BAJ, Williams DM, Hughes IA. Normal postnatal androgen production and action in isolated micropenis and isolated hypospadias. Arch Dis Child 1991; 66:1033.

4a. Ludwig G. Micropenis and apparent micropenis—a diagnostic and therapeutic challenge. Andrologia 1999; 31(Suppl 1):27. 5. Lovinger RD, Kaplan SL, Grumbach MM. Congenital hypopituitarism associated with neonatal hypoglycemia and micropenis: four cases secondary to hypothalamic hormone deficiencies. J Pediatr 1975; 87:1171. 6. Burstein S, Kaplan SI, Grumbach MM. Early determination of androgen-responsiveness is important in the management of microphallus. Lancet 1979; 2:983. 7. Husmann DA, Cain MF. Microphallus: eventual phallic size is dependent on the timing of androgen administration. J Urol 1994; 152:734. 8. Ritchey ML, Bloom D. Summary of the urology section. Pediatrics 1995; 96:138. 9. Shima H, Yabumoto H, Okamoto E, et al. Testicular function in patients with hypospadias associated with enlarged prostatic utricle. Br J Urol 1992; 69:192. 10. Nonomura K, Fujieda K, Sakakibara N, et al. Pituitary and gonadal function in prepubertal boys with hypospadias. J Urol 1984; 132:595. 11. Aarskog D. Maternal progestins as a possible cause of hypospadias. N Engl J Med 1979; 300:76. 12. Barakat AY, Seikaly MG, Der Kaloustan VM. Urogenital abnormalities in genital disease. J Urol 1986; 136:778. 13. Zaontz MR, Packer MG. Abnormalities of the external genitalia. Pediatr Clin North Am 1997; 44:1267. 14. Hutson JM, Hasthorpe S, Heyms CF. Anatomical and functional aspects of testicular descent and cryptorchidism. Endocr Rev 1997; 18:259. 15. Pouplard MD, Job JC, Luxembourger I, et al. Antigonadotropic cell and antibodies in the serum of cryptorchid children and infants and of their mothers. J Pediatr 1985; 107:26. 16. Martinetti M, Maghnie, M, Salvaneschi L, et al. Immunogenetic and hormonal study of cryptorchidism. J Clin Endocrinol Metab 1992; 74:39. 17. Mayr J, Rune GM, Holas A, et al. Ascent of the testis in children. Eur J Pediatr 1995; 154:893. 18. Prader A. Testicular size: assessment and clinical importance. Triangle 1966; 7:240. 19. Scorer CG, Farrington GH. Congenital deformities of the testis and epididymis. New York: Appleton-Century-Crofts, 1972. 20. Berkowitz GS, Lapinski RH, Dolgin SE, et al. Prevalence and natural history of cryptorchidism. Pediatrics 1993; 92:44. 21. Rezvani I. Cryptorchidism: a pediatrician's view. Pediatr Clin North Am 1987; 34:735. 22. Sizonenko PC, Schindler A-M, Cuendet A. Clinical evaluation and management of testicular disorders before puberty. In: Burger H, de Kretser D, eds. The testes. New York: Raven Press, 1981:303. 23. De Muinck SMPF, Hazebroek FWJ, Drop SLS, et al. Hormonal evaluation of boys born with undescended testes during the first year of life. J Clin Endocrinol Metab 1988; 66:159. 24. Van Vliet G, Ceufriez A, Robyn C, et al. Plasma gonadotropin values in prepubertal cryptorchid boys: similar increase of FSH secretion in uni- and bilateral cases. J Pediatr 1980; 97: 253. 25. Atlas I, Stone N. Laparoscopy for evaluation of cryptorchid testis. Urology 1992; 40:256. 26. Rivarola MA, Bergada C, Cullen M. HCG stimulation test in prepubertal boys with cryptorchidism, bilateral anorchia and in male pseudohermaph-roditism. J Clin Endocrinol Metab 1970; 31: 526.

27. Levitt SB, Kogan SJ, Schneider KM, et al. Endocrine tests in phenotypic children with bilateral impalpable testes can reliably predict “congenital” anorchism. Urology 1978; 11:11. 28. Bartone FF, Huseman CA, Maizels M, et al. Pitfalls in using human chorionic gonadotropin stimulation test to diagnose anorchia. J Urol 1984; 132:563. 28a. Kubini K, Zachman M, Albers N. Basal inhibin B and the testosterone response to human chorionic gonadotropin correlate in prepubertal boys. J Clin Endocrinal Metab 2000; 85:134. 29. 30. 31. 32. 33.

Laron Z, Dickerman Z, Ritterman I, et al. Follow-up of boys with unilateral compensatory testicular hypertrophy. Fertil Steril 1980; 33:303. Mengel W, Wronecki K, Schroeder J, et al. Histopathology of the crypt-orchid testis. Urol Clin North Am 1982; 9:331. Hadziselimovic F, Herzog B, Seguchi H. Surgical correction of cryptorchid-ism at 2 years: electron microscopic and morphometric investigations. J Pediatr Surg 1975; 10:19. Longui CA, Arnhold IJP, Mendonca BB, et al. Serum inhibin levels before and after gonadotropin stimulation in cryptorchid boys under age 4 years. J Pediatr Endocrinol Metab 1998; 11:687. Mengel W, Zimmerman FA. Immunologic aspects of cryptorchidism. In: Fonkalsaud EW, Mengel W, eds. The undescended testis. Chicago: Year Book Medical Publishers, 1981:57.

33a. Tekin A, Aygun YC, Aki FT, Ozen H. Bilateral germ cell cancer of the testis: a report of 11 patients with a long-term follow-up. BJU Int 2000; 85:864. 34. Martin DC. Malignancy in the cryptorchid testis. Urol Clin North Am 1982; 9:37. 35. Benson RC, Beard CM, Kelalis PP, Kurland LT. Malignant potential of the cryptorchid testis. Mayo Clin Proc 1991; 66:372. 36. Forest MG, David M, Lecoq A, et al. Kinetics of the HCG-induced ste-roidogenic response of the human testis. III: Studies in children of the plasma levels of testosterone and HCG: rationale for testicular stimulation test. Pediatr Res 1980; 14:819. 37. Christiansen P, Muller J, Buhl S, et al. Hormonal treatment of cryptorchid-ism—hCG or GnRH—a multicentre study. Acta Pediatr 1992; 81:605. 38. Gill B, Kogan S. Cryptorchidism: current concepts. Pediatr Clin North Am 1997; 44:1211. 39. Pyorala S, Huttunen NP, Uhari M. A review and meta-analysis of hormonal treatment of cryptorchidism. J Clin Endocrinol Metab 1995; 80:2795. 40. Rozanski TA, Bloom DA. The undescended testis. Urol Clin North Am 1995; 22:107. 41. Lee PA, O'Leary LA, Songer NJ, et al. Paternity after bilateral cryp-torchism: a controlled study. Arch Pediatr Adolesc Med 1997; 151:260. 42. Gracia J, Sanchez Zalabardo J, Sanchez Garcia J, et al. Clinical, physical, sperm and hormonal data in 251 adults operated on for cryptorchidism in childhood. BJU Int 2000; 85:1100. 43. Rogers E, Teahan S, Gallagher H, et al. The role of orchiectomy in the management of postpubertal cryptorchidism. J Urol 1998; 159:851. 44. Winter JSD, Faiman C. Pituitary-gonadal relations in male children and adolescents. Pediatr Res 1972; 6:126. 45. Cassorla FG, Golden SM, Johnsonbaugh RE, et al. Testicular volume during early infancy. J Pediatr 1981; 99:742. 46. Daniel WA, Feinstein RA, Howard-Peebles R, et al. Testicular volumes of adolescents. J Pediatr 1982; 101:1010.

CHAPTER 94 MORPHOLOGY AND PHYSIOLOGY OF THE OVARY Principles and Practice of Endocrinology and Metabolism

CHAPTER 94 MORPHOLOGY AND PHYSIOLOGY OF THE OVARY GREGORY F. ERICKSON AND JAMES R. SCHREIBER Morphology of the Ovary Follicles Interstitial Cells Corpus Luteum Physiology of the Ovary Normal Folliculogenesis Ovulation Ovarian Steroidogenesis Lipoproteins as Cholesterol Source Mechanisms of Gonadotropin Effects Steroid Hormones Produced by the Ovary Effects of Ovarian Steroids Intraovarian Steroid Effects Relaxin Conclusion Chapter References

Under normal conditions, women produce a single dominant follicle that participates in a single ovulation each menstrual cycle. The process begins when a cohort of primordial follicles is recruited to initiate growth. Successive recruitments give rise to a pool of growing follicles (i.e., primary, secondary, tertiary, graafian) in the ovaries. The ability to become dominant is not a characteristic shared by all follicles, and those that lack the property die by atresia. In the human female, only ~400 of the original 7 million follicles survive atresia. Recognition that only a few follicles survive and ovulate their eggs demonstrates the principle that folliculogenesis in mammals is a highly selective process. After the dominant follicle ovulates its ovum, the follicle wall develops into a corpus luteum by a process called luteinization. If implantation does not occur, the corpus luteum is destroyed by luteolysis. This chapter reviews the structure of the various histologic units in the ovary, and analyzes the mechanisms that cause them to change during the menstrual cycle.

MORPHOLOGY OF THE OVARY The human ovary is organized into two principal parts: a central zone called the medulla and a predominant peripheral zone called the cortex (Fig. 94-1). The characteristic feature of the cortex is the presence of follicles, containing the female gamete or oocyte, and the corpus luteum. The number and size of the follicles change as a function of the age and reproductive stage of the female. Another feature of the cortex is the presence of clusters of differentiated steroidogenic cells called secondary interstitial cells. They arise from the theca interna of atretic follicles and remain as androgen-producing cells. Characteristically, the medulla contains blood tissue, nerves, and groups of hilus or ovarian Leydig cells (see Fig. 94-1).

FIGURE 94-1. Morphology of the human ovary. The follicles, corpora lutea, and secondary interstitial cells are embedded in the outer cortex; hilus cells, autonomic nerves, and spiral arteries are found in the medulla. (From Erickson GF, Magoffin D, Dyer CA, et al. The ovarian androgen producing cells; a review of structure/function relationships. Endocr Rev 1985; 6:371.)

FOLLICLES All follicles are located in the cortex, medial to the tunica albuginea, or ovarian capsule. There are two principal classes of follicles: nongrowing and growing. The nongrowing or primordial follicles comprise 90% to 95% of the ovarian follicles throughout the life of the woman. When a primordial follicle is recruited into the pool of growing follicles, its size and position in the cortex change (Fig. 94-2). Typically, the growing follicles are divided into four classes: primary, secondary, tertiary, and graafian (Fig. 94-3; see Fig. 94-2). Intrinsic signals are required for, and are important to, the development of preantral follicles (primary, secondary, early tertiary). Hence, the preantral stages of folliculogenesis are gonadotropin-independent. By contrast, the graafian stages (small, medium, large) are gonadotropin-dependent. The growing follicles that do not participate in ovulation undergo apoptosis (programmed cell death) and become atretic follicles.

FIGURE 94-2. Photomicrographs of the adult ovary. A, High magnification of cortex shows nongrowing primordial follicles and their recruitment into the growing pool of preantral follicles. Notice the dramatic increase in oocyte size and the progressive migration of growing follicles toward the medulla. (se, surface epithelium; ta, tunica albuginea; pf, primordial follicle; prf, primary follicle; sf, secondary follicle.) B, Low magnification, showing diversity of follicles. Notice the follicle migration into the medulla, presumably by morphogenetic activities in the theca cone (tc). (gf, graafian follicles 1, 2, 3, 4; ca, corpus albicans; af, atretic follicle; m, medulla.)

FIGURE 94-3. Architecture and classification of ovarian follicles during development: preantral (gonadotropin-independent) stages: primary, secondary, early tertiary; antral or graafian (gonadotropin-dependent). Recruitment occurs within the pool of primordial follicles, and selection of the dominant preovulatory follicle occurs at the graafian stage, when the follicle is ~5 mm in diameter. (From Erickson GF, Magoffin D, Dyer CA, et al. The ovarian androgen producing cells; a review of structure/function relationships. Endocr Rev 1985; 6:371.)

PRIMORDIAL FOLLICLE The ability of a woman to have a menstrual cycle totally depends on having a pool of primordial follicles. Consequently, primordial follicles represent the fundamental reproductive units of the ovary. Histologically, primordial follicles possess a simple organization: a small oocyte arrested in diplotene of the first meiotic prophase, a surrounding layer of follicle cells (i.e., future granulosa cells), and a basal lamina (see Fig. 94-3). Primordial follicles do not have a theca and therefore do not have an independent blood supply.1 All primordial follicles are formed in the fetal ovaries2 at between 6 and 9 months of gestation (Fig. 94-4). Because each germ cell has entered meiosis, there are no gametes capable of dividing mitotically. All oocytes capable of participating in reproduction during a woman's life are formed before birth. In human females, recruitment (i.e., the initiation of primordial follicle growth) begins in the fetus and continues until menopause.2 As a result of recruitment, the size of the pool of primordial follicles becomes progressively smaller with age; between birth and menarche, the number of primordial follicles decreases from several million to several hundred thousand (see Fig. 94-4). The number of primordial follicles continues to decline until they are relatively rare at menopause.3

FIGURE 94-4. Changes in the pool of oocytes in human ovaries during aging. A, Stages of meiosis in human fetal ovaries leading to formation of primordial follicles: (1) At 3 months, oogonia are engaged in mitosis, and a few germ cells deep within the cortical cords enter meiosis; (2) at 4 months, more oocytes enter meiosis; (3) at 7 months, the cords are no longer distinct, and all germ cells are in meiotic prophase; (4) at 9 months, the cortex is packed with individual primordial follicles. (From Ohno S, Klinger HP, Atkin NB, et al. Human oogenenic. Cytogenetics 1962; 1:42.) B, Changes in the total number of germ cells in human ovaries during aging. (From Baker TG, Sum OW. Development of the ovary and oogenesis. Clin Obstet Gynecol 1967; 3:3.) C, The photomicrographs show a progressive decrease in the number of primordial follicles at different periods in a woman's life.

PRIMARY FOLLICLE A primary follicle contains a growing oocyte surrounded by one layer of granulosa cells (see Fig. 94-2 and Fig. 94-3). The process of primary follicle formation begins when the squamous granulosa cells round up and appear cuboidal.4 After this occurs, the meiotic chromosomes enter the lampbrush state, and the oocyte begins to increase in size by virtue of increased RNA and protein synthesis.2,5 Small patches of oocyte-derived material appear between granulosa cells. Eventually, this extracellular matrix (i.e., zona pellucida [ZP]) covers the entire oocyte. By the late primary stage, the oocyte, encapsulated by the ZP, is almost full-grown (~100 µm in diameter). The human ZP is composed of three glycoproteins termed ZP-1, -2, and -3.6 The ZP-3 glycoprotein functions as the primary sperm receptor and induces the acrosome reaction.7 Anti–ZP-3 antibodies can block fertilization, and attempts are under way to utilize ZP-3 as an immunogen to develop a human contraceptive vaccine.8,9 The development of primary follicles leads to an increase in the number and size of gap junctions between the granulosa cells (Fig. 94-5) as well as between the granulosa and the oocyte.10 Gap junctions consist of a family of proteins called connexins (Cx).10,11 In the case of animal follicles, Cx43 is the major gap junction protein between granulosa cells,10 while Cx37 is the major gap junction protein between the oocyte and granulosa cells.12 Cx37 is an oocyte-derived protein; results from studies of Cx37-deficient mice have shown that Cx37 is obligatory for folliculogenesis and fertility.12 The role of Cx37 in human ovary physiology and pathophysiology remains to be determined.

FIGURE 94-5. Electron micrograph shows the structure of gap junctions (arrows) between granulosa cells of a healthy graafian follicle. Inset, Replica of granulosa cell fracture shows the hexagonally ordered connexin proteins of the gap junction. (RER, rough endoplasmic reticulum.) (Courtesy of Dr. David Albertini, Tufts University, Boston, MA.)

SECONDARY FOLLICLE A secondary follicle contains two to eight layers of granulosa cells with no antrum. During secondary follicle development, the granulosa cells proliferate slowly,4 and the oocyte completes its final growth.13 By the end of the secondary stage, the follicle is a multilayered structure that is strikingly symmetric; in the center is a full-grown

oocyte (~120 µm in diameter), eight layers of stratified low columnar granulosa cells, and a basal lamina (Fig. 94-6). When the follicle has two to three layers of granulosa cells, a signal (yet to be identified) is generated that causes a stream of mesenchymal cells to migrate toward the basal lamina.14 They become organized into a layer of fibroblast-like cells (see Fig. 94-6) that ultimately develops into the theca interna and the theca externa. At about this time, the secondary follicle acquires a set of capillaries. The vessels form two sets of interconnected capillaries, an inner wreath located in the theca interna, which is supplied by branches from an outer wreath located in the theca externa.1 Call-Exner bodies develop among the granulosa cells in the secondary follicle (see Fig. 94-6). Histologically, these bodies appear to be made of extracellular matrix. The physiologic function of Call-Exner bodies is unknown; however, given the importance of the extracellular matrix in proliferation and cytodifferentiation,15 they may play a role in generating subtypes of granulosa cells by providing novel substrate-to-cell interactions.

FIGURE 94-6. Photomicrograph of a fully grown secondary follicle with six to eight layers of granulosa cells. (TE, theca externa; TI, theca interna; ZP, zona pellucida; *, germinal vesicle or egg nucleus; arrowheads, cytoplasmic process of corona radiata granulosa cells traversing the ZP.) (Adapted from Anderson E. The ovary: basic principles and concepts. In: Felig P, Baxter JD, Broadus AE, Frohman LA, eds. Endocrinology and metabolism, 3rd ed. New York: McGraw-Hill, 1994.)

TERTIARY FOLLICLE A characteristic feature of a tertiary follicle is the antrum.16 When a follicle reaches ~400 µm in diameter, follicular fluid accumulates between some granulosa cells. This results in the formation of a small cavity or antrum at one pole of the oocyte (see Fig. 94-3). The initiation of antrum formation is controlled by the follicle itself, but the nature of the regulatory factors remains unknown. As a consequence of beginning antrum formation, the follicle assumes a symmetry that remains throughout folliculogenesis (see Fig. 94-2 and Fig. 94-3). Simultaneously, histologic changes are initiated in the theca interna.14 Subpopulations of fibroblasts transform into large epithelial-like cells called theca-interstitial cells (see Fig. 94-3), which produce a variety of ligands, most notably androgens in response primarily to luteinizing hormone (LH) and insulin stimulation.17 GRAAFIAN FOLLICLE The morphology of the early tertiary and graafian follicle is similar except that the graafian follicle is larger. In women, a graafian follicle can increase as much as 75-fold in diameter, from 0.4 mm to 30 mm.18 The tremendous growth is caused by follicular fluid accumulation and proliferation of the granulosa and theca cells.16,18 Follicular fluid is an exudate of plasma plus various regulatory factors produced by the follicle cells themselves.19 The follicular fluid is the medium in which the granulosa cells are found and through which regulatory molecules must pass on their way to and from the microenvironment. By virtue of the structure of the follicle, the granulosa cells become different from one another with respect to their position in the system (see Fig. 94-3). There are four granulosa cell domains16: Granulosa cells forming the corona radiata make contact with the oocyte and ZP, those comprising the cumulus make contact with the corona and membrana granulosa cells, and those forming the membrana granulosa make contact with the basal lamina and Call-Exner bodies (Fig. 94-7). It has been shown that the position of the granulosa cells determines the direction in which they differentiate in response to follicle-stimulating hormone (FSH).16 For example, the membrana, but not cumulus, granulosa cells express the P450 enzyme aromatase (P450AROM) and LH receptor in response to FSH stimulation (see Fig. 94-7). The significance of granulosa heterogeneity is unknown. Nonetheless, it is becoming clear that oocyte-derived regulatory proteins determine the way in which the granulosa cells differentiate.20

FIGURE 94-7. Diagram of the heterogeneity of the granulosa cells in a healthy graafian follicle. By virtue of their position or location in the follicle wall, the granulosa cells express different patterns of proliferation and cytodifferentiation in response to follicle-stimulating hormone stimulation. (cAMP, cyclic adenosine monophosphate; LH, luteinizing hormone; HSD, hydroxysteroid dehydrogenase.) (From Erickson GF. The graafian follicle: a functional definition. In: Adashi EY, ed. Ovulation: evolving scientific and clinical concepts. New York: Springer-Verlag, [in press 1999].)

During the small (1–5 mm), medium (6–11 mm), and large (>12 mm) stages of graafian follicle development, the number of theca-interstitial cells increases progressively until there are five to eight layers in the dominant preovulatory follicle.4,18 The increase results from mitosis, presumably within a population of undifferentiated stem cells. The theca externa (see Fig. 94-1 and Fig. 94-3) is composed of smooth muscle cells, and these contractile cells are innervated by the autonomic nervous system.16,21 ATRETIC FOLLICLE After a primordial follicle is recruited, it develops to the pre-ovulatory stage or dies by atresia. Atresia results in the death and removal of the granulosa and the oocyte by a programmed cell death mechanism22,23 and 24 called apoptosis. The discovery that atresia involves the activation of physiologic cell suicide mechanisms has opened up new ways to investigate the regulation and mechanisms underlying follicle death. Although the field is still in its infancy, it is clear that FSH is a major suppressor of apoptosis in granulosa cells.22,23 The challenge is to understand the nature of the physiologic ligands that activate apoptosis in follicles during the menstrual cycle and in so doing understand the basis for selection of the dominant follicle. It is noteworthy that the theca cells appear to survive atresia, becoming islands of secondary interstitial cells.14 Atresia is evident at all stages of graafian follicle growth, and at all stages of the cycle; however, atresia is rare or absent in the nongrowing primordial follicles and difficult to detect in the pool of preantral follicles.4 Morphologic data suggest that the sequence of atresia may be different for preantral and antral follicles.14 Atresia of preantral follicles is most readily identified by precocious antrum formation (Fig. 94-8A); however, premature meiotic maturation and fragmentation of the oocyte are also seen in preantral atresia.25 In the graafian follicles, the earliest morphologic sign of atresia involves a major shape change in the granulosa cells, which are attached to the basal lamina and the ZP.26 In healthy graafian follicles,16 the membrana granulosa cells consist of a uniform layer of pseudostratified epithelial cells, all of which are attached to the basal lamina. At or about the time atresia is initiated, these cells contract and become a simple stratified cuboidal epithelium (see Fig 94-8B).

FIGURE 94-8. Photomicrographs showing the early histologic changes of atresia in preantral and graafian follicles. A, Preantral follicle (secondary stage), with three layers of granulosa cells (sc). A cavity (*) is already present, a phenomenon called precocious antrum formation. (ZP, zona pellucida; t, theca.) B, Graafian follicle. The membrana granulosa (outer layer) has contracted into a cuboidal epithelium that appears as a distinct bead of cells (arrowheads) around the periphery of the follicle. (ti, theca interna; te, theca externa; a, antrum.)

The cause of this coordinated cell-contraction mechanism is unknown. A similar phenomenon occurs in the corona radiata cells. Another morphologic sign of early atresia in graafian follicles is that some of the periantral granulosa cells (those at the border of the antrum) lose contact with one another and are released into the follicular fluid, where they become apoptotic. As atresia progresses, the number of apoptotic cells increases to the point that no healthy granulosa cells are visible. Oocyte death occurs relatively late in the process of graafian follicle atresia by apoptotic mechanisms.23 INTERSTITIAL CELLS THECA-INTERSTITIAL CELLS Theca-interstitial cells are located in the theca interna of all graafian follicles. Morphologically, they have the ultrastructure that is typical of active steroidogenic cells.14 In the human ovary, they are the primary site of androstenedione biosynthesis.18,27 Shortly after presumptive theca cells reach the secondary follicle, some begin a theca-interstitial cytodifferentiation.14,28 Most notably, this developmental process involves the acquisition of the steroid acute regulatory protein (StAR), a protein that transfers cholesterol from the outer to the inner mitochondrial membrane; cholesterol side-chain cleavage enzyme (P450c22); 3b-hydroxysteroid dehydrogenase-D4,5 isomerase enzyme (3b-HSD); and the LH receptor.14,28 The initial differentiation of these cells is regulated by local autocrine/paracrine mechanisms operating within the secondary follicle. In response to LH delivered by the theca capillaries, the interstitial cells transform from elongated mesenchymal cells into large epithelial-like cells that produce progesterone.14,28 As the preantral follicle grows to the graafian stage, the interstitial cells express the 17a-hydroxylase C17–20 lyase enzyme (P450c17) and transform from progesterone-producing cells to cells that produce androstenedione (Fig. 94-9). As discussed later, there is a causal relationship between this switch into an androgen-producing cell and the ability of the developing follicle to produce aromatase substrate and, thus, estradiol.

FIGURE 94-9. Flow pathway diagram of steroidogenesis of human interstitial cells during folliculogenesis. The first step is the conversion of prothecoblasts into theca-interstitial cells, which produce progesterone, which occurs in secondary follicles. The second step consists of luteinizing hormone (LH) induction of P450c17, which results in the production of androstenedione. It begins in early tertiary follicles and continues through graafian follicle development (healthy and atretic). The third step is a switch back to a progesterone-producing cell, which occurs at ovulation. The fourth step is the reinduction of P450c17 in theca-lutein cells during luteinization, which results in androstenedione production. (Adapted from Erickson GF. Normal regulation of ovarian androgen production. Semin Reprod Endocrinol 1993; 11:307.)

It should be noted that an interesting role for insulin and lipoproteins in stimulating human theca progesterone and androstenedione production has been demonstrated by in vitro studies.29 Thus, a much more complex endocrine regulation of theca-interstitial cells is emerging (Fig. 94-10).

FIGURE 94-10. Diagram of the luteinizing hormone (LH) signal transduction pathway in differentiated theca-interstitial cells leading to androstenedione biosynthesis. Other regulatory molecules including insulin and low- and high-density lipoproteins (LDL, HDL) can interact with the LH signaling pathway to increase steroidogenesis further. (GDP, guanosine diphosphate; GTP, guanosine triphosphate; PTK, protein tyrosine kinase; ATP, adenosine triphos-phate; cAMP, cyclic adenosine monophosphate; R, receptor; StAR, steroid acute regulatory protein; HSD, hydroxysteroid dehydrogenase.) (Redrawn from Erickson GF. Normal regulation of ovarian androgen production. Semin Reprod Endocrinol 1993; 11:307.)

SECONDARY INTERSTITIAL CELLS The secondary interstitial cells arise as a consequence of atresia.14 Secondary interstitial cells maintain their specialized ultrastructure and can respond to LH with increased androstenedione production.14,18 One difference between the theca and the secondary interstitial cells is that the latter are the only endocrine cells in the ovary that are innervated14 (see Fig. 94-1).

Animal studies suggest that there is a point-to-point communication between neurons in the hypothalamus and the ovarian steroidogenic cells.14,30 Moreover, in vitro studies indicate that catecholamines directly stimulate androgen synthesis in secondary interstitial cells.14 There is strong evidence for neural and neurotrophic control of androgen production by ovarian secondary interstitial cells.14,31 Evidence supports the proposition that the nervous system may play an important role in the physiology and pathophysiology of ovarian androgen production. CORPUS LUTEUM After ovulation (see Fig. 94-1), the follicle wall transforms into the corpus luteum (Fig. 94-11). Cells that make up the corpus luteum are contributed by the membrana granulosa, theca interna, theca externa, and invading blood tissue. The white blood cells produce potent regulatory ligands, such as the cytokines, that can regulate ovulation and corpus luteum function.32 Understanding how white blood cells influence these critical processes is a major goal of ovary research. Morphologically, there is a fibrin clot where the antrum and liquor folliculi were located, into which loose connective tissue and blood cells have invaded (see Fig. 94-11).

FIGURE 94-11. Photomicrographs of a section of human corpus luteum. A, Fibrin clot has formed in antrum, and collapsed follicle wall is composed of granulosa and theca-lutein cells. B, Theca externa, theca interna, and granulosa lutein tissues are readily distinguishable. (Courtesy of Dr. T. Crisp, USEPA, Washington, DC.)

During luteinization, the membrana and periantral granulosa cells attain a large size, ~35 µm in diameter.33 These cells, now called granulosa lutein cells, have an ultrastructure typical of differentiated steroidogenic cells; they contain abundant smooth endoplasmic reticulum, tubular cristae in the mitochondria, and large clusters of lipid droplets containing cholesterol esters in the cytoplasm (Fig. 94-12).

FIGURE 94-12. Electron micrograph of a section through a human granulosa lutein cell shows abundant rough and smooth endoplasmic reticulum, which synthesize proteins and steroids, respectively, and numerous lipid droplets composed of stored cholesterol esters. Notice the tubular cristae, the site of cholesterol side-chain cleavage enzyme (P450c22) in mitochondria. (Courtesy of Dr. T. Crisp, USEPA, Washington, DC.)

The theca-interstitial cells also are incorporated into the corpus luteum, becoming the theca-lutein cells (see Fig. 94-12). They can be distinguished from granulosa lutein cells because they are smaller (~15 µm in diameter) and stain more darkly.33 Theca-lutein cells also exhibit the ultrastructure of active steroid-secreting cells.33 During the ovulatory phase, the theca cells lose the P450c17 and become active in progesterone production (see Fig. 94-9). These cells reacquire the P450c17 after luteinization and once again produce androstenedione (see Fig. 94-9). By virtue of the expression of androstenedione and P450AROM in the theca and granulosa lutein cells, respectively,34 the corpus luteum also synthesizes and secretes estradiol. If implantation does not occur, the corpus luteum degenerates. This process, called luteolysis, becomes apparent histologically at 8 days after ovulation. The first histologic indication of luteolysis is shrinkage of the granulosa lutein cells. The theca-lutein cells appear selectively hyperstimulated during early luteolysis, analogous to the theca-interstitial cell hypertrophy associated with atresia. After day 23 of the cycle, apoptosis is activated,35 and the corpus luteum dies. Histologically, all that remains is a nodule of dense connective tissue called the corpus albicans (see Fig. 94-1 and Fig. 94-2). The mechanism of luteolysis in women is poorly understood, but it has been proposed that prostaglandin F2a might be involved.36,37

PHYSIOLOGY OF THE OVARY In women, the cyclic changes that occur in the menstrual cycle reflect structural and functional changes that occur within the follicle and corpus luteum. The dominant follicle begins as a primordial follicle and is slowly prepared for ovulation and luteinization by the action of the pituitary gonadotropins and ovarian growth factors. To understand the relationship of the dominant follicle to the events occurring in the menstrual cycle, the underlying physiologic mechanisms of folliculogenesis—recruitment, selection, atresia, ovulation, and luteogenesis (i.e., luteinization, luteolysis)—must be considered. NORMAL FOLLICULOGENESIS Follicular growth and development in women is a very long process.4 In each menstrual cycle, the ovulating follicle originates from a primordial follicle that was recruited to grow ~1 year earlier (Fig. 94-13). At first, the recruited primordial follicle develops very slowly, requiring ~270 days to complete the preantral period and grow to the early tertiary stage (~0.4 mm). The basis of this slow growth is the very long doubling time (~250 hours) of the granulosa cells.38

FIGURE 94-13. Chronology of folliculogenesis in the human ovary. Follicle development is typically divided into two major periods. During the preantral period, the recruited primordial follicle develops to the early antral (tertiary) stage (class 2). Antrum formation occurs at this point, and the graafian follicle enters the antral period.

Small antral (0.5–5.0 mm, class 4, 5), midantral (6–10 mm, class 6), large antral (10–15 mm, class 7), preovulatory (16–20 mm, class 8). The total time required for completion of preantral and antral periods is 355 days. The number of granulosa cells (GC), the follicle diameter in millimeters, and the percentage of atretic follicles in each class are indicated. (From Gougeon A. Dynamics of follicular growth in the human: a model from preliminary results. Hum Reprod 1986; 2:81.)

When FSH enters the microenvironment of the early tertiary follicle, follicular fluid production by the granulosa cells is increased, and the graafian follicle begins to expand.18 During the antral period, a graafian follicle may pass through the small, medium, and large stages (see Fig. 94-13). The follicles that survive to ovulate require ~85 days to complete the antral period (see Fig. 94-13). Selection of the dominant follicle is one of the last steps in this long process. The dominant follicle is selected from a cohort of rapidly growing small graafian follicles in the late luteal phase of the menstrual cycle.4,38,39 It requires ~15 to 19 days for the dominant follicle to complete its growth to the ovulatory stage (see Fig. 94-13). The 99.9% of all growing follicles that are not selected die by atresia (see Fig. 94-13). RECRUITMENT In women, recruitment is a continuous process throughout life, and the mean age for total follicular exhaustion is ~51 years of age.3,40,40a The first indication that a primordial follicle has been recruited to grow is that the granulosa cells transform from a squamous to a cuboidal shape.41 As the granulosa cells round up, they acquire the ability for DNA synthesis and division, albeit at a very slow rate.41 When more than 90% of the granulosa cells are cuboidal, there occurs a dramatic increase in RNA synthesis in the arrested oocytes.25 The increased transcription and translation lead to the marked growth of the oocyte that occurs during preantral follicle development (see Fig. 94-3). The fact that changes in the oocyte begin later than those in the granulosa suggests that the granulosa cells might produce or respond to a ligand that initiates the recruitment of the primordial follicle into the growing pool of follicles. What is known about the mechanisms of recruitment? The evidence that recruitment continues in the absence of pituitary gonadotropins argues that the process is regulated by intrinsic ovarian factors.3 Based on experiments in animals, it appears that the number of recruited primordial follicles varies with age—the highest level of recruitment occurs early in life, after which it decreases progressively with advancing age. This implies that the rate of recruitment is somehow determined by the actual number of primordial follicles and that the rate of recruitment can be suppressed by testosterone, thymectomy, starvation, and opioid peptides.3 Although inconclusive, these results suggest that recruitment is an active process that can be modulated (i.e., inhibited) by ligands. What evokes or triggers a particular follicle to grow is totally obscure. It is noteworthy that a monotropic rise in FSH (perhaps due to reduced plasma inhibin levels) occurs in women during aging, and this increase in FSH coincides with an accelerated loss in primordial follicles or ovary reserve (OR).3 In this regard, studies of aging rats indicate that the monotropic rise in plasma FSH may be involved in the accelerated loss of OR in old rats.3 This phenomenon could have clinical implications, because reduced OR leads to reduced fecundity in older women. Precisely how FSH might accelerate primordial follicle (OR) loss during aging is unknown, but it may involve the premature expression and/or activation of FSH receptors in granulosa cells of primordial follicles in aging ovaries. SELECTION Morphometric studies of normal human ovaries indicate that the dominant follicle that will ovulate its egg the next cycle is selected from a cohort of small graafian follicles (4.7 ± 0.7 mm in diameter) at the end of the luteal phase of the menstrual cycle.4,38,39 After the midluteal phase, there is an approximate doubling of the rate of mitosis in the granulosa cells of all cohort follicles. This suggests that the demise of the corpus luteum is followed by a dramatic stimulation of mitosis and granulosa cell division in the cohort follicles. The first visible sign that one of the follicles has been selected is that the granulosa cells of the chosen follicle maintain a high rate of mitosis, but the mitotic rate falls significantly in the other follicles of the cohort.4,38,39 This change becomes evident in the late luteal phase. The newly selected dominant follicle continues to grow and expand during the follicular phase and at a relatively rapid rate: 6.9 ± 0.5 mm at days 1 to 5; 13.7 ± 1.2 mm at days 6 to 10; and 18.8 ± 0.5 mm at days 11 to 14. The growth is caused by a progressive increase in follicular fluid and granulosa cell number.18 As the dominant follicle undergoes its growth and development, the cohort of nondominant follicles becomes increasingly more atretic, and rarely does an atretic follicle reach ³9 mm in diameter over the cycle.39 What do we know about the mechanisms underlying the selection process? There is compelling evidence that the secondary rise in FSH plays a central role in the selection process in rats.42 In women, the secondary rise in plasma FSH begins at or about the time plasma progesterone levels fall to basal levels at the end of the luteal phase, and it continues through the first week of the follicular phase43 (see Fig. 94-14). Evidence that suppression of the secondary rise in FSH prevents ovulation in monkeys supports the proposition that the secondary increase in FSH is critical for the continued growth and development of the dominant follicle in the primate.44 It seems likely that the secondary rise in FSH is also critical for selection in women, but further proof is needed.

FIGURE 94-14. The secondary follicle-stimulating hormone (FSH) rise during the luteal-follicular transition. Data are mean (±SEM) of daily FSH, estradiol, progesterone, and inhibin A and B of normal cycling women (n = 5). Data are centered to the day of menses in the cycle. (From Welt CK, Martin KA, Taylor AE, et al. Frequency modulation of follicle stimulating hormone [FSH] during the luteal-follicular transition: evidence for FSH control of inhibin B in normal women. J Clin Endocrinol Metab 1997; 82:2645.)

During this period, the concentration of FSH increases steadily in the microenvironment of the chosen follicle, and FSH levels become low or absent in the nondominant follicles.45 In the dominant follicle, the FSH stimulates a sharp increase in the number of granulosa cells and apoptosis is suppressed. By contrast, the growth development of nondominant follicles is suppressed in the absence of adequate levels of FSH, and apoptosis is activated in the granulosa cells. In this way, the selected follicle achieves dominance. How is dominance established? In studies in the monkey, estradiol has been found to be a causative agent by virtue of negative feedback on the pituitary gonadotrope.42,44 In this concept, the high intrafollicular levels of FSH lead to increased estradiol production, which, in turn, suppresses FSH secretion by the pituitary. When FSH becomes rate-limiting in the nondominant cohort follicles, they undergo atresia.38 Human menopausal gonadotropins can stimulate granulosa cells to divide mitotically in nondominant follicles during the early follicular phase.38 If FSH levels within the microenvironment are increased, it appears that the nondominant follicle can be rescued from atresia.38 This rescue phenomenon could be involved in the generation of multiple large follicles in women undergoing ovulation induction with exogenous FSH. The FSH rise that begins before the onset of menses also occurs with concomitant decreases in inhibin A produced by the corpus luteum.43 The inverse relationship between inhibin A and FSH suggests an endocrine role for inhibin A in the regulation of the secondary FSH rise (see Fig. 94-14). In contrast to inhibin A, inhibin B increases before menses, coincident with the FSH rise (see Fig. 94-14). Although estradiol secretion by the follicle is primarily responsible for the negative-feedback regulation of FSH during the follicular phase of the cycle, it is possible that inhibin B produced by graafian follicles might play a role as well. However, the role of inhibin B in reproduction remains unknown. It should be mentioned that there is considerable interest in follistatin because it can specifically bind to activin and inhibin and modulate their activity.46 Follistatin might, therefore, be a physiologically relevant protein in the control of folliculogenesis in the ovary. However, further work is needed

to establish this concept. TWO-CELL, TWO-GONADOTROPIN CONCEPT Because the estradiol produced by the selected follicle plays an important role in establishing follicle dominance,42 an understanding of the mechanisms of follicular estradiol production is important. The process requires two cell types (i.e., theca and granulosa) and two gonadotropins (i.e., LH and FSH); it is called the two-cell, two-gonadotropin concept for follicle estrogen synthesis (Fig. 94-15).

FIGURE 94-15. Diagram of the two-gonadotropin, two-cell concept of follicle estrogen biosynthesis. G proteins include aG-stimulatory (aGs), b, g, A kinase, cyclic adenosine monophosphate (cAMP)–dependent protein kinase A. (GTP, Guanine nucleotide triphosphate; GDP, Guanine nucleotide diphosphate; ATP, adenosine triphosphate.) (From Kettel LM, Erickson GF. Basic and clinical concepts of ovulation induction. In: Rock J, Alvarez-Murphy A, eds. Advances in obstetrics and gynecology. Chicago: Mosby, 1994.)

When FSH and LH interact with transmembrane receptors in the granulosa and theca-interstitial cells, respectively, the binding events are transduced into intracellular signals by means of the heterotrimeric G proteins. The LH-bound receptor is coupled to the aG-stimulatory (aGs), cyclic AMP (cAMP), protein kinase A (PKA) pathway. The stimulation of this signal transduction pathway in the theca-interstitial cells leads to increased transcription of those genes involved in de novo androstenedione biosynthesis28 (see Fig. 94-15). The FSH-bound receptor activates the aG s/cAMP/PKA pathway in membrana granulosa cells (see Fig. 94-15), and the signal promotes the stimulation of the genes encoding P450AROM and 17b-hydroxysteroid dehydrogenase (17b-HSD) type 1 enzyme,47,48 which then results in the aromatization of androstenedione to estradiol. Because the dominant follicle contains large numbers of granulosa cells and relatively high levels of FSH, it is capable of producing large quantities of estradiol. Although nondominant follicles produce a high level of androstenedione, they have a paucity of granulosa cells and microenvironment FSH and, thus, produce very little estradiol. The LH or human chorionic gonadotropin receptor (LH/hCG receptor) has been cloned, and its structure is similar to that of other G protein–coupled receptors.49 The N terminus, which is the extracellular domain of the receptor, is glycosylated and binds circulating LH or hCG. As with other G protein–coupled receptors, this LH/hCG receptor contains seven membrane-spanning domains. The C-terminal domain is located intracellularly and is responsible for signal generation that begins with the activation of G proteins. The FSH receptor is structurally similar.50 The extracellular domain binds FSH, there are seven membrane-spanning domains, and a short C-terminal cytoplasmic domain activates the heterotrimeric G protein. The mechanism by which an increase in cAMP leads to increased gene expression is now understood in many instances. cAMP activates the catalytic subunit of protein kinase A. The kinase A phosphorylates cAMP-response element–binding protein or other related DNA-binding proteins.51 When such proteins are phosphorylated, they bind to upstream DNA regulatory elements called cAMP response elements. The binding has been shown to increase gene transcription and the production of the LH/FSH-responsive proteins. The interaction among these various regulatory proteins is exceedingly complex but is described clearly in a review.51 FOLLICLE-STIMULATING HORMONE REGULATION OF MITOSIS AND THE LUTEINIZING HORMONE RECEPTOR Several other physiologically important effects of FSH occur in the dominant follicle, most notably the stimulation of mitosis and expression of LH receptor in the granulosa cells (Fig. 94-16). Based on in vivo4,38 and in vitro52,53 work, it appears that the rate of human granulosa cell division is stimulated directly by FSH in the human. Precisely how this occurs is unclear, but studies with growth factors indicate that human granulosa cells cultured in vitro respond to potent mitogens, such as fibroblast growth factor and epidermal growth factor, resulting in dramatic increases in mitosis.54 The question of whether growth factors mediate the FSH-induced proliferation of human granulosa cells in vivo is an interesting, but unresolved, question. The ability of the dominant follicle to respond to the LH surge with ovulation depends on the expression of high levels of LH receptor in the membrana granulosa cells (see Fig. 94-7 and Fig. 94-16). Direct evidence that FSH induces LH receptors in the primate has been provided by in vitro experiments with monkey granulosa cells.55 There is some evidence favoring this concept in women.56 In examining the level of granulosa LH receptor during the follicular phase, the researcher finds that the number is low in small and medium graafian follicles but increases sharply to very high levels at the preovulatory stage.57,58 Unlike the early effects of FSH on P450AROM enzyme and mitosis, the FSH control of LH receptor appears to be restricted to the late stages of folliculogenesis. How the stage-specific effects of FSH are achieved is unknown.

FIGURE 94-16. Diagram of the follicle-stimulating hormone (FSH) signaling pathway in granulosa cells of a dominant follicle that result in proliferation, steroid biosynthesis, and luteinizing hormone (LH)/human chorionic gonadotropin receptor expression. (GTP, guanosine triphosphate; GDP, guanosine diphosphate; cAMP, cyclic adenosine monophosphate; HSD, hydroxysteroid deoxygenase.) (From Erickson GF. Polycystic ovary syndrome: normal and abnormal steroidogenesis. In: Schats R, Schoemaker J, eds. Ovarian endocrinopathies. Proceedings of the 8th Reinier deGraaf Symposium. UK: Parthenon Publishing, 1994.)

It is clear that the ability of the ovary to generate a dominant follicle depends on having sufficient amounts of FSH within the microenvironment and that the FSH functions in stimulating granulosa cell division and differentiation. A fundamental concept of ovarian physiology is that FSH is obligatory for dominant follicle formation and that no other ligand by itself can serve in this regulatory capacity. GROWTH FACTOR CONCEPT One exciting and important concept to emerge in ovarian physiology is the awareness that folliculogenesis and luteogenesis are modulated by proteins that are produced by the ovaries themselves.59,60,60a The evidence has led to the novel idea that the actions of hormones (FSH, LH, progesterone, androgen, and estrogen) can be modulated, either amplified or attenuated, by ovary growth factors that act in autocrine/paracrine manners to control proliferation, differentiation, and apoptosis (Fig. 94-17).

FIGURE 94-17. Growth factor or autocrine/paracrine concept. (From Erickson GF. Ovarian control of follicle development. Am J Obstet Gynecol 1995; 172:736.)

Growth factors are regulatory proteins that control a wide variety of proliferative and developmental functions (see Chap. 173). All of them are ligands that interact with specific receptors in target cells, and the binding events generate signal transduction pathways that modulate cellular responses. All of the growth factors share the property of being modulators; they increase or decrease the responsiveness of target cells to ligands (e.g., hormones, growth factors, neurotransmitters). The results of a large number of studies have demonstrated that all five families of growth factors are expressed in the rat follicle, and there is increasing evidence for growth factors in the human ovary.59,60 The potential importance of this rapidly emerging field is illustrated by the gene knock-out studies in mice, which have demonstrated that specific growth factors are essential for FSH-dependent folliculogenesis and female fertility. For example, loss of function of insulin-like growth factor-I (IGF-I)61 and oocyte-derived growth differentiation factor-9 (GDF-9)62 results in the cessation of folliculogenesis at the preantral stage, and the females are infertile. Thus, it is becoming increasingly clear that ovary growth factors are fundamental players in female reproduction. The current challenges are to understand how specific growth factors affect ovarian function and how these actions are integrated into the overall effects of FSH and LH. The presence within the ovary of potent positive and negative regulatory proteins that function to modulate cell function could have far-reaching implications for physiology and pathophysiology in women; however, definitive evidence for an obligatory role of a growth factor in human fertility is still lacking. OVULATION The expulsion of a mature oocyte from the ovary is tightly coupled to the generation of proteolytic activity.63 This process occurs in a highly localized area called the stigma (Fig. 94-18). Morphologic and biochemical studies have shown that, during the ovulatory period, the surface epithelial cells in the presumptive stigma become filled with lysosome-like inclusions.64 With increasing time, the inclusions fuse with the plasma membrane and release their contents toward the tunica albuginea. This process is accompanied by the progressive destruction of the basement membrane and the theca layers. In this way, the steps leading to the formation of the stigma are initiated in a specialized population of surface epithelial cells and involve the release of hydrolytic enzymes. How does this event occur?

FIGURE 94-18. Progressive hormone-induced changes in the dominant follicle during ovulation. The preovulatory surge of follicle-stimulating hormone (FSH) causes cumulus expansion and participates in ovulation by virtue of stimulating plasminogen activator production. The preovulatory luteinizing hormone (LH) surge induces meiotic maturation, luteinization, and stigma formation; the latter depends on intrinsic progesterone and prostaglandin production. (From Erickson GF. The ovary: basic principles and concepts. In: Felig P, Baxter JD, Broudus AE, Frohman LA, eds. Endocrinology and metabolism, 3rd ed. New York: McGraw-Hill, 1994.)

The most important stimulating force in ovulation is the preovulatory surge of LH.63 Although the basic mechanisms involved in LH-induced ovulation are still under investigation in women, some insights have been generated from studies carried out in rats. First, the LH surge starts the preovulatory follicle on the path of progesterone production. An important concept is that increased progesterone is obligatory for ovulation and that the progesterone response is mediated by the progesterone receptor induced in the follicle by the preovulatory surge of LH.65,66 Thus, progesterone plays an essential physiologic role in the mechanism of ovulation, in part by acting as a mediator of LH action. Second, prostaglandins (i.e., PGE and PGF) are required for ovulation. After the ovulatory surge of LH and the stimulation of progesterone production, the synthesis of PGE and PGF is increased in the preovulatory follicle. If the follicle is injected with indomethacin or PG antibodies, ovulation is completely blocked. Furthermore, knocking out the key rate-limiting enzyme in PG synthesis, PG synthetase,67 blocks ovulation, making the female mice infertile.68,69 Morphologic studies of indomethacin-treated ovaries suggest that the prostaglandins are involved in stigma formation. Collectively, the data support the proposition that the elevated level of progesterone induced by LH serves to activate PG production, which, in turn, promotes the release of hydrolytic enzymes by a subpopulation of surface epithelial cells, which then causes stigma formation (see Fig. 94-18). Another active protease relevant to ovulation is plasmin.70 The follicular fluid of preovulatory follicles contains the plasmin precursor, plasminogen. The granulosa cells are stimulated specifically by FSH to release plasminogen activator, which converts plasminogen to the active protease, plasmin (see Fig. 94-18). After the process is initiated, holes are formed in the basal lamina, and there is a general weakening of the follicular wall, presumably caused by the proteolytic action of plasmin. More work is needed to elucidate the physiologic importance of plasminogen activator in ovulation in women. High levels of LH at midcycle stimulate meiotic maturation, and the oocyte reaches the second meiotic metaphase or first polar body stage (Fig. 94-19). During this process, the cumulus granulosa cells undergo a series of structural and functional changes called mucification or expansion. The preovulatory surge of gonadotropins induces the granulosa cells in the cumulus to secrete a hyaluronidase-sensitive mucous substance.71 This results in the dispersal of the cumulus cells and causes the oocyte-cumulus complex to expand tremendously. The specific stimulus for mucification is thought to be FSH, and the functional significance of mucification is thought to be critical for the pickup and transport of the oocyte-cumulus complex in the fallopian tube.

FIGURE 94-19. Process of meiotic maturation or resumption of meiosis. A, Germinal vesicle stage. B, Germinal vesicle breakdown followed by condensation of chromosomes into bivalents. C and D, Release of first polar body and arrest of meiotic process at metaphase II. (Courtesy of Dr. C. Banka.)

Therefore, a cascade of FSH- and LH-dependent progesterone and PG responses are involved in mediating the ovulation of a fertilizable oocyte at midcycle. DIFFERENTIATION OF THE OOCYTE Oocyte differentiation involves two interrelated processes: growth and meiotic maturation.71 Oocyte growth is associated with the accumulation and storage of nutritional and informational molecules. During growth, the oocyte increases in diameter from 20 µm to 120 µm (see Fig. 94-2 and Fig. 94-3). Oocyte growth depends on the transcription of selected genes in chromosomes that are in the so-called lampbrush stage.2,4 Initially, oocyte and follicle growth are positively and linearly correlated until the follicle reaches the early tertiary stage; then, oocyte growth ceases while follicle growth continues.23 The oocyte therefore completes its growth very early in follicle development, for example, when the early tertiary follicle reaches ~400 µm in diameter (see Fig. 94-2 and Fig. 94-6). Granulosa cells are an absolute requirement for oocyte growth.72,73 As growth progresses, the oocyte is surrounded closely by the corona radiata granulosa cells, which are metabolically coupled with the oocyte by means of gap junctions composed of Cx37. There is evidence that 85% of the metabolites in follicle-enclosed oocytes originally are taken up by the granulosa cells and then transferred into the oocyte through gap junctions.72,73 As discussed earlier, the oocyte-derived ZP (see Fig. 94-6) plays an important role in a number of vital biologic functions. It contains species-specific receptors for capacitated sperm, it participates in the block of polyspermy, and it is critical in allowing the embryo to move freely through the fallopian tube into the uterus. There is increasing evidence that oocytes express growth factor ligands that control granulosa cytodifferentiation.19 Indeed, a functional link between one such oocyte growth factor, GDF-9, and folliculogenesis and fertility has been established.62 Hence, the emerging concept is that the oocyte may be at or near the apex in the mechanisms that control folliculogenesis in rats and, perhaps, humans.74 MEIOTIC MATURATION The capacity of the oocyte to resume meiosis is acquired at a specific stage in its growth, and the ability to complete meiotic maturation is acquired subsequently.23 Meiotic maturation or resumption of meiosis (see Fig. 94-19) is a process characterized by the dissolution of the nuclear or germinal vesicle membrane, the condensation of dictyotene chromosomes into discrete bivalents, the separation of homologous chromosomes, the release of a first polar body, and the arrest of the meiotic process at metaphase II. After meiotic maturation, the completion of meiosis and release of the second polar body are triggered by fertilization. In laboratory animals, the oocyte first acquires the capacity to resume meiosis at about the time of cavitation or early antrum formation, when the oocyte has completed its growth and is surrounded by the ZP and four or five layers of granulosa cells23 (see Fig. 94-6). The acquisition of the capacity for meiotic maturation seems to be a two-step process. First, the oocyte acquires the capacity to undergo germinal vesicle breakdown and to progress to metaphase I. Subsequently, it acquires the capacity to complete the first reductional division and release the first polar body.23 The mechanisms responsible for the acquisition of meiotic potential are unknown. Although the oocyte is capable of resuming meiosis early in follicle development, it is kept from doing so by an inhibitory influence. Under physiologic conditions, meiotic maturation is a highly selective process; it occurs only in those oocytes that are in dominant preovulatory follicles, responding to the preovulatory surge of LH.23 Fully grown oocytes from any tertiary or graafian follicle undergo meiotic maturation spontaneously if the oocyte is placed in tissue culture. Apparently, there is an inhibiting substance in follicular fluid that blocks meiotic maturation and that may be overridden by high levels of LH. The nature of the putative oocyte meiotic inhibitor remains to be elucidated. DIFFERENTIATION OF THE CORPUS LUTEUM The life of the corpus luteum typically is divided into two periods: luteinization and luteolysis. Luteinization begins in response to the preovulatory surge of gonadotropins. Maximal differentiation is reached at the end of 1 week (i.e., day 21 or 22 of the cycle).37 Subsequently, the corpus luteum normally undergoes apoptosis,35 a process called luteolysis. The basis of luteal differentiation is reflected in the biphasic secretion of progesterone, 17a-hydroxyprogesterone, androstenedione, and 17b-estradiol. This biphasic steroid production is causally connected to biphasic changes of activities of key steroidogenic enzymes. LH is the inducer of luteinization, and low levels of LH are critical for maintaining active luteal tissue during the early and midluteal phase. The hCG produced by the blastocyst can prevent luteolysis and promote further differentiation of the corpus luteum. The mechanism is totally obscure. Because LH/hCG action is mediated by receptors, luteinization becomes intimately connected with how LH receptors are controlled by luteal cells.36 LH receptors in the corpus luteum undergo a predictable pattern of activity. As ovulation approaches, a striking decrease in LH receptors occurs, and during the early luteal phase, LH receptors increase sharply, reaching near-maximum levels in the midluteal phase, and remain elevated until the end of the cycle.75 It seems that LH receptors first are down-regulated and then are reinduced during luteinization. Because reinduction of LH receptors is what renders the corpus luteum sensitive to hCG secreted by the implanting blastocyst, an important question is how LH receptors are reinduced in luteal cells. In the rat, FSH and prolactin are important in the induction and reinduction of LH receptors, respectively, in the granulosa cells.76 Functional FSH and prolactin receptors are present in the highest concentration in the early human corpus luteum, when LH receptors are being replenished.77 To what extent are FSH and prolactin involved in reinducing LH receptors during luteinization? Although the answer is unknown, it is important clinically, because foremost among luteal-phase defects in women is the inability of the corpus luteum to respond to LH/hCG. Could this defect be caused by inappropriate formation of new LH/hCG receptors? At luteolysis, a program is initiated that leads to apoptosis ~5 to 7 days later.35 After luteolysis is initiated, there occurs a striking decrease in progesterone production. Despite its physiologic importance, the mechanism regulating luteolysis in women is unknown. In laboratory animals, PGF2a is a physiologic luteolysin37; however, the precise role of PGF2a in human luteolysis remains equivocal.36,37

OVARIAN STEROIDOGENESIS LIPOPROTEINS AS CHOLESTEROL SOURCE Lipoproteins are complex particles containing a lipid core surrounded by amphiphilic proteins and phospholipids (see Chap. 162). Low-density lipoprotein (LDL; density = 1.019–1.063 g/mL) is the predominant cholesterol and cholesteryl ester carrier in human plasma and provides cholesterol to cells for membrane synthesis and as substrate for steroidogenesis in steroidogenic organs such as the ovary.78 High-density lipoprotein (HDL; density = 1.063–1.21 g/mL) also carries cholesterol and cholesteryl ester. In humans, HDL cannot provide cholesterol for steroidogenesis.79,80 In rats, a model often used in the study of the ovary, HDL and LDL can provide cholesterol for steroid hormone production.81 All cells can obtain cholesterol from two sources. Cholesterol can be synthesized de novo from acetate by the cell, or the cell can obtain the cholesterol from an external source such as lipoprotein.79 The mechanism by which steroidogenic cells in the human ovary obtain LDL-cholesterol essentially is the pathway described by Brown and Goldstein82 in their studies of the human fibroblast. LDL binds to ovarian steroidogenic cell-membrane receptors with high affinity and specificity. These receptors recognize apolipoprotein B, the predominant LDL apoprotein.78 In studies of human ovarian corpora lutea membranes, the number of LDL receptors is highest in the midluteal phase, suggesting that LDL-receptor number and the rate of steroidogenesis are positively correlated.83 After the LDL particle binds to the receptor on the plasma membrane, particle and receptor are internalized by the cell inside coated vesicles, which fuse to form lysosomes. The lysosomes hydrolyze the components of the LDL particle; cholesteryl ester is hydrolyzed to free cholesterol, and apoproteins are hydrolyzed to amino acids. The free cholesterol can be stored in the cell as cholesteryl ester or can be converted to steroid hormone products. The increase in the cell concentration of free cholesterol or a metabolic product such as an oxygenated sterol acts as a major control point in cell cholesterol metabolism by decreasing the number of LDL receptors, decreasing de novo cholesterol synthesis from acetate and increasing esterification of cholesterol to cholesteryl ester.84,85 In this way, the cell prevents the overaccumulation of cholesterol (Fig. 94-20).

FIGURE 94-20. Lipoprotein-cell interaction is shown in this model for low-density lipoprotein (LDL) uptake by an ovary steroidogenic cell by means of the LDL receptor. (C, cholesterol; CE, cholesteryl ester; PL, phospholipid; Apo, apolipoprotein B.) (From Schreiber JR, Weinstein D. Receptors in stereogenic cells. In: Scanu A, Spector A, eds. Lipoproteins, receptors, and cell function. New York: Marcel Dekker Inc, 1986.)

The factor that down-regulates de novo synthesis of cholesterol and LDL receptors appears to be an oxysterol rather than cholesterol itself. A prime candidate for this regulatory role is 26-hydroxycholesterol. Luteinized human granulosa cells contain 26-hydroxylase messenger RNA (mRNA), and this enzyme is localized to mitochondria. Data suggest that, when steroidogenesis is active, the 26-hydroxylase enzyme is inhibited by the products of the side-chain cleavage enzyme, allowing de novo cholesterol synthesis and increased LDL-cholesterol uptake.86 When steroidogenic activity is decreased in the ovary (i.e., follicle or corpus luteum), 26-hydroxylase activity remains, allowing the formation of 26-hydroxycholesterol and the resultant reduction in cholesterol synthesis and LDL-receptor gene expression. This finely tuned series of regulatory steps ensures enough cholesterol for cell function (e.g., membrane synthesis and specialized activities, such as steroid hormone production), but it prevents the overaccumulation of cholesterol in the cell. Although there are receptors for HDL in the human ovary, HDL is unable to provide cholesterol for steroid hormone production. At high concentrations, HDL seems to inhibit steroidogenesis by cultured human ovarian cells.80 This differential effect of HDL and LDL on steroid hormone synthesis by human ovarian cells could be of physiologic importance because of compartmentalization within the ovary. Granulosa cells in the preovulatory graafian follicle are bathed in follicular fluid but are separated from the ovarian vasculature by the basal lamina of the follicle. Human follicular fluid contains little or no LDL, but contains levels of HDL close to those in the plasma. The lack of available LDL probably limits progesterone synthesis before ovulation. After ovulation, the follicle becomes the vas-cularized corpus luteum, exposing these granulosa cells to plasma levels of LDL. The rapid rise in ovarian progesterone production after ovulation can be explained, at least in part, by the sudden availability of LDL-cholesterol as substrate by means of the LDL receptor.87 Before ovulation, the theca is well vascularized, and LDL would be available to provide cholesterol for thecal androgen production. Androgen can cross through the basal lamina to provide substrate for estrogen production by follicle granulosa cells. The theca of the dominant follicle has the richest blood supply of all follicles, ensuring adequate substrate for estrogen synthesis.88 Rat ovarian steroidogenic cells also can use LDL-cholesterol by the pathway described for human cells. However, HDL also can provide cholesterol for rat ovarian steroidogenesis by a mechanism that is quite different from that described for LDL. For example, the HDL particle can provide cholesterol to the cell in the absence of degradation of the apolipoprotein surface coat.81 The HDL (scavenger, type 1)-receptor mRNA is localized to theca cells of the rat ovary. Gonadotropic stimulation with hCG causes a marked increase in HDL receptor mRNA in theca interstitial cells and luteinized granulosa cells, consistent with a functional role for this receptor in rodent cholesterol transport and ovarian steroidogenesis.89 MECHANISMS OF GONADOTROPIN EFFECTS The a and the b subunits of FSH and LH are required for binding to the specific membrane receptors. Theca cells have specific LH receptors, but granulosa cells contain FSH receptors. FSH stimulates production of granulosa cell LH receptors. Gonadotropin binding to its receptor stimulates the cAMP–A kinase regulatory system.90 By this mechanism, gonadotropins increase available free cholesterol by increasing lipoprotein receptor number and by increasing hydrolysis of stored cholesteryl ester. In the absence of available lipoprotein (e.g., under serum-free in vitro conditions, perhaps within the follicle in the human ovary before ovulation), gonadotropins stimulate cholesterol synthesis de novo from acetate in the cell. The free cholesterol is transferred to mitochondria, which is the location of the rate-limiting enzyme in steroidogenesis, P450 side-chain cleavage enzyme (P450scc), which converts cholesterol to pregnenolone.81 There is evidence that cholesterol is carried to the mitochondria by a carrier protein called sterol carrier protein-2 (SCP-2). Gonadotropins then facilitate the transport of cholesterol from the outer to the inner mitochondrial membrane and continued transport to the P450scc enzyme on the inner mitochondrial membrane.91 A protein named steroidogenic acute regulatory protein (StAR) has been implicated as the regulator of cholesterol translocation from the outer to inner mitochondrial membrane.92 StAR mRNA transcripts are localized in the human ovary to the theca of preovulatory follicles, and luteinized granulosa and theca cells in the corpus luteum.93 Mutations of the StAR gene result in congenital lipoid adrenal hyperplasia, in which synthesis of all gonadal and adrenal steroids is severely impaired.94 This finding establishes the critical role of StAR protein in steroidogenesis. StAR gene expression is stimulated by cAMP.93 Gonadotropins can stimulate acutely the transport of cholesterol to the P450scc enzyme and chronically increase the amount and activity of this enzyme. After conversion of cholesterol to pregnenolone in the mitochondria, the pregnenolone moves back out of the mitochondria and into the cell cytoplasm for conversion to progesterone in corpus luteum cells and androgen in theca cells. The enzymes for these conversions are located on the microsomes. A summary of cholesterol transport and metabolism within ovarian cells is shown in Figure 94-21.

FIGURE 94-21. Cholesterol metabolism in a steroidogenic cell. (ACTH, adrenocorticotropic hormone; LH, luteinizing hormone; R, membrane receptor; cyclase, adenylate cyclase; ATP, adenosine triphosphate; AMP, adenosine monophosphate; P, phosphate; kinase, cyclic AMP–dependent protein kinase; hydrolase, cholesteryl ester hydrolase; HMG, human menopausal gonadotropin; ACAT, acyl CoA: cholesteryl acyltransferase.) Another source of cholesterol is low-density lipoprotein, as described in Figure 94-20. (From Schreiber JR, Weinstein D. Receptors in stereogenic cells. In: Scanu A, Spector A, eds. Lipoproteins, receptors, and cell function. New York: Marcel Dekker Inc, 1986.)

STEROID HORMONES PRODUCED BY THE OVARY ANDROGENS Androgens, primarily androstenedione and testosterone, are secreted by interstitial and thecal cells. The secretion rate of androstenedione is ~3 mg per day, with one-half coming from the ovaries and the other half coming from the adrenal glands or from the peripheral conversion of circulating dehydroepi-androsterone.95 The plasma androstenedione concentration is 40 to 240 ng/dL, with a small peak at ovulation and higher levels in the luteal than in the follicular phase. The secretion rate of testosterone is significantly less than that of androstenedione: ~0.25 mg per day, with a plasma concentration of ~19 to 70 ng/dL. Testosterone is bound tightly to testosterone-estradiol–binding globulin (TeBG; also known as sex hormone–binding globulin [SHBG]), and only ~1% of circulating testosterone is the biologically active free component (see Chap. 114). The principal function of ovarian androgen production is to provide substrate to the granulosa cell aromatase enzyme for the synthesis of estrogen.14 When ovarian androgen production is excessive (e.g., ovarian androgen-producing tumor, hyper-thecosis) or when the conversion of androgen to estrogen in the ovary is inefficient

(i.e., polycystic ovary syndrome), the excess androgen causes hirsutism or, at a higher concentration, virilism (see Chap. 96, Chap. 101 and Chap. 102). ESTROGEN Estrogen is produced predominantly in the ovarian follicle by the granulosa cell aromatization of thecal androgens.90 The amount of estrogen production depends on the time of the menstrual cycle. During the early follicular phase, the secretion rates of estradiol and estrone are about equal at 60 to 170 µg per day. As the dominant follicle is selected in the second half of the follicular phase, the secretion rate of estradiol rises to 400 to 800 µg per day, with the estradiol coming almost exclusively from the dominant follicle.96 The corpus luteum in the human ovary also produces significant quantities of estradiol, ~250 µg per day. In the late follicular and luteal phases, estrone secretion is one-fourth that of estradiol. Plasma concentrations of estradiol range from 50 to 60 pg/mL in the early follicular phase and rise to 250 to 400 pg/mL in the late follicular phase, but levels of estrone increase to only 150 to 200 pg/mL. The dominant follicle and corpus luteum produce ~95% of circulating estradiol, and estrone has little clinical significance in cycling women. However, in the menopausal years, estrone becomes the predominant estrogen. Estrone comes from peripheral conversion (mostly in adipose tissue) of adrenal androgens, particularly of androstenedione97 (see Chap. 100). Estradiol and estrone are converted into at least 40 metabolic products. Most are conjugated to glucuronic acid for excretion in the urine, but sulfates and mixed conjugates also are found. The principal metabolites of estradiol and estrone are estriol (by 16-hydroxylation, plus reduction of the 17-keto group for estrone) and catechol estrogens (by hydroxylation at the 2 or 4 position; Fig. 94-22).

FIGURE 94-22. Structure of estrogens, catechol estrogens, and cate-cholamines. Notice the similarity of the left portion of these catechol estrogens to catecholamines.

PROGESTERONE Progesterone production remains low until ovulation and corpus luteum formation. Just before ovulation, there is a slight rise in progesterone concentration; after ovulation, with the luteinization of granulosa cells and the influx of blood vessels carrying plasma levels of LDL, the production rate of progesterone rises to 10 to 40 mg each day.95 Plasma concentrations rise to 5 to 25 ng/mL in the midluteal phase. The normal life span of the corpus luteum, in the absence of pregnancy, is ~14 days. The cause of the demise of the corpus luteum has been postulated to be the effects of intraovarian estrogen, prostaglandins, and/or cytokines, but the mechanism remains unknown.98 However, LH stimulation is required for corpus luteum progesterone production during the luteal phase.99 Progesterone is cleared rapidly from plasma and is excreted mainly in the urine as pregnanediol. EFFECTS OF OVARIAN STEROIDS ESTROGEN Estrogen binds to specific estrogen receptors in target tissues and stimulates the production of particular mRNAs that direct the production of proteins that mediate the effects of estrogen.100 Estrogen has various effects throughout the body (Table 94-1). In the rat ovary, estrogen stimulates granulosa cell division and subsequent follicle growth, increases the number of FSH receptors per follicle, and increases the follicle uptake of FSH. Estrogen has a positive effect on its own production within the ovary.101 Estrogen sensitizes the human pituitary to the effects of gonadotropin-releasing hormone (GnRH or LHRH), increasing the secretion of LH in response to a given amount of GnRH. Estrogen stimulates the secondary sexual characteristics associated with female puberty, including breast development. It also stimulates uterine endometrial gland proliferation and induces the appearance of endometrial cell progesterone receptors. The hormone stimulates vaginal epithelial growth and secretions and cervical mucus production. It also has multiple effects on the liver, including the stimulation of hepatic synthesis of proteins such as renin substrate, TeBG, and others. In the postmenopausal woman, estrogen replacement can prevent hot flushes and bone loss (i.e., osteoporosis)102 (see Chap. 100).

TABLE 94-1. Functions of Estrogens and Progestins in Nonpregnant Women

PROGESTERONE Progesterone also binds to specific receptors in target tissues. The hormone converts proliferative into secretory endometrium in estrogen-primed uteri (see Table 94-1). High levels of progesterone act as a smooth muscle relaxant and, in pregnancy, allow the uterus to expand. Progesterone counteracts the effect of estrogen on cervical mucus, making the estrogen-induced thin mucus thicker; it also has a mild thermogenic effect and is responsible for the increase in basal body temperature associated with ovulation.102 The progesterone secreted by the ovary just before ovulation may synergize with estradiol to ensure the LH surge responsible for ovulation.101 ANTIESTROGENS Compounds that compete with estrogens for binding to the estrogen receptor and have little or no estrogenic effect themselves are antiestrogens. There are two groups: steroidal and nonsteroidal. Estrogen metabolites such as estriol bind weakly to the estrogen receptor and have little estrogenic activity themselves, but they can limit access of active estrogens to the receptor and behave as antiestrogens. Nonsteroidal antiestrogens have important therapeutic value. Clomiphene citrate, which structurally is related to diethylstilbestrol (DES), has found wide use as a fertility agent. Other nonsteroidal antiestrogens, such as tamoxifen and raloxifene, are members of a new class of drugs known as selective estrogen receptor modulators (SERMs). These drugs work as estrogen in some tissues, but not in others, and offer new ways to treat osteoporosis and breast cancer.103,104

CATECHOL ESTROGENS Other metabolic products of estrogen include the catechol estrogens, which result from hydroxylation of estradiol or estrone at the 2 or 4 position (see Fig. 94-22). They have the potential for interacting with receptors for estrogen and catecholamines and with enzymes that degrade these two latter groups of compounds. Catechol estrogens are found in at least 10-fold higher concentrations in the hypothalamus and pituitary than are estradiol or estrone.105 Although the functions of catechol estrogens remain unknown, it has been postulated that they may influence GnRH or LH release by acting as antiestrogens through the blockade of the binding of active estrogens to the estrogen receptor and by increasing the activity of endogenous catecholamines by competing for available catecholamine-degrading enzyme, catechol-O-methyltrans-ferase.106 Synergism of catechol estrogens and catecholamine and gonadotropin effect on progesterone production have also been reported.107 INTRAOVARIAN STEROID EFFECTS ESTROGEN RECEPTOR A novel estrogen receptor (ER), named ER beta, has been cloned from human tissue.108 A comparison of the amino-acid sequence of ER beta with the classic ER (ER alpha) shows a high degree of conservation in the DNA-binding and ligand-binding domains. The rest of the protein is not conserved.108 ER beta is expressed in human thymus, spleen, ovary, and testis. Estradiol-17 binds and activates both ER alpha and ER beta. ER alpha has been identified in the ovarian surface epithelium cells. ER beta is most abundant in human fetal ovaries, suggesting that it perhaps plays a role in ovarian development.109 Analysis of whole human ovaries shows equal amounts of ER alpha and ER beta mRNA.108 Granulosa cells obtained at oocyte aspiration contain high levels of ER beta mRNA109 and functional ERs.110 These data are consistent with a direct role of estrogen on follicular granulosa cells. This is, however, an evolving and complex field. PROGESTERONE RECEPTOR The progesterone receptor (PR) exists in two isoforms in the human, PRA and PRB. PRA is an amino-terminal truncated variant of PRB. PR has been identified in the epithelial and stromal cells in the ovaries of multiple species, including human.111 The periovulatory expression of the PR gene in granulosa cells has been detected in all species examined thus far.11 However, there are distinct species differences in the expression of PR in the corpus luteum (CL). PR has not been detected in the rat CL,112 but is present in the CL of primates. PR mRNA is lowest in the primate CL cells in the early luteal phase, and increases threefold by the mid to late luteal phase.113 hCG stimulates luteinized granulosa cell PR mRNA and PR expression, which is modulated by progesterone or a progesterone metabolite.114 Interestingly, ovarian PR is not under estrogen control, as it is in most tissues.111 These data suggest a role for intraovarian progesterone in primate ovarian function. The significance of the PR in the rodent ovary has been examined in the PR knock-out mouse. In the absence of PR, the ovary contains normal-appearing granulosa cells, but no corpora lutea, and no ovulation occurs in response to exogenous gonadotropin. These data suggest that PR plays a critical role in rodent ovulation and corpus luteum formation.66 ANDROGEN RECEPTOR Androgen receptor (AR) localization and abundance have been studied in primate ovary. In the ovary of the rhesus monkey, in situ hybridization demonstrates AR mRNA to be most abundant in granulosa cells of healthy preantral and antral follicles.115 Theca interna and stromal cells also express AR mRNA, but to a lesser extent than granulosa cells. Granulosa cell AR mRNA abundance is positively correlated with granulosa cell proliferation and is negatively correlated with granulosa cell apoptosis, suggesting that intraovarian androgen stimulates early follicular development in the primate ovary.115 Similar results were found in studies of the marmoset ovary, using immunohistochemistry. Granulosa cells of immature follicles had a 4.2-fold higher level of immunoreactive AR, as compared to preovulatory follicles. These and other data suggest that the paracrine action of androgen on granulosa cells converts from stimulation to inhibition as the follicle matures. It has been hypothesized that androgens stimulate early follicular development, and a development-related reduction in AR could then protect the follicle against a late inhibitory action of androgen, thus promoting preovulatory follicle dominance in the primate ovarian cycle.116 RELAXIN Relaxin is a 6-kilodalton (kDa) peptide hormone. The amino-acid sequence of relaxin from several species, including pig, rat, and shark, has been determined, and although the structure is conserved poorly across species lines, all the relaxins consist of two dissimilar peptide chains linked by disulfide bridges (Fig. 94-23). There is significant structural homology to insulin; as is the case with insulin, relaxin is derived from a larger precursor in which the two chains are connected by a C peptide.117

FIGURE 94-23. Structure of relaxin. The amino- and carboxyl-terminal sequences of the A and B chains are based on probable sites of proteolytic cleavage. These sites have been confirmed for porcine relaxin (Schwabe C, McDonald JK, et al. Primary structure of the B-chain of porcine relaxin. Biochem Biophys Res Commun 1977; 75:503). Human relaxin-2 is a form of the hormone found in human ovaries during pregnancy. Cystine bridges for the A and B chains of relaxin are in the same relative amino-acid positions as in human insulin. (Sequence information obtained from Hudson P, John M, Crawford R, et al. Relaxin gene expression in human ovaries and the predicted structure of a human preprorelaxin by analysis of cDNA clones. EMBO J 1984; 3:2333.) Single-letter abbreviations for amino acids: A, alanine; B, asparagine or aspartic acid; C, cysteine; D, aspartic acid; E, glutamic acid; F, phenylalanine; G, glycine, H, histidine; I, isoleucine; K, lysine; L, leucine; M, methionine; N, asparagine, P, proline; Q, glutamine; R, arginine; S, serine; T, threonine; V, valine; W, tryptophan; Y, tyrosine; Z, glutamine or glutamic acid.

Relaxin serves various functions and has different tissue sources in the species studied; therefore, extrapolation across species lines is unwise. Relaxin has been found in the peripheral blood of pregnant animals in all species studied (including women), and the principal tissue source is the corpus luteum, except in the pregnant mare, in which relaxin comes from the placenta. Relaxin has an obligatory role in cervical ripening, softening, and dilatation in pigs and rats, but no such role has been identified in women. Relaxin occasionally can be detected in the serum of nonpregnant women during the luteal phase and is found in the corpus luteum of nonpregnant women, but the concentration is only one-hundredth that in the corpus luteum of pregnant women. In the rat, relaxin biologic activity is detected on day 14 of pregnancy and is maximal on day 20 (i.e., the last day of pregnancy in the rat). Relaxin is undetectable on day 1 of lactation. By immunocytochemical techniques, relaxin is localized exclusively in the rat ovary to the corpora lutea cells.118 In the human, relaxin mRNA is localized to the corpus luteum of the cycling ovary, and is found in even much higher levels in the corpus luteum of pregnancy.119 Relaxin is detectable in serum by postconception day 14. Relaxin concentrations in sera and corpora lutea are maximal in the first trimester and then fall by 20% at the end of the first trimester and remain stable through the end of pregnancy. Three days after delivery, relaxin falls to undetectable levels. Studies in the primate demonstrate that chorionic gonadotropin stimulates corpus luteum relaxin secretion.120 The effects of relaxin on the reproductive tract include actions that lead to changes that allow pregnancy to progress and that facilitate delivery. There are important species differences, and many questions remain. Relaxin facilitates delivery in rodents by causing a breakdown in the interpubic ligament that binds the pubic symphysis together. Inhibition of uterine activity has been observed in guinea pigs, sheep, and hamsters. In the human, progesterone and relaxin synergistically decrease the amplitude of

spontaneous myometrial contractions in vitro. Relaxin causes cervical softening in several species, including the rat and pig, but it does not seem to have this effect in primates.117 Relaxin is also present in the male reproductive tract, and in seminal fluid, it may play a role in sperm motility.

CONCLUSION Knowledge of ovarian physiology leans heavily on findings in rodent species. The methods have been adapted by researchers working on the human ovary. Unfortunately, although many molecular processes found in the rat are also found in the human, many are not. The difficulty in this situation is the tendency to generalize the findings to all animals. The conclusion is that it is dangerous to attempt to generalize between the rodent and the human. Much more needs to be learned before the menstrual cycle can be fully understood. An understanding of folliculogenesis can lead to an understanding of fertility and the ability to alleviate the afflictions of infertility. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Reynolds SRM. The vasculature of the ovary and ovarian function. Recent Prog Horm Res 1950; 5:65. Baker TG, Sum W. Development of the ovary and oogenesis. Clin Obstet Gynaecol 1976; 3:3. Erickson GF. Basic biology: ovarian anatomy and physiology. In: Lobo R, Marcus R, Kelsey J, eds. Menopause. San Diego: Academic Press (in press 1999). Gougeon A. Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr Rev 1996; 17:121. Bachvarova R. Gene expression during oogenesis and oocyte development in mammals. In: Browder L, ed. Developmental biology: a comprehensive synthesis. New York: Plenum Publishing, 1985. Moos J, Faundes D, Kopf GS, Schultz RM. Composition of the human zona pellucida and modifications following fertilization. Hum Reprod 1995; 10:2467. Wassarman PM. Zona pellucida glycoproteins. Annu Rev Biochem 1988; 57:415. Skinner SM, Prasad SV, Ndolo TM, Dunbar BS. Zona pellucida antigens: targets for contraceptive vaccines. Am J Reprod Immunol 1996; 35:163. Bagavant H, Fusi FM, Baisch J, et al. Immunogenicity and contraceptive potential of a human zona pellucida 3 peptide vaccine. Biol Reprod 1997; 56:764. Grazul-Bilska AT, Reynolds LP, Redmer DA. Gap junctions in the ovaries. Biol Reprod 1997; 57:947. Kumar NM, Gilula NB. The gap junction communication channel. Cell 1996; 84:381. Simon AM, Goodenough DA, Li F, Paul DL. Female infertility in mice lacking connexin 37. Nature 1997; 385:525. Green SH, Zuckerman S. Quantitative aspects of the growth of the human ovum and follicle. J Anat 1951; 85:373. Erickson GF, Magoffin DA, Dyer C, Hofeditz C. The ovarian androgen producing cells: a review of structure/function relationships. Endocr Rev 1985;6:371. Meredith JE Jr, Winitz S, Lewis JM, et al. The regulation of growth and intracellular signaling by integrins. Endocr Rev 1996; 17:207. Erickson GF. The graafian follicle: a functional definition. In: Adashi EY, ed. Ovulation: evolving scientific and clinical concepts. New York: Springer-Verlag (in press 1999). Erickson GF. Ovarian androgen biosynthesis: endocrine regulation. In: Azziz R, Nestler JE, Dewailly D, eds. Androgen excess disorders in women. New York: Lippincott–Raven Publishers, 1997. McNatty KP, Moore-Smith D, Osathanondh R, Ryan KJ. The human antral follicle: functional correlates of growth and atresia. Ann Biol Anim Biochim Biophys 1979; 19:1547. Edwards RG. Follicular fluid. J Reprod Fertil 1974; 37:189. Eppig JJ, Chesnel F, Hirao Y, et al. Oocyte control of granulosa cell development: how and why. Hum Reprod 1997; 12:127. Erickson GF. The ovary: basic principles and concepts. In: Felig P, Baxter JD, Frohman LA, eds. Endocrinology and metabolism, 3rd ed. New York: McGraw-Hill, 1995. Hsueh AJ, Billig H, Tsafriri A. Ovarian follicle atresia: a hormonally controlled apoptotic process. Endocr Rev 1994; 15:707. Tilly JL. Apoptosis and ovarian function. Rev Reprod 1996; 1:162. Erickson GF. Defining apoptosis: players and systems. J Soc Gynecol Invest 1997; 4:219. Erickson GF. Analysis of follicle development and ovum maturation. Semin Reprod Endocrinol 1986; 4:233. Hirshfield AN, Midgley AR. Morphometric analysis of follicular development in the rat. Biol Reprod 1978; 19:597. Erickson GF, Yen SSC. New data on follicle cells in polycystic ovaries: a proposed mechanism for the genesis of cystic follicles. Semin Reprod Endocrinol 1984; 2:231. Magoffin DA, Erickson GF. Control systems of theca-interstitial cells. In: Findlay JK, ed. Molecular biology of the female reproductive system. New York: Academic Press, 1994. Erickson GF, Magoffin DA, Jones KL. Theca function in polycystic ovaries of a patient with virilizing congenital adrenal hyperplasia. Fertil Steril 1989; 51:173. Kawakami M, Kubo K, Uemura T, et al. Involvement of ovarian innervation in steroid secretion. Endocrinology 1981; 109:136. Dissen GA, Dees WL, Ojeda SR. Neural and neurotrophic control of ovarian development. In: Adashi EY, Leung PCK, eds. The ovary. New York: Raven Press, 1993. Brännström M, Norman RJ. Involvement of leukocytes and cytokines in the ovulatory process and corpus luteum function. Hum Reprod 1993; 8:1762. Crisp TM, Dessouky DA, Denys FR. The fine structure of the human corpus luteum of early pregnancy and during the progestational phase of the menstrual cycle. Am J Anat 1970; 127:37. Suzuki T, Sasano H, Tamura M, et al. Temporal and spatial localization of steroidogenic enzymes in premenopausal human ovaries: in situ hybridization and immunohistochemical study. Mol Cell Endocrinol 1993; 97:135. Shikone T, Yamoto M, Kokawa K, et al. Apoptosis of human corpora lutea during cyclic luteal regression and early pregnancy. J Clin Endocrinol Metab 1996; 81:2376. Michael AE, Abayasekara DR, Webley GE. Cellular mechanisms of luteolysis. Mol Cell Endocrinol 1994; 99:R1. Auletta FJ, Flint AP. Mechanisms controlling corpus luteum function in sheep, cows, nonhuman primates, and women especially in relation to the time of luteolysis. Endocr Rev 1988; 9:88. Gougeon A. Dynamics of follicular growth in the human: a model from preliminary results. Hum Reprod 1986; 1:81. Gougeon A, Lefèvre B. Evolution of the diameters of the largest healthy and atretic follicles during the human menstrual cycle. J Reprod Fertil 1983; 69:497. Gougeon A, Ecochard R, Thalabard JC. Age-related changes of the population of human ovarian follicles: increase in the disappearance rate of non-growing and early-growing follicles in aging women. Biol Reprod 1994; 50:653.

40a.McGee EA, Hsueb AJW. Initial and cyclic recruitment of ovarian regulatory system in health and disease. Endocr Rev 1999; 13:1018. 41. Gougeon A, Chainy GBN. Morphometric studies of small follicles in ovaries of women at different ages. J Reprod Fertil 1987; 81:433. 42. Zeleznik AJ. Dynamics of primate follicular growth: a physiologic perspective. In: Adashi EY, Leung PCK, eds. The ovary. New York: Raven Press, 1993. 43. Welt CK, Martin KA, Taylor AE, et al. Frequency modulation of follicle-stimulating hormone (FSH) during the luteal-follicular transition: evidence for FSH control of inhibin B in normal women. J Clin Endocrinol Metab 1997; 82:2645. 44. Zeleznik AJ. Premature elevation of systemic estradiol reduces serum levels of follicle-stimulating hormone and lengthens the follicular phase of the menstrual cycle in rhesus monkeys. Endocrinology 1981; 109:352. 45. McNatty KP, Hunter WM, McNeilly AS, Sawers RS. Changes in the concentration of pituitary and steroid hormones in the follicular fluid of human graafian follicles throughout the menstrual cycle. J Endocrinol 1975; 64:555. 46. Phillips DJ, deKretser DM. Follistatin: a multifunctional regulatory protein. Front Neuroendocrinol 1998; 19:287. 47. Sawetawan C, Milewich L, Word RA, et al. Compartmentalization of type I 17 beta-hydroxysteroid oxidoreductase in the human ovary. Mol Cell Endocrinol 1994; 99:161. 48. Zhang Y, Word RA, Fesmire S, et al. Human ovarian expression of 17 beta-hydroxysteroid dehydrogenase types 1, 2, and 3. J Clin Endocrinol Metab 1996; 8 1:3594. 49. Segaloff DL, Ascoli M. The lutropin/choriogonadotropin receptor...4 years later. Endocr Rev 1993; 14:324. 50. Simoni M, Gromoll J, Nieschlag E. The follicle-stimulating hormone receptor: biochemistry, molecular biology, physiology, and pathophysiology. Endocr Rev 1997; 18:739. 51. Meyer TE, Habener JF. Cyclic adenosine 3',5'-monophosphate response element binding protein (CREB) and related transcription-activating deoxyribonucleic acid–binding proteins. Endocr Rev 1993; 14:269. 52. McNatty KP, Sawers RS. Relationship between the endocrine environment within the graafian follicle and the subsequent rate of progesterone secretion by human granulosa cells in vitro. J Endocrinol 1975; 66:391. 53. Yong EL, Baird DT, Hillier SG. Mediation of gonadotrophin-stimulated growth and differentiation of human granulosa cells by adenosine-3',5'-monophos-phate: one molecule, two messages. Clin Endocrinol (Oxf) 1992; 37:51. 54. Gospodarowicz D, Bialecki H. Fibroblast and epidermal growth factors are mitogenic agents for cultured granulosa cells of rodent, porcine, and human origin. Endocrinology 1979; 104:757. 55. Shaw HJ, Hillier SG, Hodges JK. Developmental changes in luteinizing hormone/human chorionic gonadotropin steroidogenic responsiveness in marmoset granulosa cells: effects of follicle-stimulating hormone and androgens. Endocrinology 1989; 124:1669. 56. Bar-Ami S, Haciski RC, Channing CP. Increasing 125I-human chorionic gonadotrophin specific binding in human granulosa cells by follicle-stimulating hormone and follicular fluid. Hum Reprod 1989; 4:876. 57. Channing CP. Steroidogenesis and morphology of human ovarian cell types in tissue culture. J Endocrinol 1969; 45:297. 58. Yamoto M, Shima K, Nakano R. Gonadotropin receptors in human ovarian follicles and corpora lutea throughout the menstrual cycle. Horm Res 1992; 37(Suppl 1):5. 59. Adashi F, Leung PCK, eds. The ovary: comprehensive endocrinology. New York: Raven Press, 1993. 60. Erickson GF, Danforth DR. Ovarian control of follicle development. Am J Obstet Gynecol 1995; 172:736. 60a.Abulafia O, Sherer DM. Angiogenesis of the ovary. Am J Obstet Gynecol 2000; 182:240. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80.

Poretsky L, Catakio NA, Roseriwaks Z, et al. The insulin-related ovarian regulatory system in health and disease. Endocr Rev 1999; 20:535. Elvin JA, Yan C, Wang P, et al. Molecular characterization of the follicle defects in the growth differentiation factor 9-deficient ovary. Mol Endo-crinol 1999; 13:1018. Tsafriri A, Chun SY, Reich R. Follicular rupture and ovulation. In: Adashi EY, Leung PCK, eds. The ovary: comprehensive endocrinology. New York: Raven Press, 1993. Okamura H, Takenaka A, Yajima Y, Nishimura T. Ovulatory changes in the wall at the apex of the human graafian follicle. J Reprod Fertil 1980; 58:153. Graham JD, Clarke CL. Physiological action of progesterone in target tissues. Endocr Rev 1997; 18:502. Lydon JP, DeMayo FJ, Funk CR, et al. Mice lacking progesterone receptor exhibit pleiotropic reproductive abnormalities. Genes Dev 1995; 9:2266. Herschman HR. Prostaglandin synthase 2. Biochim Biophys Acta 1996; 1299:125. Dinchuk JE, Car BD, Focht RJ, et al. Renal abnormalities and an altered inflammatory response in mice lacking cyclooxygenase II. Nature 1995; 378:406. Lim H, Paria BC, Das SK, et al. Multiple female reproductive failures in cyclooxygenase 2-deficient mice. Cell 1997; 91:197. Beers WH. Follicular plasminogen and plasminogen activator and the effect of plasmin on ovarian follicle wall. Cell 1975; 6:379. Eppig JJ. Regulation of mammalian oocyte maturation. In: Adashi EY, Leung PCK, eds. The ovary: comprehensive endocrinology. New York: Raven Press, 1993. Helter DT, Cahill DM, Schultz RM. Biochemical studies of mammalian oogenesis: metabolic cooperativity between granulosa cells and growing mouse oocytes. Dev Biol 1981; 84:455. Brower PT, Schultz RM. Intercellular communication between granulosa cells and mouse oocytes: existence and possible nutritional role during oocyte growth. Dev Biol 1982; 90:144. McGrath SA, Esquela AF, Lee S-J. Oocyte-specific expression of growth/differentiation factor-9. Mol Endocrinol 1995; 9:131. Nishimori K, Dunkel L, Hsueh AJ, et al. Expression of luteinizing hormone and chorionic gonadotropin receptor messenger ribonucleic acid in human corpora lutea during menstrual cycle and pregnancy. J Clin Endocrinol Metab 1995; 80:1444. Richards JS. Hormonal control of gene expression in the ovary. Endocr Rev 1994; 15:725. McNeilly AS, Kerin J, Swanston IA, et al. Changes in the binding of human chorionic gonadotrophin/luteinizing hormone, follicle-stimulating hormone and prolactin to human corpora lutea during the menstrual cycle and pregnancy. J Endocrinol 1980; 87:315. Gwynne JT, Strauss JF 3rd. The role of lipoproteins in steroidogenesis and cholesterol metabolism in steroidogenic glands. Endocr Rev 1982; 3:299. Miller GJ. High density lipoproteins and atherosclerosis. Ann Rev Med 1980; 31:97. Tureck RW, Strauss JF 3rd. Progesterone synthesis by luteinized human granulosa cells in culture: the role of de novo sterol synthesis and lipoprotein-carried sterol. J Clin Endocrinol Metab 1982; 54:367.

81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120.

Schreiber JR, Weinstein DB. Lipoprotein receptors in steroidogenesis. In: Scanu AM, Spector A, eds. Lipoproteins, receptors, and cell function. New York: Marcel Dekker Inc, 1985. Brown MS, Goldstein JL. Receptor-mediated control of cholesterol metabolism. Science 1976; 191:150. Ohashi M, Carr BR, Simpson ER. Lipoprotein-binding sites in human corpus luteum membrane fractions. Endocrinology 1982; 110:1477. Kandutsch AA, Chen HW, Heiniger HJ. Biological activity of some oxygenated sterols. Science 1978; 201:498. Brown MS, Goldstein JL. Receptor-mediated endocytosis: insights from the lipoprotein receptor system. Proc Natl Acad Sci U S A 1979; 76:3330. Rennert H, Fischer RT, Alvarez JG, et al. Generation of regulatory oxy-sterols: 26-hydroxylation of cholesterol by ovarian mitochondria. Endocrinology 1990; 127:738. Carr BR, MacDonald PC, Simpson ER. The role of lipoproteins in the regulation of progesterone secretion by the human corpus luteum. Fertil Steril 1982; 38:303. diZerega GS, Hodgen GD. Fluorescence localization of luteinizing hormone/human chorionic gonadotropin uptake in the primate ovary. II. Changing distribution during selection of the dominant follicle. J Clin Endocrinol Metab 1980; 51:903. Li X, Peegel H, Menon KM. In situ hybridization of high density lipoprotein (scavenger, type 1) receptor messenger ribonucleic acid (mRNA) during folliculogenesis and luteinization: evidence for mRNA expression and induction by human chorionic gonadotropin specifically in cell types that use cholesterol for steroidogenesis. Endocrinology 1998; 139:3043. Richards JS, Hedin L. Molecular aspects of hormone action in ovarian follicular development, ovulation, and luteinization. Annu Rev Physiol 1988; 50:441. Privalle CT, Crivello JF, Jefcoate CR. Regulation of intramitochondrial cholesterol transfer to side-chain cleavage cytochrome P-450 in rat adrenal gland. Proc Natl Acad Sci U S A 1983; 80:702. Clark BJ, Wells J, King SR, Stocco DM. The purification, cloning, and expression of a novel luteinizing hormone-induced mitochondrial protein in MA-10 mouse Leydig tumor cells. J Biol Chem 1994; 269:28314. Kiriakidou M, McAllister JM, Sugawara T, Strauss JF 3rd. Expression of steroidogenic acute regulatory protein StAR in the human ovary. J Clin Endo-crinol Metab 1996; 81:4122. Lin D, Sugawara T, Strauss JF 3rd, et al. Role of steroidogenic acute regulatory protein in adrenal and gonadal steroidogenesis. Science 1995; 267:1828. Lipsett M. Steroid hormones. In: Yen SSC, Jaffe RB, eds. Reproductive endocrinology. Philadelphia: WB Saunders, 1978. Hodgen GD. The dominant ovarian follicle. Fertil Steril 1982; 38:281. Grodin JM, Siiteri PK, MacDonald PC. Source of estrogen production in postmenopausal women. J Clin Endocrinol Metab 1973; 36:207. Wuttke W, Theiling K, Hinney B, Pitzel L. Regulation of steroid production and its function within the corpus luteum. Steroids 1998; 63:299. Hutchison JS, Zeleznik AJ. The rhesus monkey corpus luteum is dependent on pituitary gonadotropin secretion throughout the luteal phase of the menstrual cycle. Endocrinology 1984; 115:1780. King WJ, Greene GL. Monoclonal antibodies localize oestrogen receptor in the nuclei of target cells. Nature 1984; 307:745. Schreiber J. Current concepts of human follicular growth and development. Contemp Obstet Gynecol 1983; 26:125. Henzl M. Natural and synthetic female sex hormones. In: Yen SSC, Jaffe RB, eds. Reproductive endocrinology. Philadelphia: WB Saunders, 1978. Baker VL, Jaffe RB. Clinical uses of antiestrogens. Obstet Gynecol Surv 1996; 51:45. Baynes KC, Compston JE. Selective oestrogen receptor modulators: a new paradigm for HRT. Curr Opin Obstet Gynecol 1998; 10:189. Paul SM, Axelrod J. Catechol estrogens: presence in brain and endocrine tissues. Science 1977; 197:657. Ball P, Knuppen R, Haupt M, Breuer H. Interactions between estrogens and catechol amines. 3. Studies on the methylation of catechol estrogens, catechol amines and other catechols by the catechol-O-methyltransferases of human liver. J Clin Endocrinol Metab 1972; 34:736. Spicer LJ, Hammond JM. Mechanism of action of 2-hydroxyestradiol on steroidogenesis in ovarian granulosa cells: interactions with catecholamines and gonadotropins involve cyclic adenosine monophosphate. Biol Reprod 1989; 40:87. Mosselman S, Polman J, Dijkema R. ER beta: identification and characterization of a novel human estrogen receptor. FEBS Lett 1996; 392:49. Brandenberger AW, Tee MK, Jaffe RB. Estrogen receptor alpha (ER-alpha) and beta (ER-beta) mRNAs in normal ovary, ovarian serous cystadenocarcinoma and ovarian cancer cell lines: down-regulation of ER-beta in neoplastic tissues. J Clin Endocrinol Metab 1998; 83:1025. Hurst BS, Zilberstein M, Chou JY, et al. Estrogen receptors are present in human granulosa cells. J Clin Endocrinol Metab 1995; 80:229. Pinter JH, Deep C, Park-Sarge OK. Progesterone receptors: expression and regulation in the mammalian ovary. Clin Obstet Gynecol 1996; 39:424. Park-Sarge OK, Parmer TG, Gu Y, Gibori G. Does the rat corpus luteum express the progesterone receptor gene? Endocrinology 1995; 136:1537. Duffy DM, Stouffer RL. Progesterone receptor messenger ribonucleic acid in the primate corpus luteum during the menstrual cycle: possible regulation by progesterone. Endocrinology 1995; 136:1869. Duffy DM, Molskness TA, Stouffer RL. Progesterone receptor messenger ribonucleic acid and protein in luteinized granulosa cells of rhesus monkeys are regulated in vitro by gonadotropins and steroids. Biol Reprod 1996; 54:888. Weil SJ, Vendola K, Zhou J, et al. Androgen receptor gene expression in the primate ovary: cellular localization, regulation, and functional correlations. J Clin Endocrinol Metab 1998; 83:2479. Hillier SG, Tetsuka M, Fraser HM. Location and developmental regulation of androgen receptor in primate ovary. Hum Reprod 1997; 12:107. Weiss G. Relaxin. Annu Rev Physiol 1984; 46:43. Golos TG, Weyhenmeyer JA, Sherwood OD. Immunocytochemical localization of relaxin in the ovaries of pregnant rats. Biol Reprod 1984; 30:257. Bogic LV, Mandel M, Bryant-Greenwood GD. Relaxin gene expression in human reproductive tissues by in situ hybridization. J Clin Endocrinol Metab 1995; 80:130. Duffy DM, Hutchison JS, Stewart DR, Stouffer RL. Stimulation of primate luteal function by recombinant human chorionic gonadotropin and modulation of steroid, but not relaxin, production by an inhibitor of 3 beta-hydroxysteroid dehydrogenase during simulated early pregnancy. J Clin Endocrinol Metab 1996; 81:2307.

CHAPTER 95 THE NORMAL MENSTRUAL CYCLE AND THE CONTROL OF OVULATION Principles and Practice of Endocrinology and Metabolism

CHAPTER 95 THE NORMAL MENSTRUAL CYCLE AND THE CONTROL OF OVULATION ROBERT W. REBAR, GARY D. HODGEN, AND MICHAEL ZINGER General Characteristics of the Normal Menstrual Cycle Hormonal Changes During the Normal Menstrual Cycle Cyclic Changes in the Target Organs of the Reproductive Tract Endometrium Cervix and Cervical Mucus Vaginal Mucosa Central and Gonadal Feedback Mechanisms in the Control of Ovulation Hypothalamic-Pituitary Signals Steroidal Feedback Presumed Mechanism for the Luteinizing Hormone Surge Effect of the Ovary on Gonadotropin Secretion Recruitment and Selection of the Dominant Follicle in Ovulatory Menstrual Cycles Role of Estradiol in Follicular Development Mechanisms of Ovulation Luteal Function as the Sequel to Folliculogenesis Chapter References

Perhaps the single feature that most clearly distinguishes the reproductive endocrinology of the female from that of the male is the dependence of female reproductive function on an entirely different set of endocrine rhythms. In considering abnormal female reproductive function, it is imperative to remember what is normal for that moment in the life of that individual. An overview of the patterns of circulating concentrations of luteinizing hormone (LH), follicle-stimulating hormone (FSH), and estradiol (E2) throughout the life of the normal woman is depicted in Figure 95-1. Rhythmic changes occur in the hormones secreted by all levels of the reproductive system. Moreover, hormonal secretion is modified through several phases of the life cycle.1 Gonadotropin secretion is low in the prepubertal years, increases before and during pubertal development, assumes the characteristic monthly cyclicity of the reproductive years, and finally increases to high levels after the menopause (i.e., the final menstrual period). These changes are both temporally and causally related to simultaneous rhythms in the secretion of ovarian (especially E2) and hypothalamic (particularly gonadotropin-releasing hormone [GnRH]) hormones. Superimposed on these long-term changes are the shorter-term rhythms that are so important to female reproduction.

FIGURE 95-1. Changing patterns of luteinizing hormone (LH), follicle-stimulating hormone (FSH), and estradiol (E2) concentrations in peripheral blood throughout the life of a typical woman. The pubertal period has been expanded to depict the sleep-induced increases in LH and FSH followed by morning increases in E2 that are observed during puberty. Such sleep-associated increases also occur during the early follicular phase of the menstrual cycle. (From Rebar RW. Normal physiology of the reproductive system. In: Endocrine metabolism continuing education and quality control program. American Association of Clinical Chemistry, 1982.)

Several distinctive rhythms become prominent as a female child progresses to sexual maturity. Female puberty is characterized by the resetting of the classic negative ovarian steroid feedback loop, the establishment of new circadian (24-hour) and ultradian (60- to 90-minute) gonadotropin rhythms, and the development of a positive estrogen feedback loop controlling the infradian (monthly) rhythm as an interdependent cyclic expression of the gonadotropins and the ovarian steroids.2 Sleep-related increases in gonadotropins and gonadal steroids become evident during puberty and appear to play an important role in pubertal maturation (see Chap. 91). Although changes in reproductive hormones during the menstrual cycle in adult women are common knowledge, it is not widely recognized that basal concentrations of several other hormones, including growth hormone (GH),3,4 prolactin,5,6 corticotropin and cortisol, and parathyroid hormone and calcitonin,7 also are influenced by the stage of the menstrual cycle. Although the physiologic significance of these changes is unclear, they suggest that the menstrual cycle affects systems and functions throughout the body. What is most remarkable about these separate and yet interdependent rhythms is that in most women, they are coordinated in an as yet incompletely defined manner to ensure ovulation and pregnancy. Because available data indicate the existence of finely controlled rhythms at all levels of the reproductive system, an abnormality at any level may lead to an abnormal state with altered rhythms.

GENERAL CHARACTERISTICS OF THE NORMAL MENSTRUAL CYCLE A series of cyclic and closely related events involving the reproductive organs occur in normal, nonpregnant adult women at about monthly intervals between menarche, at approximately age 12 years, and menopause, at approximately age 51 years. These events constitute the menstrual cycle. During each normal menstrual cycle, an ovum matures, is ovulated, and enters the uterine lumen through the fallopian tubes. Steroids secreted by the ovaries effect endometrial changes, allowing implantation if the ovum is fertilized. In the absence of fertilization, ovarian secretion of progesterone and E2 declines, the endometrium sloughs, and menstruation begins. The menstrual cycle requires the coordinated, functional interaction of the hypothalamus, the pituitary gland, and the ovaries (the hypothalamic–pituitary–ovarian axis) to produce associated changes in the target tissues of the reproductive tract (endometrium, cervix, and vagina), which then permit pregnancy and perpetuation of the species.8 Although the individual units of the hypothalamic–pituitary–ovarian axis are innervated, the mediators of the communication also include autocrine, paracrine, and hemocrine mechanisms. By definition, a menstrual cycle begins with the first day of genital bleeding and ends just before the next menstrual period begins. Although the median menstrual cycle length is 28 days, normal menstrual cycles may vary from ~21 to 40 days in length. Menstrual cycle length varies most in the years immediately after menarche and in those immediately preceding menopause.9 The average duration of menstrual flow is 5 ± 2 days, with typical blood loss ranging from 30 to 80 mL.10,11,12 and 13 Tampons and pads each absorb an estimated 20 to 30 mL or more. The normal menstrual cycle can be divided into the follicular and luteal phases. Sometimes an ovulatory phase is delineated as well. The follicular phase, also known as the proliferative or preovulatory phase, begins with the onset of menstruation and ends with ovulation (Fig. 95-2). It is variable in duration and accounts for the range in menstrual cycle length found in ovulatory women. The luteal phase, sometimes termed the postovulatory or secretory phase, begins with ovulation and ends with the onset of menses. It is the more constant half of the menstrual cycle and averages 14 days in length. The ovulatory phase extends from 1 day before the LH surge to the time of ovulation, ~32 to 34 hours after the onset of the preovulatory LH surge.

FIGURE 95-2. Idealized cyclic changes observed in pituitary secretion of gonadotropins, ovarian secretion of estradiol (E2) and progesterone (P), and the uterine endometrium during the reproductive life of the woman. The data have been centered around the day of the luteinizing hormone (LH) surge (day 0). (M, days of menstrual bleeding; FSH, follicle-stimulating hormone.) (From Rebar RW. Normal physiology of the reproductive system. In: Endocrine and metabolism continuing education and quality control program. American Association of Clinical Chemistry, 1982.)

Some women experience dull, unilateral pelvic pain of a few minutes' to a few hours' duration near the time of ovulation. Such pain has been termed mittelschmerz. The pain may occur before, during, or after actual ovulation. Although the pain occurs on the side containing the ovary that releases the oocyte, it is not clear that the pain actually is caused by the physical act of ovulation itself.

HORMONAL CHANGES DURING THE NORMAL MENSTRUAL CYCLE Circulating concentrations of FSH begin to increase in the late luteal phase of the previous menstrual cycle14,15 and 16 (see Fig. 95-2). The increase in FSH levels continues into the early follicular phase and is responsible for initiating the growth and development of a group of follicles. The oocyte that will be ovulated is selected from this cohort undergoing development, but the manner of selection is not understood. FSH levels then fall after the early follicular phase increase. Except for a brief peak at midcycle, FSH levels continue to fall until they reach their lowest levels in the midluteal phase, just before they begin to increase again before menses. Circulating LH concentrations also begin to increase in the late luteal phase of the previous menstrual period.14,15 and 16 However, in contrast to FSH levels, LH concentrations continue to increase gradually throughout the follicular phase. At midcycle, there is a significant increase in circulating LH levels that lasts 1 to 3 days and triggers ovulation. Currently available urinary LH-testing kits detect the LH surge with reasonable accuracy. LH levels gradually decrease in the luteal phase to reach their lowest levels just before beginning to increase again before menses. As is true for virtually all hormones, the gonadotropins (especially LH) are secreted in a pulsatile manner, with intervals of 1 to 4 hours between pulses, depending on the phase of the menstrual cycle.17,18 and 19 (Fig. 95-3). LH pulse frequency is lowest during the luteal phase of the menstrual cycle, apparently because of the effects of progesterone.20 The pulsatile secretion of the gonadotropins is dependent on the pulsatile secretion of GnRH by the hypothalamus.21

FIGURE 95-3. A, Variation in the frequency and magnitude of the pulsatile patterns of circulating luteinizing hormone (LH) and follicle-stimulating hormone (FSH) during different phases of the menstrual cycle. Results are presented in terms of the Second International Reference Preparation for Human Menopausal Gonadotropin (2nd IRP-HMG). B, LH and FSH concentrations by stages of sleep on cycle days 13 and 14. (From Yen SSC, Rebar RW, Vandenberg G, et al. Pituitary gonadotropin responsiveness to synthetic LRF in subjects with normal and abnormal hypothalamic-pituitary-gonadal axis. J Reprod Fertil 1973; 20[Suppl]:137.)

The ovary secretes numerous steroidal and nonsteroidal hormones. Several of the steroids secreted by the ovary also are secreted by the adrenal gland, and some are formed by peripheral conversion from other steroid precursors (Fig. 95-4). Consequently, circulating concentrations do not reflect ovarian production rates. Several steroids secreted by the ovary do vary throughout the menstrual cycle (Fig. 95-5, see Chap. 94).

FIGURE 95-4. Pathways of steroidogenesis. The necessary enzymes are depicted as well as the steroids.

FIGURE 95-5. Steroid patterns during the menstrual cycle. All changes are those observed in circulating concentrations except for total estrogens (Total E) and

2-hydroxyestrone (20HE1), for which changes in 24-hour urinary excretion are shown. (E, estrone; E2, estradiol; E3, estriol; T, testosterone; D4A, D4-androstenedione; DHT, dihydrotestosterone; DHEA, dehydroepiandrosterone; 17OH prog, 17-hydroxyprogesterone; PROG, progesterone; LH, luteinizing hormone.) (From Rebar RW, Yen SSC. Endocrine rhythms in gonadotropins and ovarian steroids with reference to reproductive processes. In: Krieger DT, ed. Endocrine rhythms. New York: Raven Press, 1979:259.)

E2 may be the most important steroid secreted by the ovary because of its biologic potency and its many effects on peripheral target tissues. Circulating levels of E2 are low during the first half of the follicular phase, begin to increase ~7 to 8 days before the preovulatory LH surge, and generally peak at levels of 250 to 350 pg/mL the day before or the day of the LH surge.22,23,24 and 25 As peak LH levels are reached during the ovulatory phase, E2 levels fall rapidly, only to increase again to a secondary peak 6 to 8 days after the LH surge during the midluteal phase26 (Fig. 95-6). Parallel, but smaller, changes occur in circulating estrone levels. The dominant follicle and corpus luteum synthesize ~95% of circulating E2. In contrast, a significant portion of the circulating estrone is converted from E2 and from the peripheral conversion of androstenedione.

FIGURE 95-6. Mean (± SE) luteinizing hormone (LH), follicle-stimulating hormone (FSH), estradiol (E2), and progesterone (P) concentrations in five women, measured at 2-hour intervals for 5 days at midcycle. The initiation of the LH surge has been used as the reference point (at time 0) from which data have been tabulated. The hormone concentrations are plotted on a logarithmic scale. (From Hoff JD, Quigley MD, Yen SSC. Hormonal dynamics at midcycle: a reevaluation. J Clin Endocrinol Metab 1983; 57:792.)

Androstenedione and testosterone, secreted by the interstitial and theca cells, are the primary ovarian androgens. Androstenedione, the major ovarian androgen, also can be converted to testosterone and estrogens in peripheral tissues. Both androstenedione and testosterone also are secreted in significant amounts by the adrenal gland, and both peak at the time of the midcycle LH surge, no doubt because of increased ovarian secretion.23,24,27 In contrast, dehydroepiandrosterone and its sulfate, which are secreted almost entirely by the adrenal gland, do not vary with the menstrual cycle. Circulating levels of progesterone and progesterone secretion remain low throughout the follicular phase and begin to increase just before the onset of the LH surge26,28,29 (see Fig. 95-6). During the luteal phase, progesterone secretion increases to peak 6 to 8 days after the LH surge. Progesterone levels decrease toward menses unless the ovum is fertilized. Serum progesterone levels of 10 ng/mL or greater 1 week before menses generally indicate normal ovulation. Moreover, because progestins increase morning basal body temperature, an “upward shift” of more than 0.3°C orally after a midcycle nadir is a presumptive sign of ovulation and progesterone secretion. Unfortunately, daily measurement of basal temperature is tedious and not very reliable. 17a-Hydroxyprogesterone concentrations actually begin to increase at midcycle, before progesterone and parallel changes in progesterone levels in the luteal phase. Insulin-like growth factor-I (IGF-I) appears to play an important role in stimulating follicular growth and maturation as well as in augmenting steroid production. Although the majority of IGF-I production is peripheral, 30 IGF-I production by the ovary also occurs.31 IGF-I levels in peripheral serum are lowest during menses and highest during the periovulatory period and the luteal phase.32,33 and 34 Serum levels of IGF-I decline with age.35 The main antagonist of IGF-I action is IGF-binding protein-3 (IGFBP-3), which decreases the amount of bioactive IGF-I through binding of the hormone. Serum concentrations of GH have been correlated with those of E2, peaking during the periovulatory period.33,36 GH can potentiate the stimulatory effects of gonadotropins on ovarian follicles. It also stimulates the corpus luteum to increase progesterone production. These effects can be mediated by IGF-I. Inhibin and activin are composed of a family of polypeptide subunits, a, b-A, and b-B, which are produced by follicular granulosa cells as well as the corpus luteum. Because earlier assays for inhibin were specific for the a subunit, they did not distinguish between inhibin A, which is composed of an a subunit and a b-A subunit, and inhibin B, which is composed of an a subunit and a b-B subunit. Therefore, care must be taken in interpreting older studies that used this assay. Thus, although it was previously recognized that inhibin is often a negative regulator of FSH, current assays, which distinguish between inhibin A and B, provide a clearer delineation of their individual roles. Inhibin A begins to increase as the follicle grows and reaches its peak during the luteal phase. However, estradiol, not inhibin A, seems to be the predominant suppressor of FSH release during the luteal phase.37 Serum inhibin B peaks during the follicular phase, apparently in response to increasing FSH.38 Inhibin B is then critical for suppressing FSH release and inducing a plateau in FSH serum levels in the midfollicular phase; however, estradiol is instrumental in further reducing FSH levels.37 The subsequent rise in LH and fall in FSH in the late follicular phase also are believed to be due in part to the fact that E2 is more inhibitory of FSH than of LH secretion.19 Unlike inhibin, activin, a dimer composed of two b subunits in any combination, does not vary during the menstrual cycle.39 The primary function of follistatin, which is concentrated 100-fold within follicular fluid, appears to be regulation of activin action by binding the hormone. FSH and prostaglandin E2 promote follistatin production in granulosa cells.40 However, reports of menstrual cycle variation of follicular fluid levels of follistatin have been inconsistent. Circulating levels of several other hormones that do not seem to be directly related to ovulation (i.e., prolactin, cortisol, parathyroid hormone, calcitonin, estrogen-sensitive neurophysin, and catechol estrogens) appear to peak at midcycle.20,41,42,43 and 44 Although the physiologic significance of such changes is unclear, the cyclic fluctuation of these hormones suggests that endogenous estrogen may modulate the secretion of other hormones.

CYCLIC CHANGES IN THE TARGET ORGANS OF THE REPRODUCTIVE TRACT ENDOMETRIUM During the menstrual cycle, the endometrium undergoes a series of histologic and cytologic changes that culminate in menstruation if pregnancy does not result45,46 and 47 (see Fig. 95-2). The basal layer of the endometrium, nearest the myometrium, undergoes little change during the menstrual cycle and is not shed during menses. The basal layer regenerates an intermediate spongiosa layer and a superficial compact epithelial cell layer, both of which are sloughed at each menstruation. Under the influence of estrogen, with IGF-I likely acting as a paracrine mediator, the endometrial glands in these two functional layers proliferate during the follicular phase, leading to thickening of the mucosa. During the luteal phase, the glands become coiled and secretory under the influence of progesterone, with IGF-II being the suspected paracrine mediator.48 The endometrium becomes much more edematous and vascular, largely because spiral arteries develop in the functional layers. With the decline of both E2 and progesterone in the late luteal phase, endometrial and blood vessel necrosis occurs, and menstrual bleeding begins. The local secretion of prostaglandins appears to initiate vasospasm and consequent ischemic necrosis of the endometrium, as well as the uterine contractions that frequently occur with menstruation.49 Thus, prostaglandin synthetase inhibitors can relieve dysmenorrhea (i.e., menstrual cramping).50 Fibrinolytic activity in the endometrium also peaks during menstruation, thus explaining the noncoagulability of menstrual blood.51 Because of the characteristic histologic changes that occur during the menstrual cycle, endometrial biopsies can be used to date the stage of the menstrual cycle and to assess the tissue response to gonadal steroids.46,47 Transvaginal ultrasound is a less invasive modality that has a 76% accuracy in assessing endometrial stage as compared to biopsy.52

CERVIX AND CERVICAL MUCUS Under the influence of estrogen, cyclic changes occur in the diameter of the external cervical os, the dimensions of the external cervical canal, the vascularity of the cervical tissues, and the amount and biophysical properties of cervical mucus53,54 (Fig. 95-7). Normally, sex steroids are present in cervical mucus.52a During the follicular phase, there is a progressive increase in cervical vascularity, congestion, and edema, as well as in the secretion of cervical mucus. The external cervical os opens to a diameter of 3 mm at ovulation and then decreases to 1 mm. Under the influence of increasing levels of estrogen, several changes in cervical mucus occur. There is a 10- to 30-fold increase in the amount of cervical mucus. The elasticity of the mucus, also known as spinn-barkeit, increases. Just before ovulation, “palm leaf” arborization or ferning becomes prominent when cervical mucus is allowed to dry on a glass slide and is examined microscopically (Fig. 95-8). This ferning is a result of the increased sodium chloride concentration in the cervical mucus induced by rising levels of estrogen. The pH of the mucus increases to ~8.0 at midcycle as well. Under the influence of progesterone, during the luteal phase, cervical mucus thickens, becomes less watery, and loses its elasticity and ability to fern. These characteristics of cervical mucus can be used clinically to help evaluate the stage of the menstrual cycle, as in the Billings method,55 to help a woman time her ovulation (although with poor precision).

FIGURE 95-7. Changes in the composition and properties of cervical mucus during the menstrual cycle. (From Goldfien A, Monroe S. The ovaries. In: Greenspan FS, Forsham PH, eds. Basic and clinical endocrinology, 2nd ed. Los Altos, CA: Lange Medical Publications, 1986:400.)

FIGURE 95-8. A, Absence of arborization (i.e., ferning) in a smear of cervical mucus obtained during the immediate postmenstrual period (day 5) from a normally menstruating woman (×88). B, Ferning in a smear obtained from the same woman just before ovulation (×88). The mucus was allowed to dry thickly on a microscope slide and then photographed through a microscope without staining or fixing. (From Rebar RW. Practical evaluation of hormonal status. In: Yen SSC, Jaffe RB, eds. Reproductive endocrinology: physiology, pathology and clinical management, 3rd ed. Philadelphia: WB Saunders, 1991:830.)

VAGINAL MUCOSA Both proliferation and maturation of the vaginal epithelium are influenced by estrogens and progesterone.56,57 Three types of vaginal cells are exfoliated: (a) mature superficial cells, which are squamous epithelial cells with pyknotic nuclei; (b) intermediate cells, which are relatively mature squamous epithelial cells with vesicular, nonpyknotic nuclei; and (c) parabasal cells, which are thick, small, round immature cells with large vesicular nuclei. Parabasal cells predominate before puberty, after menopause, and in women with estrogen-deficient forms of amenorrhea. When ovarian estrogen secretion is low in the early follicular phase, the vaginal epithelium is thin and pale. As E2 levels increase in the follicular phase, there is an increase in the number of cells, the thickness of the epithelium, and the number of cornified superficial cells. Under the influence of progesterone during the luteal phase, the percentage of cornified cells decreases while the number of precornified intermediate cells increases, and there are increased numbers of polymorphonuclear leukocytes and increased cellular debris and clumping of shed (desquamated) cells. Thus, the ratio among the various desquamated vaginal cells, obtained by fixing a scraping from the upper third of the lateral vaginal wall on a microscope slide, can be used to evaluate estrogen effect. This ratio has been termed the maturation index. Cytologic changes similar to those observed in vaginal cells also are seen in the lower urinary tract, particularly the urethra. This fact has been used clinically to examine the cells in the urinary sediment from children suspected of having prematurely increased estrogen levels.58

CENTRAL AND GONADAL FEEDBACK MECHANISMS IN THE CONTROL OF OVULATION HYPOTHALAMIC-PITUITARY SIGNALS For normal reproduction, the principal hormone that allows gonadotropin secretion from the pituitary gland is the decapeptide GnRH (also known as luteinizing hormone–releasing hormone).21,59 Although GnRH is present in several hypothalamic regions, its greatest concentration is localized to the arcuate nucleus. The physiologic importance of this association is apparent from numerous observations that specific anatomic lesions within the arcuate nucleus result in hypogonadotropic hypogonadism secondary to either the absence of or deficiencies in gonadotropin secretion (see Chap. 8, Chap. 16 and Chap. 96). GnRH from the arcuate nucleus is transported to the base of the hypothalamus, and more specifically to the median eminence, where it is released into the pituitary portal vascular bed (Fig. 95-9). This relatively closed vascular system directly links the hypothalamus to the pituitary and allows high concentrations of GnRH to affect pituitary gonadotropes directly without systemic passage. Thus, peripheral blood levels of GnRH and the other releasing hormones do not accurately reflect hypothalamic-pituitary interaction. Extremely high levels of GnRH may be present in the portal circulation when peripheral levels are undetectable.60

FIGURE 95-9. Neuroanatomic relationships among various neurons within the preoptic–anterior hypothalamic and arcuate nucleus–median eminence regions of the

brain that affect gonadotropin secretion. (GnRH, gonadotropin-releasing hormone.) (Modified from Yen SSC. Neuroendocrine regulation of the menstrual cycle. Hosp Pract 1979; 14:83.)

GnRH-secreting neurons in culture secrete GnRH in a pulsatile manner.61 Furthermore, it is apparent that GnRH must be provided to the pituitary gland in such a pulsatile manner to stimulate normal adult ovarian function and ovulation.62 The requirement for pulsatile GnRH presentation to the pituitary was proved when monkeys with endogenous GnRH secretion eliminated by selective placement of lesions in the hypothalamic arcuate nucleus were given replacement therapy with exogenous GnRH. GnRH given for 6 minutes every 60 minutes effectively induced ovulation and resulted in normal corpus luteum function.21 Continuously administered GnRH was ineffective in inducing ovulation. From the classic experimental model,21 it is clear that the frequency and amplitude of infused GnRH pulses alter the quantitative secretion of LH and FSH. Continuous GnRH infusion results in less gonadotropin secretion (so-called down-regulation), either because of occupation of all receptors so that stimulation is impossible or because of internalization of receptors such that there is overt refractoriness to stimulation. Information about responsiveness to exogenous GnRH is being used clinically. With commercially available portable pump technology, exogenous GnRH can be administered at regular intervals (60–120 minutes) either intravenously or subcutaneously over a dose range of 1 to 20 µg per pulse to induce ovulation63,64 (see Chap. 97). Following an initial stimulation of gonadotropin secretion, long-acting GnRH agonists will, within 3 weeks of treatment, down-regulate gonadotropin release, creating a state termed reversible menopause.65,66 and 67 The use of GnRH antagonists has been limited by incidental activation of mast-cell receptors, causing histamine release. Newer-generation GnRH antagonists, however, show great promise, in producing much less mast-cell activation. Down-regulation with GnRH analogs can be used to treat steroid-dependent pathologic processes, including, among others, leiomyomas, endometriosis, hirsutism, precocious puberty, and dysfunctional uterine bleeding. Future applications of similar technologies may provide new strategies for female and male contraception (see Chap. 104 and Chap. 123). It appears that classic neurotransmitters (e.g., norepinephrine, dopamine, serotonin) as well as neuromodulators (e.g., endogenous opiates, prostaglandins) influence the secretion of GnRH by the hypothalamus.62,68,69,70,71,72,73,74 and 75 In addition, estrogens and androgens can bind to receptors in cells in the hypothalamus and anterior pituitary,76 and progestins77 can bind to cells in the hypothalamus, to influence hypothalamic-pituitary regulation of ovarian function. Nonsteroidal ovarian factors, as noted previously, also play a role in the central control of ovulation. In humans and other primates, the reproductive axis is seemingly dormant until puberty. It has become clear that this so-called immaturity of the reproductive axis is at the neural level of the brain, above the median eminence.78 When exogenous GnRH is administered in appropriate fashion to juvenile primates, ovulation can be induced despite the chronologic and developmental age.21 This postnatal and prepubertal “off” stage of GnRH secretion in the juvenile has been likened to an electronic control system and termed a gonadostat.78 The nature of this neuronal “clamp” is unclear, but such disparate stimuli as increased intracranial pressure or premature exposure to sex steroids (such as occurs in congenital adrenal hyperplasia) can prematurely lift the normal inhibition of GnRH secretion, leading to precocious puberty. With the onset of normal puberty, GnRH secretion first begins at night (see Chap. 91). STEROIDAL FEEDBACK Gonadal steroids can exert both negative and positive feedback effects on gonadotropin secretion. Among ovarian steroids, E2 is the most potent inhibitor of gonadotropin secretion. Thus, ovariectomy leads to a rapid increase in gonadotropins,79 and the infusion of 17b-estradiol into women with hypoestrogenemia leads to almost immediate decreases in both LH and FSH80 (Fig. 95-10). Although low concentrations of estrogen inhibit the secretion of gonadotropins by the pituitary, they also stimulate synthesis and storage of gonadotropins.

FIGURE 95-10. The inhibitory effect (i.e., negative feedback) of 17b-estradiol (E2) on the release of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) in agonadal women. The fall in circulating gonadotropin concentrations during and after a 4-hour infusion of E2 (50 µg per hour) is expressed as percent of basal (i.e., control) concentrations. (From Yen SSC, Tsai CC, Vandenberg G, Rebar RW. Gonadotropin dynamics in patients with gonadal dysgenesis: a model for the study of gonadotropin regulation. J Clin Endocrinol Metab 1972; 35:897.)

For women to ovulate, E2 must have the ability to elicit a positive as well as a negative effect on gonadotropin secretion.81,82 and 83 Although the ability of ovarian estrogen to exert both negative and positive feedback effects may seem paradoxical, the development of positive feedback is known to require an estrogenic stimulus of increasing strength and duration.84 High concentrations of estrogen stimulate synthesis and storage of gonadotropin but also augment the effect of GnRH in eliciting release of gonadotropin.85,86 In the normal menstrual cycle, the positive feedback action of E2 leading to the LH surge is preceded by a period when lower E2 levels are present, with their negative feedback effects (Fig. 95-11).

FIGURE 95-11. Differences in the feedback effects of exogenous estradiol (E2) administered to normal women during the early (low endogenous estrogen) and midfollicular (moderately high endogenous estrogen) phases of the menstrual cycle. E2 was administered at a dosage of 200 µg per day for 3 days on both occasions. Although these data are generally presented as depicting the stimulatory effects (i.e., positive feedback) of E2 on gonadotropin secretion, the effects are really biphasic. (FSH, follicle-stimulating hormone; LH, luteinizing hormone.) (From Yen SSC, Vandenberg G, Tsai CC, Parker DC. Causal relationship between the hormonal variables in the menstrual cycle. In: Ferin M, Hal-berg F, Richart RM, Van de Wiele RL, eds. Biorhythms and human reproduction. New York: John Wiley and Sons, 1974:219.)

PRESUMED MECHANISM FOR THE LUTEINIZING HORMONE SURGE The basic activity of the pituitary gonadotropes is determined by the direct input of GnRH but is modulated by the feedback effects of E2. Data now suggest the existence of two functionally separate pools of gonadotropin: one that is acutely releasable (and has been termed sensitivity) and a second that is released only with sustained stimulation (termed reserve).86 Together, these pools define pituitary capacity. During the early follicular phase, when E2 levels are low, both gonadotropin sensitivity and reserve are at a minimum. As E2 levels increase during the midfollicular phase, a preferential increase in reserve occurs first. As E2 increases further toward midcycle, both sensitivity and reserve become maximal until the sensitivity becomes such that the midcycle release of LH occurs. The surge also is partly a result of what is an estrogen-dependent self-priming effect of GnRH: A second pulse of GnRH elicits greater release of gonadotropin than does the first pulse in the estrogen-primed state87 (Fig. 95-12).

FIGURE 95-12. The self-priming effect of gonadotropin-releasing hormone (GnRH; i.e., luteinizing hormone–releasing hormone [LHRH]) appears to be separate from the ability of GnRH to release LH. Minute doses of GnRH, infused continuously for 4 hours, did not increase circulating levels of LH but did increase the response to a bolus of GnRH (10 µg intravenously administered at the arrows) compared to control responses to the bolus alone (open circles). (From Hoff JD, Lasley BL, Yen SSC. The functional relationship between priming and releasing actions of LRF [LHRH]. J Clin Endocrinol Metab 1979; 49:8.)

Whether the pulsatile release of GnRH is increased at midcycle in women has not been determined. Studies with exogenous GnRH in both nonhuman primates and humans have demonstrated conclusively that the midcycle surge can occur without any increase in GnRH release.21,63,64 However, a surge in GnRH has been shown to accompany the E2-induced gonadotropin surge in monkeys.88 On the other hand, the onset of the midcycle surge may merely reflect a rapid estrogen-stimulated increase in the number of GnRH receptors on the gonadotropes and the attainment of maximal capacity by the gonadotropes.89 Estrogen alone, when administered in a manner designed to mimic the physiologic blood levels that normally occur in the late luteal phase, induces an LH surge in women and monkeys. This surge, however, is not identical to that observed in ovulatory women.26 The characteristics of the artificially induced surge are much more similar to those of the physiologic midcycle surge if progesterone is administered as well as E2.90 It is believed that the rising levels of progesterone at midcycle bring the LH surge to its end by inhibiting GnRH secretion in the hypothalamus and diminishing the sensitivity of the gonadotropes to GnRH in the pituitary.91,92

EFFECT OF THE OVARY ON GONADOTROPIN SECRETION According to Knobil,21 the ovary is the “zeitgeber” for the timing of ovulation, with the hypothalamus stimulating pulsatile release of the gonadotropes. In turn, the follicular complex and corpus luteum of the ovary develop in response to gonadotropin stimulation as detailed in Chapter 94. For appropriate ovarian regulation of reproductive function in women, at least five biologic characteristics appear necessary: 1. 2. 3. 4. 5.

Appropriate negative and positive feedback actions of gonadal steroids on gonadotropin secretion Differential feedback effects of ovarian secretions on the release of LH and FSH Local intraovarian controls on follicular growth and maturation, separate from but interrelated to the effects of gonadotropins on the ovary Appropriate development of oocytes so that ovulation may occur Appropriate development of the endometrium so that implantation may occur if fertilization results

RECRUITMENT AND SELECTION OF THE DOMINANT FOLLICLE IN OVULATORY MENSTRUAL CYCLES In the absence of pharmacologic intervention, multiple ovulation is extremely atypical in women. The primordial follicle must slowly develop and grow for many months before it becomes a 5- to 8-mm antral follicle at the beginning of the cycle during which it will potentially ovulate.93 Normally, many follicles reach this stage at the first half of each follicular phase (see Chap. 94). The process that follows has been termed recruitment94,94a (Fig. 95-13). Several morphologically identical follicles may be observed within the ovary before cycle days 5 to 7. The destruction of any one of these follicles does not delay ovulation. In contrast, after about cycle day 7, the multipotentiality of these follicles is lost. Henceforth, only one follicle is capable of progressing to ovulation in the current cycle. This one follicle, destined to ovulate and form the corpus luteum, is known as the dominant follicle. Destruction of the dominant follicle, such as by selective cautery, delays ovulation by approximately the number of days that have passed from cycle onset to follicle destruction.95 The point in time in the cycle at which all of the recruited follicles become qualitatively unequal in potential is the time of selection. That the process of selection is predetermined by some intrinsic aspect of a particular follicle seems unlikely. However, once acquired, dominance cannot be transferred. On selection of the single dominant follicle, all other follicles become destined for atresia.

FIGURE 95-13. Time course for recruitment, selection, and ovulation of the dominant ovarian follicle (DF), with onset of atresia among other follicles of the cohort (N-1) in the natural ovarian/menstrual cycle. (From Hodgen GD. Fertil Steril 1982; 38:218.)

Atresia of the nondominant follicles occurs through apoptosis or regulated cell death, most notably observed in granulosa cells. In the rodent model, a number of agents have a role in regulating apoptosis and in rescuing the dominant follicle from this fate (Table 95-1 and Table 95-2). These actions are thought to be similar in humans. Activin, while it serves to help rescue the dominant follicle, counteracts the vital rescuing effects of FSH in preantral follicles. This seems to be part of the mechanism by which secondary follicles block the stimulation of primary follicles.96

TABLE 95-1. Granulosa Cell Survival Factors

TABLE 95-2. Granulosa Cell Apoptotic Factors

The growth of the follicle, followed by the decrease in follicular diameter (which signals ovulation), and formation of the corpus luteum can be observed most accurately by transvaginal ultrasound.97 Data from ultrasonographic studies in normal women indicate that the site of ovulation occurs randomly in consecutive cycles and does not alternate between the two ovaries.98 With removal of one ovary, ovulation occurs in the single remaining ovary each month. At present, there is no good evidence that removal of one ovary (and even a small portion of the second) decreases the number of ovulatory menstrual cycles that the average woman has during her reproductive years. The process of selection of the dominant follicle is overridden in the presence of supraphysiologic gonadotropin stimulation of the ovary. Typically, exogenous gonadotropin therapy allows several recruited follicles to avert atresia. Although the development of several stimulated follicles often is not perfectly synchronous, a few are likely to be mature enough for ovulation, fertilization, and implantation (Fig. 95-14). The ability of supraphysiologic stimulation to recruit several follicles is used effectively in ovulation induction and in vitro fertilization (see Chap. 97).

FIGURE 95-14. Pattern of follicle growth during sustained elevations of follicle-stimulating hormone and luteinizing hormone (LH, with human menopausal gonadotropin [hMG]). Only a few follicles can develop quasisynchronously. Accordingly, if human chorionic gonadotropin (hCG) is given too late in the attempt to mimic the endogenous LH surge, follicles 1 and 2 may be postmature; if hCG is given too soon, other follicles will be immature. (From Hodgen GD. Physiology of follicular maturation. In: Jones HW Jr, Jones GF, Hodgen GD, Rosen-waks Z, eds. In vitro fertilization—Norfolk. Baltimore: Williams & Wilkins, 1986:8.)

Of interest is the observation that in women undergoing supraphysiologic stimulation with exogenous gonadotropins, the high levels of endogenous estrogens, with or without progesterone, frequently fail to promote a timely or full LH surge. This effect is likely due to a nonsteroidal substance, different from inhibin, that has been termed gonadotropin surge-inhibiting or attenuating factor (GnSI/AF) and can be isolated from follicular fluid by charcoal extraction. GnSI/AF has the ability to suppress GnRH-induced LH secretion in humans without affecting basal FSH production. Whether the putative factor is the C-terminal fragment of human serum albumin has not yet been confirmed.99

ROLE OF ESTRADIOL IN FOLLICULAR DEVELOPMENT It is clear that the secretion of E2 by granulosa cells is critical for the occurrence of normal menstrual cycles. E2 plays an essential role in feedback to the central nervous system and in preparing the endometrium for implantation of the developing blastocyst. It had been believed that E2 also played a central role within the ovary in the developing dominant ovarian follicle. This concept, derived from experiments involving rodent models, has been questioned based on findings in women. It has been observed that normal follicular growth and development, and successful fertilization in vitro, could be achieved using exogenous gonadotropins in a person with 17a-hydroxylase deficiency who, therefore, had no ability to synthesize androgens or estrogens.100 Thus, in contrast to the rodent paradigm, estrogens are not essential for follicular development in humans. However, the finding that estradiol receptors are present in human granulosa cells has opened the possibility that estrogen may play a facilitory role in follicle development. A number of already mentioned peptides appear to play more critical roles in intraovarian regulation. The physiologic roles of these peptides remain to be defined precisely, but data documenting their modulatory actions are accumulating.101 It appears that FSH initially stimulates activin and inhibin synthesis by granulosa cells. In immature granulosa cells, activin augments FSH action, especially FSH-receptor expression and aromatase activity.102 Inhibin appears to enhance LH stimulation of androgen synthesis in theca cells to serve as substrate for aromatization to estrogen in granulosa cells. Inhibin also inhibits FSH secretion centrally at the level of the gonadotrope. In the luteinizing granulosa cells of the dominant follicle just before ovulation, inhibin synthesis appears to come under control of LH. In the granulosa cells of growing follicles, IGF-I and IGF-II stimulate aromatase activity and cell proliferation.103 These effects are modulated by several IGFBPs.

MECHANISMS OF OVULATION In normal ovulatory cycles, the mean interval from the late follicular phase peak estrogen level to the peak LH level is ~24 hours. Ovulation follows ~9 hours later, for a total interval of ~33 hours.104 In in vitro fertilization protocols with human menopausal gonadotropin-induced ovarian stimulation and human chorionic gonadotropin (hCG) supplementation, follicular rupture seldom occurs until 36 hours after hCG administration (mimicking the endogenous LH surge). Similarly, initial ovulation

seldom occurs until 34 or more hours after treatment with clomiphene citrate and hCG. The actual events that lead to expulsion of the oocyte and that occur after either the natural or the induced LH surge are incompletely understood. It is not surprising that many of the biochemical and biophysical processes surrounding ovulation can be mimicked in vitro by cyclic adenosine monophosphate, in view of the fact that gonadotropins act by binding to receptors that are linked to cyclic adenosine monophosphate (see Chap. 94). Several possible factors may play roles in extrusion of the oocyte and the granulosa cells immediately surrounding it (i.e., the oocyte-cumulus complex): 1. Granulosa cells produce large amounts of plasminogen activator, apparently stimulated by gonadotropins.105 Because plasminogen is present in follicular fluid and plasmin can weaken follicle wall strips in vitro, LH-mediated enzymatic digestion of the follicle wall may be important in ovulation. Both interleukin-1b (IL-b) and tumor necrosis factor (TNF), which are in follicular fluid, are known to suppress the plasminogen activator system, thereby likely helping to prevent premature follicular rupture.106 However, they also stimulate prostaglandins, which are involved in promoting ovulation. 2. Maturation of the oocyte-cumulus complex involves the dispersal and expansion of the corona radiata (see Chap. 94). Before actual follicular rupture and oocyte extrusion, the oocyte-cumulus complex detaches itself from the granulosa cells of the follicle wall (i.e., the membrana granulosa). In mice, FSH-dependent deposition of a glycosaminoglycan is closely associated with cumulus expansion107 and may be necessary for ovulation. 3. Because the outer structure of the follicle wall contains smooth muscle, muscular contractions may be important in oocyte extrusion. In this regard, LH and IL-1b can stimulate prostaglandin synthesis, and prostaglandin F2a (PGF2a) stimulates ovarian smooth muscle activity.108,109 4. Follicular fluid has been found to contain an angiogenic factor.110 It seems likely that this substance plays a role in the delivery of gonadotropins to the developing dominant follicle. Whether this substance also might promote follicular rupture remains to be explored. It is possible that abnormalities in the process of oocyte extrusion account for the infrequent occurrence of “luteinized unruptured follicles” (see Chap. 96).

LUTEAL FUNCTION AS THE SEQUEL TO FOLLICULOGENESIS In most respects, the events of the luteal phase are consequences of preceding follicular phase activities. Indeed, the process of luteinization begins even before the time of ovulation (see Chap. 94). The granulosa and theca cells that, together with the oocyte, form the dominant follicle and secrete large amounts of estrogens and regulatory peptides are transformed into the corpus luteum after ovulation. In the luteal phase, these same cells produce progesterone as their primary secretory product, but estrogens still are produced in large amounts as well. The synergism of high estrogen levels and FSH in the late follicular phase induces LH receptors on granulosa cells and leads to progesterone secretion even before the LH surge. This change in granulosa cell function is known as luteinization.111 It would seem, therefore, that the greater the proliferation of FSH-stimulated granulosa cells in the follicular phase, the greater will be the transformed luteinized cell mass for progesterone production and early pregnancy support. Luteinization of granulosa and theca cells occurs only within the dominant follicle; nearby granulosa cells remain unaffected. Thus, intraovarian regulators have been suggested as important in causing this localized phenomenon as well. A luteinizing inhibitor is a convenient concept for explaining why other nearby ovarian cells do not undergo luteinization, especially because all granulosa cells removed from the ovary and cultured in vitro appear to luteinize spontaneously. Such a substance remains to be isolated and characterized. The corpus luteum is not an autonomously functioning unit that is independent of gonadotropin stimulation. Primate models have provided evidence that LH is essential for the maintenance of the corpus luteum. If a GnRH antagonist that blocks LH secretion is administered during the midluteal phase, that menstrual cycle is significantly shortened and menses begins prematurely.112 The neutralization of LH by antibody administration also causes premature demise of the corpus luteum.113 As is true for most other hormones, progesterone is secreted by the corpus luteum in a pulsatile fashion.114,115 These pulses are most frequent in the early and midluteal phases, when progesterone secretion is greatest. The pulses appear to correlate with pulses of LH as well. Still, as noted previously, pulse frequency for LH is lowest during the luteal phase, so that progesterone pulses typically occur at 4- to 8-hour intervals. Progesterone production is also enhanced by IGF-I and IGF-II, likely through increasing levels of prostaglandin E2 (PGE2).116 The mechanisms responsible for regression of the corpus luteum (i.e., luteolysis) in women are unknown and are considered in more detail in Chapter 94.117,118 The life span of the corpus luteum may depend in part on prostaglandins and prolactin, as well as on progestin. If fertilization does occur, hCG, which is biologically similar to LH, is secreted by the developing blastocyst and helps to support the corpus luteum until the fetoplacental unit can support itself (see Chap. 108). However, in the absence of a viable fetus, hCG supports the corpus luteum for only a short time in humans. Despite the limits of current knowledge, an understanding of the corpus luteum and luteolysis is essential if rational therapies are to be provided to women with luteal phase defects (see Chap. 96). CHAPTER REFERENCES 1. Rebar RW, Yen SSC. Endocrine rhythms in gonadotropins and ovarian steroids with reference to reproductive processes. In: Krieger DT, ed. Endocrine rhythms. New York: Raven Press, 1979:259. 2. Rebar RW. Normal physiology of the reproductive system. In: Endocrine and metabolism continuing education and quality control program. American Association of Clinical Chemistry, 1982. 3. Genazzini AR, Lemarchand-Beraud T, Aubert ML, et al. Patterns of plasma ACTH, hGH and cortisol during the menstrual cycle. J Clin Endocrinol Metab 1975; 41:431. 4. Yen SSC, Vela P, Rankin J, et al. Hormonal relationships during the menstrual cycle. JAMA 1970; 211:1513. 5. Ehara Y, Suer T, VandenBerg G, et al. Circulating prolactin levels during the menstrual cycle: episodic release and diurnal variation. Am J Obstet Gynecol 1973; 117:962. 6. Vekemans M, Delvoye P, L'Hermite M, et al. Serum prolactin levels during the menstrual cycle. J Clin Endocrinol Metab 1977; 44:989. 7. Pitkin RM, Reynolds WA, Williams GA, et al. Calcium metabolism during the human menstrual cycle. Gynecol Obstet Invest 1977; 8:60. 8. Hodgen GD. Neuroendocrinology of the normal menstrual cycle. J Reprod Med 1989; 34(Suppl 1):68. 9. Treloar AE, Boynton RE, Benn BG, Brown BW. Variation of human menstrual cycle through reproductive life. Int J Fertil 1967; 12:77. 10. Baldwin RM, Whalley PJ, Pritchard JA. Measurements of menstrual blood loss. Am J Obstet Gynecol 1961; 81:739. 11. Hytten FE, Cheyne GA, Klopper AI. Iron loss at menstruation. J Obstet Gynaecol Br Commonw 1964; 71:255. 12. Hallberg L, Hogdahl A, Nilsson L, Rybo G. Menstrual blood loss—a population study. Acta Obstet Gynecol Scand 1966; 45:320. 13. Rybo G. Menstrual blood loss in relation to parity and menstrual pattern. Acta Obstet Gynecol Scand 1966; 7:119. 14. Midgley AR Jr, Jaffe RB. Regulation of gonadotropins. IV. Correlations of serum concentrations of follicle-stimulating and luteinizing hormones during the menstrual cycle. J Clin Endocrinol Metab 1968; 28:1699. 15. Ross GT, Cargille CM, Lipsett MB, et al. Pituitary and gonadal hormones in women during spontaneous and induced ovulatory cycles. Recent Prog Horm Res 1970; 26:1. 16. Vande Wiele RL, Bogumil RJ, Dyrenfurth I, et al. Mechanisms regulating the menstrual cycle in women. Recent Prog Horm Res 1970; 26:63. 17. Midgley AR Jr, Jaffe RB. Regulation of human gonadotropins. X. Episodic fluctuation of LH during the menstrual cycle. J Clin Endocrinol Metab 1971; 33:962. 18. Yen SSC, Tsai CC, Naftolin F, et al. Pulsatile patterns of gonadotropin release in subjects with and without ovarian function. J Clin Endocrinol Metab 1972; 34:671. 19. Yen SSC, VandenBerg G, Tsai CC, Parker DC. Ultradian fluctuations of gonadotropins. In: Ferin M, Halberg F, Richart RM, Vande Wiele RL, eds. Biorhythms and human reproduction. New York: John Wiley and Sons, 1974:203. 20. Soules MR, Steiner RA, Clifton DK, et al. Progesterone modulation of pulsatile luteinizing hormone secretion in normal women. J Clin Endocrinol Metab 1984; 58:378. 21. Knobil E. The neuroendocrine control of the menstrual cycle. Recent Prog Horm Res 1980; 36:53. 22. Mikhail G. Hormone secretion by the human ovaries. Gynecol Invest 1970; 1:5. 23. Lloyd CW, Lobotsky J, Baird DT, et al. Concentration of unconjugated estrogens, androgens and gestagens in ovarian and peripheral venous plasma of women: the normal menstrual cycle. J Clin Endocrinol Metab 1971; 32:155. 24. Tagatz GE, Gurpide E. Hormone secretion by the normal human ovary. In: Greep PO, Astwood E, eds. Handbook of physiology, section 7. Endocrinology, vol II. Female reproductive system, part I. Washington, DC: American Physiological Society, 1973:603. 25. Baird DT, Fraser IS. Blood production and ovarian secretion rates of estradiol-17b and estrone in women throughout the menstrual cycle. J Clin Endocrinol Metab 1974; 38:1009. 26. Hoff JD, Quigley MD, Yen SSC. Hormonal dynamics at midcycle: a reevaluation. J Clin Endocrinol Metab 1983; 57:792. 27. Judd HL, Yen SSC. Serum androstenedione and testosterone levels during the menstrual cycle. J Clin Endocrinol Metab 1973; 36:475. 28. Yen SSC, Vela P, Rankin J, Littell AS. Hormonal relationships during the menstrual cycle. JAMA 1970; 211:1513. 29. Aido A-R, Landgren B-M, Cekan Z, Diczfalulsy E. Studies on the pattern of circulating steroids in the normal menstrual cycle. 2. Levels of 20a-dihy-droprogesterone, 17-hydroxyprogesterone and 17-hydroxypregnenolone and the assessment of their value for ovulation prediction. Acta Endocrinol (Copenh) 1976;82:600. 30. Pellegrini S, Fuzzi B, Pratesi S, et al. In-vivo studies on ovarian insulin-like growth factor I concentrations in human preovulatory follicles and human ovarian circulation. Hum Reprod 1995; 10:1341. 31. Devoto L, Kohen P, Castro O, et al. Multihormone regulation of progesterone synthesis in cultured human midluteal cells. J Clin Endocrinol Metab 1995; 80:1566. 32. Helle SI, Anker GB, Meadows KA, et al. Alterations in the insulin-like growth factor system during the menstrual cycle in normal women. Maturitas 1998;28:259. 33. Ovesen P, Vahl N, Fisker S, et al. Increased pulsatile, but not basal, growth hormone secretion rates and plasma in insulin-like growth factor I levels during the periovulatory interval in normal women. J Clin Endocrinol Metab 1998; 83:1662. 34. Juul A, Scheike T, Pedersen AT, et al. Changes in serum concentrations of growth hormone, insulin, insulin-like growth factor and insulin-like growth factor–binding proteins 1 and 3 and urinary growth hormone excretion during the menstrual cycle. Hum Reprod 1997; 12:2123. 35. Klein NA, Battaglia DE, Miller PB, et al. Ovarian follicular development and the follicular fluid hormones and growth factors in normal women of advanced reproductive age. J Clin Endocrinol Metab 1996; 81:1946. 36. Amato G, Izzo A, Tucker A, Bellastella A. Insulin-like growth factor binding protein-3 reduction in follicular fluid in spontaneous and stimulated cycles. Fertil Steril 1998; 70:141. 37. Lahlou N, Chabbert-Buffet N, Christin-Maitre S, et al. Main inhibitor of follicle stimulating hormone in the luteal-follicular transition: inhibin A, oestradiol, or inhibin B? Hum Reprod 1999; 14:1190.

38. Fraser HM, Groome NP, McNeilly AS. Follicle-stimulating hormone–inhibin B interactions during the follicular phase of the primate menstrual cycle revealed by gonadotropin-releasing hormone antagonist and antiestrogen treatment. J Clin Endocrinol Metab 1999; 84:1365. 39. Demura R, Suzuki T, Tajima S, et al. Human plasma free activin and inhibin levels during the menstrual cycle. J Clin Endocrinol Metab 1993; 76:1080. 40. Tuuri T, Ritovs O. Regulation of the activin-binding protein follistatin in cultured human luteinizing granulosa cells: characterization of the effects of follicle stimulating hormone, prostaglandin E2, and different growth factors. Biol Reprod 1995; 53:1508. 41. Ebara Y, Siler T, VandenBerg G, et al. Circulating prolactin levels during the menstrual cycle: episodic release and diurnal variation. Am J Obstet Gynecol 1973; 117:962. 42. Genazzini AR, Lemarchand-Beraud TH, Aubert ML, Felber JP. Patterns of plasma ACTH, hGH and cortisol during the menstrual cycle. J Clin Endocrinol Metab 1975; 41:431. 43. Legros JJ, Franchimont P, Burger H. Variations of neurohypophyseal function in normally cycling women. J Clin Endocrinol Metab 1975; 41:54. 44. Vekemans M, Delvoye P, L'Hermite M, Robyn C. Serum prolactin levels during the menstrual cycle. J Clin Endocrinol Metab 1977; 44:989. 45. Noyes RW, Hertig AT, Rock J. Dating the endometrial biopsy. Fertil Steril 1950; 1:3. 46. Tredway DR, Mishell DR Jr, Moyer DL. Correlation of endometrial dating with luteinizing hormone peak. Am J Obstet Gynecol 1973; 117:1030. 47. Bulleti C, Gallassi A, Parmeggiani R, Polli V. Dating the endometrial biopsy by flow cytometry. Fertil Steril 1994; 62:96. 48. Guidice LC, Mark SP, Irwin JC. Paracrine actions of insulin-like growth factors and IGF binding protein-1 in non-pregnant human endometrium and at the decidual-trophoblast interface. J Reprod Immunol 1998; 39:133. 49. Henzl MR, Smith RE, Boost G, Tyler ET. Lysosomal concept of menstrual bleeding in humans. J Clin Endocrinol Metab 1972; 34:860. 50. Ylikorkala O, Dawood MY. New concepts in dysmenorrhea. Am J Obstet Gynecol 1978; 130:833. 51. Todd AS. Localization of fibrinolytic activity in tissues. Br Med Bull 1964; 20:210. 52. Forrest TS, Elyaderani MK, Muilenburg MI, et al. Cyclic endometrial changes: US assessment with histologic correlation. Radiology 1988; 167:233. 52a.Adamopoulos DA, Kapolla N, Abrahamian A, et al. Sex steroids in cervical mucus of spontaneous or induced ovulatory cycles. Steroids 2000; 65:1. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94.

Moghissi KS, Syner FN, Evans TN. A composite picture of the menstrual cycle. Am J Obstet Gynecol 1972; 114:405. Moghissi KS. Composition and function of cervical secretion. In: Greep RO, ed. Handbook of physiology. Endocrinology, vol II, part 2. Washington, DC: American Physiological Society, 1973:25. Billings EL, Billings JJ, Brown JB, Burger HG. Symptoms and hormonal changes accompanying ovulation. Lancet 1972; 1:282. Rakoff AE. Hormonal cytology in gynecology. Clin Obstet Gynecol 1961; 4:1045. Frost JK. Gynecologic and obstetrics cytopathology. In: Novak ER, Woodruff JD, eds. Novak's gynecologic and obstetrics pathology, 7th ed. Philadelphia: WB Saunders, 1974:634. Collett-Solberg PR, Grumbach MM. A simplified procedure for evaluating estrogenic effects and the sex chromatin patterns in exfoliated cells in urine: studies in premature thelarche and gynecomastia of adolescence. J Pediatr 1965; 66:883. Halasz B. Hypothalamic mechanisms controlling pituitary function. Prog Brain Res 1972; 38:122. Cannel PD, Araki S, Ferin M. Prolonged stalk portal blood collection in rhesus monkeys: pulsatile release of gonadotropin-releasing hormone (GnRH). Endocrinology 1976; 99:243. Wetsel WC, Valenca MM, Merchenthaler I, et al. Intrinsic pulsatile secretory activity of immortalized luteinizing hormone-releasing hormone-secreting neurons. Proc Natl Acad Sci U S A 1992; 89:4149. Pohl CR, Knobil E. The role of the central nervous system in the control of ovarian function in higher primates. Annu Rev Physiol 1982; 44:583. Crowley WF, McArthur JW. Stimulation of the normal menstrual cycle in Kallmann's syndrome by pulsatile administration of luteinizing hormone-releasing hormone (LHRH). J Clin Endocrinol Metab 1980; 51:173. Miller DS, Reid RR, Cetel NS, et al. Pulsatile administration of low-dose gonadotropin-releasing hormone: ovulation and pregnancy in women with hypothalamic amenorrhea. JAMA 1983; 250:2937. Meldrum DR, Chang RJ, Lu J, et al. “Medical oophorectomy” using a long-acting GnRH agonist: a possible new approach to the treatment of endometriosis. J Clin Endocrinol Metab 1982; 54:1081. Yen SSC. Clinical applications of gonadotropin-releasing hormone analogs. Fertil Steril 1983; 39:257. Reissmann T, Felberbaum R, Diedrich K, et al. Development and application of luteinizing hormone-releasing hormone antagonists in the treatment of infertility: an overview. Hum Reprod 1995; 10:1974. Fink G. Neuroendocrine control of gonadotropin secretion. Br Med Bull 1979; 35:155. Gallo RV. Neuroendocrine control of pulsatile luteinizing hormone release in the rat. Neuroendocrinology 1980; 30:122. Barraclough CA, Wise PM. The role of catecholamines in the regulation of pituitary luteinizing hormone and follicle-stimulating hormone secretion. Endocr Rev 1982; 3:91. Yen SSC. Studies of the role of dopamine in the control of prolactin and gonadotropin secretion in humans. In: Fuxe K, Hökfelt T, Luff R, eds. Central regulation of the endocrine system. New York: Plenum Publishing, 1979:387. Ropert JF, Quigley ME, Yen SSC. The dopaminergic inhibition of LH secretion during the menstrual cycle. Life Sci 1984; 34:2067. Eskay RL, Warbert J, Mical RS, Porter JC. Prostaglandin E 2-induced release of LHRH into hypophyseal portal blood. Endocrinology 1975; 97:816. Quigley ME, Yen SSC. The role of endogenous opiates on LH secretion during the menstrual cycle. J Clin Endocrinol Metab 1980;51:179. Ropert JF, Quigley ME, Yen SSC. Endogenous opiates modulate pulsatile luteinizing hormone release in humans. J Clin Endocrinol Metab 1981; 52:583. Stumpf WE, Sar M, Keeper DA. Anatomical distribution of estrogen in the CNS of mouse, rat, tree shrew and squirrel monkey. In: Raspe G, Bernhard A, eds. Central actions of estrogenic hormones. Oxford: Pergamon Press, 1975:77. Sar M, Stumpf WE. Neurons of the hypothalamus concentrate ( 3H) progesterone or its metabolites. Science 1973; 182(Suppl):1266. Reiter EO, Grumbach MM. Neuroendocrine control mechanisms and the onset of puberty. Annu Rev Physiol 1982; 44:595. Yen SSC, Tsai CC. The effect of ovariectomy on gonadotropin release. Clin Invest 1971; 50:1149. Yen SSC, Tsai CC, VandenBerg G, Rebar RW. Gonadotropin dynamics in patients with gonadal dysgenesis: a model for the study of gonadotropin regulation. J Clin Endocrinol Metab 1972; 35:897. Yen SSC, VandenBerg G, Tsai CC, Siler T. Causal relationship between the hormonal variables in the menstrual cycle. In: Fern M, Halberg F, Richart RM, Vande Wiele RL, eds. Biorhythms and human reproduction. New York: John Wiley and Sons, 1974:219. Tsai CC, Yen SSC. Acute effects of intravenous infusion of 17b-estradiol on gonadotropin release in pre- and post-menopausal women. J Clin Endocrinol Metab 1971; 32:766. Tsai CC, Yen SSC. The effect of ethinyl estradiol administration during early follicular phase of the cycle on the gonadotropin levels and ovarian function. J Clin Endocrinol Metab 1971; 33:917. Lasley BL, Wang CF, Yen SSC. The effects of estrogen and progesterone on the functional capacity of the gonadotrophs. J Clin Endocrinol Metab 1975; 41:820. Yen SSC, Lein A. The apparent paradox of the negative and positive feedback control system on gonadotropin secretion. Am J Obstet Gynecol 1976; 126:942. Hoff JD, Lasley BL, Wang CF, Yen SSC. The two pools of pituitary gonadotropin: regulation during the menstrual cycle. J Clin Endocrinol Metab 1977; 44:302. Hoff JD, Lasley BL, Yen SSC. The functional relationship between priming and releasing actions of LRF. J Clin Endocrinol Metab 1979; 49:8. Xia L, Van Vugt D, Alston EJ, et al. A surge of gonadotropin-releasing hormone accompanies the estradiol-induced gonadotropin surge in the rhesus monkey. Endocrinology 1992; 131:2812. Urban RJ, Veldhuis JD, Dufau ML. Estrogen regulates the gonadotropin-releasing hormone–stimulated secretion of biologically active luteinizing hormone. Clin Endocrinol Metab 1991; 72:660. Liu JH, Yen SSC. Induction of midcycle gonadotropin surge by ovarian steroids in women: a critical evaluation. J Clin Endocrinol Metab 1983; 57:797. Kasa-Vuvu JZ, Dahl GE, Evans NP, et al. Progesterone blocks the estradiol-induced gonadotropin discharge in the ewe by inhibiting the surge of gonadotropin-releasing hormone. Endocrinology 1992; 131:208. Araki S, Chikazawa K, Motoyama M, et al. Reduction in pituitary desensitization and prolongation of gonadotropin release by estrogen during continuous administration of gonadotropin-releasing hormone in women: its antagonism by progesterone. J Clin Endocrinol Metab 1985; 60:590. Gougeon A. Regulation of ovarian follicular development in primates: facts and hypothesis. Endocr Rev 1996; 17:121. di Zerega GA, Hodgen GD. Folliculogenesis in the primate ovarian cycle. Endocr Rev 1981; 2:27.

94a.Yap C. Ontogeny: the evolution of an oocyte. Obstet Gynecol Surv 2000; 55:449. 95. Goodman AL, Hodgen GD. Between ovary interaction in the regulation of follicle growth, corpus luteum function, and gonadotropin secretion in the primate ovarian cycle. I. Effects of follicle cautery and hemiovariectomy during the follicular phase in cynomolgus monkeys. Endocrinology 1979; 104:1304. 96. Mizunuma H, Liu X, Andoh K, et al. Activin from secondary follicles causes small preantral follicles to remain dormant at the resting stage. Endocrinology 1999; 140:37. 97. Belaisch-Allart J, Dufetre C, Allan JP, De Mouzon J. Comparison of transvaginal and transabdominal ultrasound for monitoring follicular development in an in-vitro fertilization program. Hum Reprod 1991; 6:688. 98. Baird DT. A model for follicular selection and ovulation: lessons from superovulation. J Steroid Biochem 1987; 27:15. 99. Pappa A, Seferiadis K, Fotsis T, et al. Purification of a candidate gonadotropin surge attenuating factor from human follicular fluid. Hum Reprod 1999; 14:1449. 100. Rabinovici J, Blankstein J, Goldman B, et al. In vitro fertilization and primary embryonic cleavage are possible in 17a-hydroxylase deficiency despite extremely low intrafollicular 17b-estradiol. J Clin Endocrinol Metab 1989; 68:693. 101. Hillier SG. Paracrine control of follicular estrogen synthesis. Semin Reprod Endocrinol 1991; 9:332. 102. Miro F, Hillier SG. Relative effects of activin and inhibin on steroid hormone synthesis in primate granulosa cells. J Clin Endocrinol Metab 1992; 75:1556. 103. Yoshimura Y. Insulin-like growth factors and ovarian physiology. J Obstet Gynaecol Res 1998; 24:305. 104. Pauerstein CJ, Eddy CA, Croxatto HD, et al. Temporal relationships of estrogen, progesterone, and luteinizing hormone levels to ovulation in women and infrahuman primates. Am J Obstet Gynecol 1978; 130:876. 105. Strickland S, Beers WH. Studies on the role of plasminogen activator in ovulation: in vitro response of granulosa cells to gonadotropins, cyclic nucleotides, and prostaglandins. J Biol Chem 1976; 251:5694. 106. Terranova PF, Rice VM. Review: cytokine involvement in ovarian process. Am J Reprod Immunol 1997; 37:50. 107. Eppig JJ. Regulation of cumulus oophorus expansion by gonadotropins in vivo and in vitro. Biol Reprod 1980; 23:545. 108. Wallach EE, Wright KH, Hamada Y. Investigation of mammalian ovulation with an in vitro perfused rabbit ovary preparation. Am J Obstet Gynecol 1978; 132:728. 109. Adashi EY. The potential role of interleukin-1 in the ovulatory process: an evolving hypothesis. Mol Cell Endocrinol 1998; 140:77 110. Frederick J, Shimanuki T, di Zerega GS. Initiation of angiogenesis by human follicular fluid. Science 1984; 224:389. 111. Hsueh AJW, Adashi EY, Jones PBC, Welsh TH Jr. Hormonal regulation of the differentiation of cultured ovarian granulosa cells. Endocr Rev 1984; 5:76. 112. Collins RL, Sopelak VM, Williams RF, Hodgen GD. Pulsatile GnRH treatment in midluteal phase: timely luteolysis despite enhanced steroidogenesis. In: Toft DO, Ryan RJ, eds. Proceedings of the fifth ovarian workshop. Champaign, IL: Ovarian Workshops, 1985:59. 113. Groff TR, Raj HGM, Talbert LM, Willis DL. Effects of neutralization of luteinizing hormone on corpus luteum function and cyclicity in Macaca fascicularis. J Clin Endocrinol Metab 1984; 59:1054. 114. Healy DL, Schenken RS, Lynch A, et al. Pulsatile progesterone secretion: its relevance to clinical evaluation of corpus luteum function. Fertil Steril 1984; 41:114. 115. Filicori M, Butler JT, Crowley WF. Neuroendocrine regulation of the corpus luteum in the human: evidence for pulsatile progesterone secretion. J Clin Invest 1984; 73:1638. 116. Apa R, Miceli F, Pierro E, et al. Paracrine regulation of insulin-like growth factor I (IGF-I) and IGF-II on prostaglandins F2a and E2 synthesis by human corpus luteum in vitro: a possible balance of luteotropic and luteolytic effects. J Clin Endocrinol Metab 1999; 84:2507. 117. Michael AE, Abayasekara DR, Webley GE. Cellular mechanisms of luteolysis. Mol Cell Endocrinol 1994; 99:R1. 118. Devoto L, Vega M, Kohen P, et al. Endocrine and paracrine-autocrine regulation of the human corpus luteum during the mid-luteal phase. J Reprod Fertil Suppl 2000; 55:13.

CHAPTER 96 DISORDERS OF MENSTRUATION, OVULATION, AND SEXUAL RESPONSE Principles and Practice of Endocrinology and Metabolism

CHAPTER 96 DISORDERS OF MENSTRUATION, OVULATION, AND SEXUAL RESPONSE ROBERT W. REBAR Amenorrhea Definition Clinical Evaluation Laboratory Evaluation Hypergonadotropic Amenorrhea (Primary Hypogonadism) Types of Premature Ovarian Failure Chronic Anovulation Disorders of the Luteal Phase Abnormal Genital Bleeding Prepubertal Years Reproductive Years Evaluation of Abnormal Bleeding Management of Dysfunctional Uterine Bleeding Postmenopausal Years Dysmenorrhea Idiopathic or Cyclic Edema Sexual Function and Dysfunction Chapter References

Disorders of menstruation and ovulation are relatively common in women of reproductive age. Possible disorders range from minor to potentially life threatening. To diagnose and treat such menstrual disorders appropriately, an understanding of normal puberty and of the normal menstrual cycle is required (see Chap. 91 and Chap. 95). Although disorders of sexual response also occur with some frequency, they are often overlooked by physicians. Women seeking assistance may have other complaints, and only a sensitive clinician is able to discern the true reason for the visit. However, disorders of sexual response, although not life threatening, may significantly affect the life of the patient and her sexual partner. A knowledge of normal sexual responses and the willingness to discuss such issues openly with patients contribute to successful resolution of the problems.

AMENORRHEA DEFINITION Amenorrhea is generally defined as the absence of menstruation for 3 or more months in women with past menses or a failure to menstruate by girls 16 years of age who have never menstruated. Amenorrhea is merely a sign; it may suggest several disorders involving any of several organ systems. If the genital outflow tract is intact, amenorrhea indicates failure of the hypothalamicpituitarygonadal axis to interact to induce the cyclic changes in the endometrium that normally cause menses. Amenorrhea may be the result of an abnormality at any level of the reproductive tract. Traditionally, amenorrhea is regarded as primary in women who have never menstruated and as secondary in women who have menstruated previously. Because such categorization may lead to diagnostic omission, whether the amenorrhea is primary or secondary should not be a major factor in the evaluation of an amenorrheic woman. Similarly, use of the term “postpill” amenorrhea to refer to women who fail to resume menses within 3 months of discontinuing oral contraceptives conveys nothing about the cause. Women who have fewer than 9 menstrual periods per calendar year should be evaluated identically to those with amenorrhea. CLINICAL EVALUATION Most important to the clinical evaluation are the history and physical examination, with special attention to the possible effects of alterations in hormonal secretion on pubertal development. In general, the clinician should view the patient as a bioassay subject in whom gonadal steroids lead normally to the development of secondary sex characteristics. Breast development indicates exposure to estrogens. The presence of pubic and axillary hair indicates exposure to androgens. Any abnormality of the outflow tract should be eliminated by physical examination. Patients should be questioned regarding the timing of pubertal milestones, and any abnormalities of growth and development should be pursued (see Chap. 91 and Chap. 92). Patients also should be asked about dietary and exercise habits; other aspects of lifestyle, environmental, and psychological stresses; and any family history of amenorrhea or genetic anomalies. It is also important to search for any signs of increased levels of androgen, including acne, hirsutism (i.e., increased sexually stimulated terminal hair; see Chap. 101), and even virilization, such as increased masculine and decreased feminine secondary sexual characteristics, including hirsutism, temporal balding, deepening of the voice, increased muscle mass, clitoromegaly, decreased breast size, and vaginal atrophy. Any history of galactorrhea (i.e., nonpuerperal secretion of milk) should be elicited. Body dimensions and habitus, the distribution and extent of body hair, breast development and secretions, and the external and internal genitalia should be carefully evaluated. Because disorders of sexual development and reproduction frequently are manifested by changes in habitus, it is important to consider the patient's overall appearance. In normal adults, the arm span is similar to the height; in hypogonadal individuals, the arm span typically exceeds the height by 5 cm or more. In congenital hypothyroidism, the extremities are significantly shorter than in normal individuals. The distribution and quantity of body hair should be evaluated, especially with reference to the family history. Hypertrichosis, or the excessive growth of terminal hair on the back and extremities, is almost invariably familial and must be differentiated from true hirsutism. Hypertrichosis is common in women of Mediterranean ancestry, but any facial hair growth in Asian and American Indian women demands evaluation. Although several semiquantitative methods of scoring hirsutism have been developed, it is perhaps most practical to grade facial hirsutism only (because this usually is of most concern to the patient) from 0 to 4+, assigning one point each for excess chin, upper lip, or sideburn hair, and 4+ for a complete beard.1 For documentation, there is no substitute for photographs. Breast development should be staged according to the method of Tanner2 (see Chap. 91). The breasts should be examined for any secretion by applying pressure to all sections of the breast, beginning lateral to the nipple and working toward the nipple while the patient is seated. Secretions should be examined microscopically as a wet mount for the presence of thick-walled, perfectly round fat globules of various sizes, establishing that the discharge is milk (Fig. 96-1).

FIGURE 96-1. Perfectly round, thick-walled fat globules of various sizes are characteristic of galactorrhea when the breast secretion is viewed as a wet preparation under the microscope (original magnification, × 88). For photography, the oil red O stain was added to the specimen, accounting for the dark character of the fat droplets.

The female genitalia are the most sensitive indicators of hormonal status. The Tanner stage of pubic hair development should be recorded.2 The extent of any virilization present indicates the stage in development when exposure to androgens occurred; in general, the sensitivity of the genitalia to androgens decreases with time from the early stages of fetal development to adulthood. The most significant changes, including fusion of the labia and enlargement of the clitoris with or without formation of a penile urethra, are found in women exposed to excess androgens during the first few months of fetal development, as in congenital adrenal hyperplasia (see Chap. 77). The development of significant clitoromegaly in an adult requires marked androgenic stimulation and strongly suggests the presence of an androgen-secreting neoplasm (see Chap. 102). The glans clitoris is definitely enlarged if it is 1 cm or more in diameter. A clitoral index, defined as the product of the sagittal and transverse diameters of the glans at the base, greater than 35 mm2 falls outside the 95% confidence interval.3 Under the influence of estrogen, the labia minora develop at puberty. Examination of the internal genitalia should reveal any overt anomalies of müllerian duct derivatives, including imperforate hymen, vaginal and uterine aplasia, and vaginal septum (see Chap. 90). Obstruction to the escape of menstrual blood can cause hematocolpos (i.e., collection of blood in the vagina) and hematometra (i.e., distention of the uterus with blood). Although a bulging perineum and a pelvic mass are typically found on examination, differentiating vaginal agenesis from a vaginal septum or an imperforate hymen may be difficult. In all of these cases, the normal development of the external genitalia and of other secondary sex characteristics indicates normal ovarian function. The occurrence of intermittent abdominal pain suggests intra-abdominal bleeding. Müllerian dysgenesis (i.e., Rokitansky-Küster-Hauser syndrome) may be accompanied by bony abnormalities of the lumbar spine (e.g., spina bifida occulta), renal anomalies, and disorders of the eighth cranial nerve.4 If there is asynchronous pubertal development with significant breast development in the absence of much pubic and axillary hair, androgen insensitivity (i.e., 46,XY male pseudohermaphroditism) must be excluded. These disorders, including complete testicular feminization, are generally inherited as X-linked recessive or sex-linked autosomal-dominant traits. Complete virilization does not occur despite the presence of testes located inguinally or intraabdominally. Patients have a typical female habitus with normal female external genitalia, but breasts develop only to Tanner stage 3, and the vagina is absent or ends blindly (see Chap. 90). Outflow tract obstruction associated with a normal uterus should be treated surgically to prevent tubal damage from intraabdominal menstruation. Individuals with testicular feminization should be reared as females and treated with an estrogen and a progestin after surgical removal of the testes. The testes should be removed because of the risk of malignancy. Girls lacking a vagina may undergo vaginoplasty (i.e., McIndoe procedure) when regular sexual activity is anticipated.5 In motivated individuals, a vagina can also be created gradually by the daily use of dilators of increasing size (i.e., Frank nonoperative method).6 For individuals with a normal genital tract, visual inspection of the quality of the vaginal mucosa and of the cervical mucus is important because the two are sensitive to estrogen. In response to this hormone, the vaginal mucosa is transformed at puberty from a tissue with a shiny, smooth, bright red appearance to a dull, gray-pink, rugated surface. The cervical mucus increases in quantity and elasticity (i.e., spinnbarkeit) when estrogen is present. Pelvic examination may also reveal pelvic pathologic processes, including neoplasms. The history and physical examination can differentiate several causes of amenorrhea in women of reproductive age, including disorders of sexual differentiation (e.g., distal genital tract obstruction such as müllerian agenesis and dysgenesis, gonadal dysgenesis, ambiguity of external genitalia as in male and female pseudohermaphroditism); other peripheral causes (e.g., pregnancy, gestational trophoblastic disease, amenorrhea traumatica as in Asherman syndrome); and chronic anovulation or ovarian failure (e.g., hypothalamic-pituitary dysfunction, inappropriate feedback because of polycystic ovarian syndrome, adrenal or thyroid dysfunction, abnormal prolactin secretion, premature ovarian failure). Any sexual ambiguity indicates the need for chromosomal karyotyping and the measurement of serum 17-hydroxy-progesterone to rule out 21-hydroxylase deficiency (e.g., congenital adrenal hyperplasia; see Chap. 77). Pregnancy and gestational trophoblastic disease may be confirmed by determining if circulating levels of human chorionic gonadotropin (hCG) are elevated. The existence of intrauterine synechiae or adhesions (i.e., Asherman syndrome) must be suspected in women who develop oligomenorrhea or amenorrhea after curettage or endometritis; tuberculous endometritis may also lead to this disorder.7 The diagnosis can be made by performing hysterosalpingography or hysteroscopy. Hysteroscopic lysis of adhesions is effective in treating Asherman syndrome in more than 80% of affected individuals. Unless serum follicle-stimulating hormone (FSH) levels are measured, it is frequently impossible to differentiate individuals with chronic anovulation, in whom hypothalamic–pituitary–ovarian function is disrupted, from those patients with ovarian failure in whom the ovaries are generally devoid of oocytes. However, it should be possible to form strong clinical impressions about the cause of the amenorrhea. To determine with certainty whether the outflow tract is intact and to evaluate the levels of endogenous estrogen, exogenous progestin, in the form of progesterone in oil (100–200 mg given intramuscularly) or medroxyprogesterone acetate (5–10 mg taken orally each day for 5 to 10 days), can be administered. Any genital bleeding within 10 days of the completion of progestin administration makes the diagnosis of Asherman syndrome unlikely (although still possible) and suggests the presence of chronic anovulation rather than hypothalamic-pituitary or ovarian failure. If the patient does not bleed in response to the progestin, an estrogen and a progestin together (e.g., 2.5 mg of oral conjugated estrogen daily for 25 days with 5 to 10 mg of oral medroxyprogesterone acetate or 200 mg of micronized progesterone also given for the last 10 days) should produce bleeding if the endometrium is normal. Withdrawal bleeding in response to progestin does not exclude the diagnosis of hypergonadotropic amenorrhea, associated with ovarian failure. LABORATORY EVALUATION After appropriate clinical evaluation, measurements of basal serum levels of FSH, prolactin, and thyroid-stimulating hormone (TSH) are indicated in all amenorrheic women to confirm the clinical impression (Fig. 96-2). Whenever the basal prolactin level is elevated (generally >20 ng/mL) on initial testing, the measurement should be repeated, because prolactin levels are increased by a number of nonspecific stimuli, including stress, sleep, and food ingestion. If thyroid function is normal and prolactin levels are elevated, further evaluation is warranted to rule out a pituitary tumor and other causes (see Chap. 13). Basal prolactin concentrations should be determined in all amenorrheic women, not just in those with galactorrhea, because prolactin levels are elevated in more than one-third of all amenorrheic women.8

FIGURE 96-2. Flow diagram for the laboratory evaluation of amenorrhea. Such a scheme must be considered as an adjunct to the clinical evaluation of the patient. (CAH, congenital adrenal hyperplasia; DHEAS, dehydroepiandrosterone sulfate; FSH, follicle-stimulating hormone; HCA, hypothalamic chronic anovulation; PCO, polycystic ovarian syndrome; PRL, prolactin; T, thyroxine; TSH, thyroid-stimulating hormone.) (Reprinted from Rebar RW. The ovaries. In: Smith LH Jr, ed. Cecil textbook of medicine, 18th ed. Philadelphia: WB Saunders, 1992:1367.)

Increased serum TSH levels (generally >5 µU/mL utilizing sensitive assays) with or without increased levels of prolactin indicate primary hypothyroidism (see Chap. 15 and Chap. 45). The increased secretion of thyrotropin-releasing hormone (TRH) in this disorder stimulates increased secretion of prolactin and TSH in some affected women. High serum FSH levels (>30 mIU/mL in most laboratories) imply ovarian failure. Chromosomal evaluation is indicated in all women with increased serum FSH levels who are younger than 30 years of age when the amenorrhea begins, because a number of karyotypic abnormalities have been identified in such women. Gonadectomy is indicated in any such individual who has a portion of a Y chromosome because of the malignant potential of the gonads.9 If prolactin, TSH, and FSH levels are normal or low, further evaluation is based on the clinical presentation. Circulating thyroid hormone levels should be determined if there is any suggestion of thyroid dysfunction. Serum total testosterone levels should be determined whether or not the patient is hirsute; not all hyperandrogenic

women are hirsute because of relative insensitivity of the hair follicles to androgens in some women. Although slightly increased levels of serum testosterone and perhaps of dehydroepiandrosterone sulfate (DHEAS) suggest polycystic ovarian syndrome (PCO), androgen levels occasionally are not elevated in PCO, because of alterations in the metabolic clearance rates of androgens and in sex-hormone-binding-globulin (SHBG) concentrations.10 Circulating levels of luteinizing hormone (LH) may also aid in differentiating PCO from hypothalamic-pituitary dysfunction or failure. LH levels often are increased in PCO such that the ratio of LH to FSH is increased, but this too is not always so.11 However, LH and FSH levels are normal or slightly reduced in women with hypothalamic-pituitary dysfunction.12 There is some overlap between women with PCO-like disorders and those with hypothalamic-pituitary dysfunction. In an effort not to miss a serious cause of amenorrhea, some radiographic assessment of the region of the sella turcica is indicated in all amenorrheic women in whom LH and FSH levels are low (generally 50 pg/mL or if the LH level is significantly greater than the FSH level (in terms of mIU/mL) in any sample, the probability of viable oocytes is considerable. Irregular uterine bleeding, as an indication of estrogen stimulation, also provides good evidence of remaining functional ovarian follicles. It is not uncommon to identify women with intermittent menstruation, hypoestrogenism, and hypergonadotropinism. Because a number of pregnancies have occurred after biopsy of ovaries devoid of oocytes, ovarian biopsy cannot be recommended for affected women. Even in women with intermittent ovarian failure, estrogen replacement is appropriate to prevent the accelerated bone loss that occurs in affected women.42 The estrogen should always be given sequentially with a progestin to prevent endometrial hyperplasia (see Chap. 100). Because women with ovarian failure may conceive while on estrogen therapy (including combined oral contraceptive agents), affected women should be counseled appropriately and cautioned to have a pregnancy test if withdrawal bleeding does not occur or if signs and symptoms develop that are suggestive of pregnancy. Despite these considerations, probably no other contraceptive agent is required for those women who do not wish pregnancy but who are sexually active, because pregnancy occurs in far less than 10%.13 Although rare pregnancies in women with premature ovarian failure have occurred after ovulation induction with human menopausal and chorionic gonadotropins, the low likelihood should lead the physician to discourage patients from selecting such therapy. Hormone replacement treatment to mimic the normal menstrual cycle, with oocyte donation for embryo transfer, may provide the greatest possibility for pregnancy in women desiring pregnancy.43,44 CHRONIC ANOVULATION Chronic anovulation may be viewed as a steady state in which the monthly rhythms associated with ovulation are not functional. Although amenorrhea is common, irregular menses and oligomenorrhea may occur as well. Chronic anovulation further implies that viable oocytes remain in the ovary and that ovulation can be induced with appropriate therapy. Chronic anovulation is the most common endocrine cause of oligomenorrhea or amenorrhea in women of reproductive age (Table 96-3). Appropriate management requires determination of the cause of the anovulation. However, anovulation can be interrupted transiently by nonspecific induction of ovulation in most affected women.

TABLE 96-3. Causes of Chronic Anovulation

CHRONIC ANOVULATION OF CENTRAL ORIGIN Hypothalamic Chronic Anovulation. Hypothalamic chronic anovulation may be defined as anovulation in which dysfunction of hypothalamic signals to the pituitary gland causes failure to ovulate. It remains unclear whether the primary abnormality is always present within the hypothalamus or sometimes occurs as a result of altered inputs to the hypothalamus. The term is used to refer to women who may be affected with suprahypothalamic or hypothalamic chronic anovulation. Although isolated gonadotropin deficiency frequently is caused by hypothalamic dysfunction, it is preferable to consider such individuals separately. However, it may be virtually impossible to differentiate partial forms of isolated gonadotropin deficiency from hypothalamic chronic anovulation. Some reports have documented an increased incidence of amenorrhea in women who exercise strenuously, diet excessively, or are exposed to severe emotional or physical stresses of any kind45,46 and 47 (see Chap. 128). Such amenorrheic persons fall into this group of women considered as having hypothalamic chronic anovulation, which is sometimes called functional amenorrhea. The diagnosis of hypothalamic chronic anovulation is suggested by the abrupt cessation of menses in women younger than 30 years of age who have no clinically evident anatomic abnormalities of the hypothalamic–pituitary–ovarian axis or any other endocrine abnormalities. The term hypothalamic amenorrhea was first proposed by Klinefelter and colleagues in 1943 for anovulation in which hypothalamic dysfunction is thought to interfere with the pituitary secretion of gonadotropin.48 Although hypothalamic chronic anovulation is a common cause of oligomenorrhea and amenorrhea, relatively little is known about its pathophysiologic basis. The diversity of women with hypothalamic chronic anovulation indicates that this is a heterogeneous group of disorders with similar manifestations. Compared with a matched control population, young women with secondary amenorrhea are more likely to be unmarried, to engage in intellectual occupations, to have had stressful life events, to use sedative and hypnotic drugs, to be underweight, and to have a history of previous menstrual irregularities.45 Although it has been suggested that the percentage of body fat controls the maintenance of normal menstrual cycles, it is more likely that diet, exercise, stress, body composition, and other unrecognized nutritional and environmental factors contribute in various proportions to amenorrhea (Fig. 96-3).

FIGURE 96-3. Schematic representation of postulated associations among various forms of hypothalamic chronic anovulation and common linked factors. These disorders appear to be closely interrelated. (Reprinted from Rebar RW. The reproductive age: chronic anovulation. In: Serra BG, ed. The ovary. New York: Raven, 1983:217.)

Hormonally, basal circulating concentrations of pituitary (i.e., LH, FSH, TSH, prolactin, growth hormone), ovarian (i.e., estrogens, androgens), and adrenal hormones (i.e., dehydroepiandrosterone, DHEAS, cortisol) typically are within the normal range for women of reproductive age.49 However, mean serum gonadotropin, gonadal steroid, and DHEAS levels often are slightly decreased, and circulating and urinary cortisol levels are generally increased compared with those in normal women in the early follicular phase of the menstrual cycle.47,50 Despite low levels of circulating estrogen, affected women rarely have symptoms related to estrogen deficiency. Typically, the pulsatile secretion of gonadotropin is diminished, but these individuals respond normally to exogenous gonadotropin-releasing hormone (GnRH; Fig. 96-4).

FIGURE 96-4. Basal concentrations of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) and their pulsatile patterns during the early follicular phase of the normal menstrual cycle are compared with exaggerated patterns in subjects without ovarian feedback (hypogonadal), in patients with pseudocyesis, and in the absence of pulsatile fluctuations observed in various forms of hypothalamic hypogonadotropism. (IGD, isolated gonadotropin deficiency.) (Reprinted from Yen SSC, Lakely BL, Wang CF. The operating characteristics of the hypothalamic-pituitary system during the menstrual cycle and observations of biological action of somatostatin. Recent Prog Horm Res 1975; 31:321.)

ANOREXIA NERVOSA. Anorexia nervosa may represent the severest form of functional hypothalamic chronic anovulation, or it may have one or more distinct pathophysiologic bases. The constellation of amenorrhea often preceding the weight loss, a distorted and bizarre attitude toward eating, food, or weight, extreme inanition, and a disordered body image makes the diagnosis of anorexia nervosa obvious in almost all cases51,52,53 and 54 (see Chap. 128). Demographically, 90% to 95% of anorectic women are white and come from middle- and upper-income families. Peripheral gonadotropin and gonadal steroid levels generally are lower than in the early follicular phase of the menstrual cycle.55 As patients undergo therapy, gain weight, and improve psychologically, sequential studies of the ultradian gonadotropin rhythms show progressive gonadotropin changes paralleling those normally seen during puberty. Initially, there is a nocturnal rise in gonadotropins, followed by an increase in mean basal gonadotropin levels throughout the 24-hour period.56,57 and 58 The responses of severely ill anorectics to GnRH are also similar to those observed in prepubertal children and become adult-like with recovery or with treatment with pulsatile GnRH.59,60 Because the metabolism of estradiol and testosterone is also abnormal, normalizing with weight gain, some of the gonadotropin changes may be secondary to peripheral alterations in steroids.61 Several abnormalities indicate hypothalamic dysfunction, including mild diabetes insipidus and abnormal thermoregulatory responses to heat and cold.54 Affected individuals have altered body images as well.62 Still other central and peripheral abnormalities exist. There is evidence of chemical hypothyroidism, with affected patients having decreased body temperature, bradycardia, low serum triiodothyronine (T3) levels, and increased reverse T3 concentrations.63,64 and 65 Circulating cortisol levels also are elevated, but the circadian cortisol rhythm is normal.66 The increased cortisol seems to be caused by the reduced metabolic clearance of cortisol as a result of the reduced affinity constant for corticosteroid binding globulin (CBG) present in such patients.67 Moreover, like women with endogenous depression, anorectics suppress significantly less after dexamethasone administration than do normal subjects.68 Anorectics also have reduced ACTH responses to exogenous corticotropin-releasing hormone (CRH), suggesting normal negative pituitary feedback by the increased circulating cortisol.69 Although rigorous studies have not been performed of women with bulimia, presumably such individuals have endocrine disturbances similar to those of women with anorexia nervosa. SIMPLE WEIGHT LOSS AND AMENORRHEA. Societal attitudes encourage dieting and pursuit of thinness, particularly in young women. Several reproductive problems, including hypothalamic chronic anovulation, have been associated with simple weight loss. Affected women are distinctly different from anorectics in that they do not fulfill the psychiatric criteria for anorexia.52 The cessation of menses does not occur before significant weight loss in such women, although this sequence is common in anorectics. The few studies that have been conducted in amenorrheic women with simple weight loss suggest that the abnormalities are similar to those observed in anorectics, but are more minor and more easily reversible with weight gain.70 Although it has been suggested that the amenorrhea in these women is secondary to metabolic defects resulting from undernutrition, the possibility of separate central defects has not been excluded.70 The importance of normal body weight to normal reproductive function is evident in studies of a tribe of desert-dwelling huntergatherers in Botswana.71 The weights of the women vary markedly with the season, being greatest in the summer, and the peak incidence of parturition follows exactly 9 months after the attainment of maximal weight. EXERCISE-ASSOCIATED AMENORRHEA. Regular endurance training in women is associated with at least three distinct disorders of reproductive function: delayed menarche, luteal dysfunction, and amenorrhea.72,73 The American College of Sports Medicine has coined the term the “female athletic triad” to describe the three disorders recognized as sometimes occurring together in female athletes: disordered eating, amenorrhea, and osteoporosis.74,74a Activities associated with an increased frequency of reproductive dysfunction include those favoring a slimmer, lower-body-weight physique such as middle and long distance running, ballet dancing, and gymnastics. Swimmers and bicyclists appear to have lower rates of amenorrhea despite comparable training intensities. The cause of these disorders remains to be established and may involve many factors. Dietary changes, the hormonal effects of acute and chronic exercise, alterations in hormone metabolism because of the increased lean to fat ratio, and the psychological and physical “stress” of exercise itself may all contribute and may vary in importance in different individuals (Fig. 96-5; see Chap. 128 and Chap. 132).

FIGURE 96-5. Some factors apparently involved in the pathophysiology of exercise-associated amenorrhea. (Reprinted from Rebar RW. Effect of exercise on reproductive function in females. In: Givens JR, ed. The hypothalamus in health and disease. Chicago: Year Book Medical Publishers, 1984:245.)

Menstrual dysfunction was induced in untrained women who underwent a program of strenuous aerobic exercise (running 4–10 miles per day) combined with caloric restriction.75 The spectrum of abnormalities in these women included luteal phase dysfunction, loss of the midcycle LH surge, prolonged menstrual cycles, altered patterns of gonadotropin secretion, and amenorrhea. Subsequent studies have indicated that luteal phase defects can occur soon after endurance training is begun in the majority of untrained women.76 However, in contrast to these findings, others observed that a progressive exercise program of moderate intensity did not affect the reproductive system of gynecologically mature (mean age, 31.4 years), untrained, eumenorrheic women.77 It was suggested that relatively young gynecologic age or an earlier age of training onset in particular adversely affects menstrual cyclicity. Many amenorrheic athletes welcome the onset of amenorrhea. However, significant osteopenia, usually affecting trabecular bone, has been reported in these women.78,79 and 80 It appears that the loss in bone density secondary to hypoestrogenism nullifies the beneficial effects of weight-bearing exercise in strengthening and remodeling bone.79,81 Such women are at risk for stress fractures, particularly in the weight-bearing lower extremities, and bone density may remain below those of eumenorrheic athletes even after resumption of menses.82 Stress is generally acknowledged to play a role in the cause of this form of amenorrhea, even though it remains difficult to define the term stress. Amenorrheic runners subjectively associate greater stress with running than do runners with regular menses.83 However, no increase in amenorrhea was observed in a competitive group of young classical musicians, who presumably were experiencing similar stress, compared with a group of young ballet dancers, in whom the incidence of amenorrhea was quite high.84 Basal levels of circulating cortisol and urinary free cortisol excretion, which are indicative of increased stress, are increased in eumenorrheic and amenorrheic runners.85 Because levels of CBG, the disappearance rate of cortisol from the circulation, and the response of cortisol to adrenocorticotropin (ACTH) were not altered in the women runners compared with sedentary control subjects, secretion of ACTH and possibly of CRH must be increased in women who run. Abnormalities of the hypothalamic–pituitary–adrenal axis also are indicated by the observations that serum ACTH and cortisol responses to exogenous CRH are blunted, as are the responses to meals.86 The observation that amenorrheic runners also have subtle abnormalities in hypothalamic–pituitary–adrenal function provides support for the concept that exercise-associated amenorrhea is similar to other forms of hypothalamic amenorrhea.87 PSYCHOGENIC HYPOTHALAMIC AMENORRHEA. Amenorrhea may occur in women with a definite history of psychological and socioenvironmental trauma.45,78,79 and 80 The incidence of amenorrhea is quite high among depressed women, and it is difficult to differentiate the effects of lifestyle and nutritional status from variables such as stress. Studies of individuals in whom a definite psychological traumatic event preceded the onset of amenorrhea have revealed low to normal basal levels of serum gonadotropins with normal responses to GnRH, prolonged suppression of gonadotropins in response to estradiol, and failure of a positive feedback response to estradiol.78,79,80,81,82,83,84,85,86,87,88,89,90,91,92,93,94,95,96 and 97 Increased basal levels of cortisol and decreased levels of DHEAS also have been noticed in women with psychogenic amenorrhea compared with eumenorrheic women.46 The mean levels of circulating cortisol are increased in such women largely because of an increase in the amplitude of the pulses of cortisol.98 Moreover, studies of depressed women have revealed abnormal circadian rhythms of cortisol and early “escape” from dexamethasone suppression.99,100 and 101 The mechanism by which emotional states or stressful experiences cause psychogenic amenorrhea is not yet established. Evidence suggests, however, that a cascade of neuroendocrine events that may begin with limbic system responses to psychic stimuli impairs hypothalamic-pituitary activity.102 It has been suggested that increased amounts of hypothalamic b-endorphin is important in inhibiting gonadotropin secretion.102 Psychological studies have found several social and psychological correlates of psychogenic amenorrhea: a history of previous pregnancy losses, including spontaneous abortion103,104 and 105; stressful life events within the 6-month period preceding the amenorrhea106,107 and 108; and poor social support or separation from significant family members during childhood and adolescence.101,108,109 Many women with psychogenic amenorrhea report stressful events associated with psychosexual problems and socioenvironmental stresses during the teenage years.89 Women with psychogenic amenorrhea also tend to have negative attitudes toward sexually related body parts, more partner-related sexual problems, and greater fear of or aversion to menstruation than do eumenorrheic women.107 Distortions of body image and confusion about basic bodily functions, especially sexuality and reproduction, are common.104 DIMINISHED GONADOTROPIN-RELEASING HORMONE AND LUTEINIZING HORMONE SECRETION IN ALL FORMS. The various forms of hypothalamic chronic anovulation associated with altered lifestyles have several features in common. Altered GnRH and LH secretion seems to be the common result from altered hypothalamic input. It remains unclear if these disorders form a single disorder or several closely related disorders. Moreover, similar forms of amenorrhea are sometimes seen in women with severe systemic illnesses or with hypothalamic damage from tumors, infection, irradiation, trauma, or other causes. TREATMENT. The treatment of patients with hypothalamic chronic anovulation is controversial. Psychological therapy and support or a change in lifestyle may cause cyclic ovulation and menses to resume. However, ovulation does not always resume, even after the lifestyle is altered. The treatment of affected women in whom menses do not resume and who do not desire pregnancy is difficult. Most physicians now advocate the use of exogenous sex steroids to prevent osteoporosis. Therapy consisting of oral conjugated estrogens (0.625–1.250 mg), ethinyl estradiol (20 µg), micronized estradiol-17a (1–2 mg), or estrone sulfate (0.625–2.500 mg) or of transdermal estradiol-17a (0.05–0.10 mg) continuously with oral medroxyprogesterone acetate (5–10 mg) or oral micronized progesterone (200 mg) added for 12 to 14 days each month is appropriate. Sexually active women can be treated with oral contraceptive agents. These women appear to be particularly sensitive to the undesired side effects of sex steroid therapy, and close contact with the physician may be required until the appropriate dosage is established. If sex steroid therapy is provided, patients must be informed that the amenorrhea may still be present after therapy is discontinued. Some physicians believe that only periodic observation of affected women is indicated, with barrier methods of contraception recommended for fertility control. Contraception is necessary for sexually active women with hypothalamic chronic anovulation because spontaneous ovulation may resume at any time (before menstrual bleeding) in these mildly affected individuals. Women who refuse sex steroid therapy should be encouraged to have spinal bone density evaluated at intervals to document that bone loss is not accelerated. Adequate calcium ingestion should be encouraged in all affected women. For women who desire pregnancy but who do not ovulate spontaneously, clomiphene citrate (50–100 mg per day for 5 days beginning on the third to fifth day of withdrawal bleeding) can be used. Treatment with human menopausal and chorionic gonadotropins (hMG-hCG) or with pulsatile GnRH may be effective in women who do not ovulate in response to clomiphene. Because the underlying defect in hypothalamic amenorrhea is decreased endogenous GnRH secretion, administration of pulsatile GnRH to induce ovulation can be viewed as physiologic; it offers the additional advantages of decreased need for ultrasonographic and serum estradiol monitoring, a decreased risk of multiple pregnancies, and a virtual absence of ovarian hyperstimulation. A starting intravenous dose of GnRH of 5 µg every 90 minutes is effective.110 After ovulation is detected by urinary LH testing or ultrasound, the corpus luteum can be supported by continuation of pulsatile GnRH or by hCG (1500 IU every 3 days for four doses). Ovulation rates of 90% and conception rates of 30% per ovulatory cycle can be expected.111 Isolated Gonadotropin Deficiency. As originally described in 1944, Kallmann syndrome consisted of the triad of anosmia, hypogonadism, and color blindness in men.112 Women may be affected as well, and other midline defects may be associated.113,114,115 and 116 Because autopsy studies have shown partial or complete agenesis of the olfactory bulb, the term olfactogenital dysplasia has also been used to describe the syndrome.117 Because isolated gonadotropin deficiency may also occur in the absence of anosmia, the syndrome is considered to be quite heterogeneous.

Data indicate that the defect is a failure of GnRH neurons to form completely in the medial olfactory placode of the developing nose or the failure of GnRH neurons to migrate from the olfactory bulb to the medial basal hypothalamus during embryogenesis.118 In some patients, structural defects of the olfactory bulbs can be seen on magnetic resonance imaging.119 It appears likely that this disorder forms a structural continuum with other midline defects, with septooptic dysplasia representing the severest disorder. Clinically, affected individuals typically present with sexual infantilism and an eunuchoidal habitus, but moderate breast development may also occur. Primary amenorrhea is the rule. The ovaries usually are small and appear immature, with follicles rarely developed beyond the primordial stage.120 These immature follicles respond readily to exogenous gonadotropin with ovulation and pregnancy, and exogenous pulsatile GnRH can also be used to induce ovulation.121,122 Replacement therapy with estrogen and progestin should be given to affected women who do not desire pregnancy. Circulating LH and FSH levels generally are quite low. The response to exogenous GnRH is variable, sometimes being diminished and sometimes normal in magnitude, but rarely may be absent.123,124 Although the primary defect in most individuals appears to be hypothalamic, with reduced GnRH synthesis or secretion, a primary pituitary defect may occasionally be present. In addition, partial gonadotropin deficiency may be more frequent than has been appreciated (see Chap. 115). Hyperprolactinemic Chronic Anovulation. Approximately 15% of amenorrheic women have increased circulating concentrations of prolactin, but prolactin levels are increased in more than 75% of patients with galactorrhea and amenorrhea.8 Radiologic evidence of a pituitary tumor is present in ~50% of hyperprolactinemic women, and primary hypothyroidism must always be considered. Individuals with galactorrhea-amenorrhea (i.e., hyperprolactinemic chronic anovulation) frequently complain of symptoms of estrogen deficiency, including hot flushes and dyspareunia. However, estrogen secretion may be essentially normal.125 It is not clear if it is the hyperprolactinemia or the “hypoestrogenism” that causes the accelerated bone loss seen in such individuals.126 Signs of androgen excess are observed in some women with hyperprolactinemia; androgen excess may rarely result in PCO. In hyperprolactinemic women, serum gonadotropin and estradiol levels are low or normal. Most hyperprolactinemic women have disordered reproductive function, and it appears that the effects on gonadotropin secretion are primarily hypothalamic. The mechanism by which hypothalamic GnRH secretion is disrupted is unknown but may involve an inhibitory effect of tuberoinfundibular dopaminergic neurons.125,127 It has been proposed that increased hypothalamic dopamine is present in hyperprolactinemic women with pituitary tumors but is ineffective in reducing prolactin secretion by adenomatous lactotropes. The dopamine can, however, reduce pulsatile LH secretion and produce acyclic gonadotropin secretion through a direct effect on hypothalamic GnRH secretion (see Chap. 13). It has been suggested that mild nocturnal hyperprolactinemia may be present in some women with regular menses and unexplained infertility.128 Galactorrhea in women with unexplained infertility may reflect increased bioavailable prolactin and may be treated appropriately with bromocriptine.129 Bromocriptine or cabergoline therapy may also be indicated in normoprolactinemic women with amenorrhea and increased prolactin responses to provocative stimuli.130 Hypopituitarism. Hypopituitarism may be obvious on cursory inspection or it may be quite subtle (see Chap. 12, Chap. 13, Chap. 14, Chap. 15, Chap. 16, Chap. 17 and Chap. 18). The clinical presentation depends on the age at onset, the cause, and the woman's nutritional status (Fig. 96-6). Loss of axillary and pubic hair and atrophy of the external genitalia should lead the physician to suspect hypopituitarism in a previously menstruating young woman who develops amenorrhea. In such cases, a history of past obstetric hemorrhage suggesting post-partum pituitary necrosis (i.e., Sheehan syndrome) should be sought.131 Failure to develop secondary sexual characteristics or to progress in development once puberty begins must always prompt a workup for hypopituitarism (see Chap. 18).

FIGURE 96-6. Hypopituitarism in a 28-year-old woman with a craniopharyngioma diagnosed at age 16 years. She had received total replacement therapy since the time of diagnosis. Breast development has not advanced beyond Tanner stage 3, little pubic hair is present, and the body habitus is not that of a mature adult. The deep pigmentation of the areolae occurred during therapy several years earlier with fluoxymesterone in an attempt to induce pubic and axillary hair growth.

Individuals with pituitary insufficiency often complain of weakness, easy fatigability, lack of libido, and cold intolerance. Short stature may occur in individuals developing hypopituitarism during childhood. Symptoms of diabetes insipidus may be observed if the posterior pituitary gland is involved. On physical examination, the skin is generally thin, smooth, cool, and pale (i.e., “alabaster skin”) with fine wrinkling about the eyes; the pulse is slow and thready; and the blood pressure is low. An evaluation of thyroid and adrenal function is of paramount importance in these individuals. Thyroid replacement therapy must be instituted and the patient must be euthyroid before adrenal testing is initiated (see Chap. 14, Chap. 15, Chap. 18 and Chap. 74). Serum gonadotropin and gonadal steroid levels typically are low in hypopituitarism. Responses to exogenously administered hypothalamic hormones have failed to localize the cause to the hypothalamus or the pituitary gland in affected patients. Radiographic evaluation of the sella turcica is indicated in any individual with suspected hypopituitarism. The ovaries appear immature and unstimulated, but because oocytes still are present, ovulation can be induced with exogenous gonadotropins when pregnancy is desired. Exogenous pulsatile GnRH may also be used to induce ovulation if the disorder is hypothalamic. Moreover, oocytes may undergo some development, and even ovarian cysts may appear in the absence of significant gonadotropic stimulation (see Chap. 94). When pregnancy is not desired, maintenance therapy with cyclic estrogen and progestin is indicated to prevent signs and symptoms of estrogen deficiency (see Chap. 100). CHRONIC ANOVULATION RESULTING FROM INAPPROPRIATE FEEDBACK IN POLYCYSTIC OVARY SYNDROME Heterogeneous Disorder. In 1935, Stein and Leventhal focused attention on a common disorder in which amenorrhea, hirsutism, and obesity were frequently associated.132 With the development of radioimmunoassays for measuring reproductive hormones, it became clear that women with what is called PCO shared several distinctive biochemical features. Compared with eumenorrheic women in the early follicular phase of the menstrual cycle, affected women typically have elevated serum LH levels and low to normal FSH levels11 (Fig. 96-7). Virtually all serum androgens are moderately increased, and estrone levels are generally greater than estradiol levels133 (Fig. 96-8). Ovarian inhibin physiology is normal.134

FIGURE 96-7. A, Mean (± SE) daily levels of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) in 16 women with polycystic ovary (PCO) syndrome are

compared with values in 16 normally menstruating women. (From Yen et al. J Clin Endocrinol Metab 1970; 30:435.) B, Single measurements of LH and FSH obtained at the time of the first office visit from 37 women with PCO and 34 normal women in the early follicular phase of the menstrual cycle. Means (± SE) are shown as well as individual data points. Although mean LH levels are elevated and mean FSH levels are decreased compared with normal values, there is considerable overlap. Dashed lines indicate the lower limits of detectability of the assays. (From Rebar RW. Semin Reprod Endocrinol 1984; 2:223.) C, Representative pulsatile LH (but not FSH) release in six women with PCO. (From Rebar RW et al. J Clin Invest 1976; 57:1320.)

FIGURE 96-8. Mean (± SE) circulating levels of peptide and steroid hormone levels in women with polycystic ovary (PCO) syndrome compared with women in the early follicular phase of the menstrual cycle (days 2–4). (FSH, follicle-stimulating hormone; LH, luteinizing hormone.) (Reprinted from DeVane GW, Czekala NM, Judd HL, Yen SSC. Circulating gonadotropins, estrogens, and androgens in polycystic ovarian disease. Am J Obstet Gynecol 1975; 121:496.)

Many women with the biochemical features of PCO have small or even morphologically normal ovaries and are not over-weight or hirsute. Not all women with PCO have the characteristic features. Moreover, excess androgen from any source or increased conversion of androgens to estrogens can lead to the constellation of findings observed in PCO.10 Included are such disorders as Cushing syndrome, congenital adrenal hyperplasia, virilizing tumors of ovarian or adrenal origin, hyperthyroidism and hypothyroidism, obesity, and type 2 diabetes.134a In all of these disorders, the ovaries may be morphologically polycystic. Although no clinical and biochemical criteria describe the syndrome strictly, a conference convened by the National Institutes of Health135 developed diagnostic criteria for PCO: 1. Clinical evidence of hyperandrogenism (e.g., hirsutism, acne, androgenetic alopecia) and/or hyperandrogemia (e.g., elevated total or free testosterone). 2. Oligoovulation (i.e., cycle duration >35 days or 1400 of these women had died. The investigators found that the women who had ever taken postmenopausal estrogens had 20% less all-cause mortality compared with women who did not (relative risk [RR] of death, 0.80; 95% CI, 0.70–0.87). The greatest reductions in mortality were seen with current use and with long durations of use: current use for more than 15 years was associated with a 40% reduction in mortality rates. This reduction in mortality rates was not dependent on the dosage of estrogen used: both high (i.e., ³1.25 mg daily) and low (i.e., £0.625 mg daily) doses of oral conjugated equine estrogens (the most common estrogen used) were associated with nearly equal reductions in mortality. Few women took progestins or parenteral estrogens; therefore, the effect of those hormones on mortality cannot be determined from this study. Most of the reduced mortality in estrogen users seen in this study15 was the result of fewer deaths from occlusive arteriosclerotic vascular disease. Estrogen users were also found to have 20% less cancer mortality, which was observed for many malignancies, including breast cancer (RR, 0.81). One possible explanation is that estrogen users may have had greater health awareness and/or increased medical surveillance and consequently had less extensive disease at the time of diagnosis. As expected, estrogen users had excess mortality from endometrial cancer (RR, 3.0). Women who underwent menopause before age 45 years showed the greatest benefit from estrogen use.15 For the group of women whose menopause occurred after age 54 years, estrogen treatment did not reduce mortality rates. Estrogen use also appeared to reduce mortality for women who smoked, who had hypertension, or who had a history of angina or myocardial infarction, approaching that of healthy women who did not use estrogen. This appears to be a very important finding: at one time, hypertension, tobacco use, and coronary disease were thought to be relative contraindications to estrogen replacement. This was based on the increased incidence of stroke and heart attack seen with high-dose oral contraceptives, as well as with high-dose conjugated estrogens prescribed to men as secondary prevention of myocardial infarction.16 Conceivably, the results of this study15 may argue for offering estrogen replacement to nearly all postmenopausal women, particularly those who underwent a relatively early menopause or who have risk factors for CVD. Perhaps the results may further argue for continuous, long-term treatment, because increasing durations of treatment were associated with further reductions in mortality. It should be remembered, however, that this study is an epidemiologic observation of estrogen users and nonusers and is not a clinical trial. Although the investigators controlled for many potential confounding factors, the possibility nevertheless exists that healthier women are more likely to seek and to be prescribed estrogens. Only randomized, placebo-controlled clinical trials are free of this bias (see later in this chapter). Women with preexisting atherosclerosis may benefit from estrogen replacement.17 In one study, estrogen users undergoing coronary catheterization were less likely to have demonstrable disease compared with nonusers (RR, 0.44; 95% CI, 0.29–0.67).18 The investigators performed a retrospective analysis of the all-cause mortality of women who have undergone catheterization during the preceding 10 years. Relatively few women in this study were estrogen users, which was defined as estrogen use at the time of catheterization (5% subjects) or beginning some time thereafter (another 5% subjects). The adjusted 10-year survival of women with severe coronary stenosis who used estrogens was 97%, but it was only 60% for nonusers. For mild to moderate coronary stenosis, 10-year survival was 95% for users and 85% for nonusers. For women with normal coronary arteries, the 10-year survival was 98% for users and 91% for nonusers, a difference that was not statistically significant. These findings suggest that women with severe coronary atherosclerosis may substantially benefit from estrogen use. However, this retrospective study may be biased by the fact that the decreased mortality seen in estrogen users may have been, in part, a self-fulfilling prophecy: estrogen nonusers who lived the longest after catheterization had the greatest opportunity to begin estrogen treatment and therefore became “estrogen users.” Initially, the only attempts to reduce CVD by estrogen treatment had all been performed in men. Early trials in which men were enrolled after a myocardial infarction showed estrogen treatment to reduce serum cholesterol but not the incidence of a second event.19 The Coronary Drug Project, consisting of 1101 survivors of myocardial infarction, was terminated when excess thrombotic events (particularly pulmonary emboli) were seen in the group treated with estrogen; the incidence of CVD was not reduced.16 This experience is similar to that seen in men with prostatic cancer treated with an estrogen, diethylstilbestrol, which appeared to increase CVD, possibly by causing excessive fluid accumulation leading to congestive heart failure or by increasing thromboembolism.20 This adverse action of estrogen in men may have been the consequence of the high estrogenic potency of the doses used and may not reflect the physiologic action of estrogens. Because men and women have an equal incidence of CVD when matched for lipoprotein concentrations,21 the sex difference in CVD may be a consequence of the characteristic sex differences in serum lipoprotein concentrations. Thus, premenopausal women appear to be protected against CVD by their typically lower low-density lipoprotein (LDL) levels and higher high-density lipoprotein (HDL) levels compared with men of the same age. However, coincident with the loss of estrogen, female LDL levels rise at the time of the menopause and eventually exceed those of men.22 It has been suggested that the loss of estrogen at the menopause causes this increase in LDL, because postmenopausal estrogen replacement has been found to lower LDL levels by 15% to 19% by increasing the clearance of LDL from the circulation.23 In contrast, HDL levels in women decline by only 5% at menopause.22 Thus, the HDL-raising effect of oral estrogens (typically 16% to 18%)23 appears to be a pharmacologic action of the high portal estrogen concentrations presented to the liver after intestinal absorption, which stimulates the production of HDL particles.24 Therefore, if endogenous estrogens protect against CVD, an effect on LDL levels that is greater than that on HDL levels is the likely mechanism. In contrast, the lower incidence of CVD among postmenopausal estrogen users may be the result of both increases in HDL levels and decreases in LDL levels. The magnitude of these lipid changes induced by oral estrogen treatment would be expected to lower the incidence of CVD by as much as 40%, using the regression coefficients determined by clinical trials in which cholesterol levels were improved by drug treatment. Reductions of this magnitude have been observed among estrogen users.14,15 In addition, postmenopausal estrogen treatment has also been found to reduce plasma levels of lipoprotein(a), a highly atherogenic particle.25 Estrogens may also protect against CVD independent of their beneficial actions on lipoprotein levels. Estrogens may retard the oxidation of LDL, thereby decreasing its atherogenicity. This was demonstrated in healthy postmenopausal women who were infused with estradiol intravenously; LDL oxidation was significantly delayed.26 Estrogens may also suppress the uptake of LDL by blood vessel walls, thereby impairing the development of endothelial atheroma.27 There is also evidence that estrogens act directly to promote vasodilatation, as demonstrated in estrogen-treated castrated female monkeys.28 This may be mediated indirectly by estrogen-induced alterations in prostacyclin metabolism, increasing the levels of prostacyclin, a vasodilator, and decreasing the levels of thromboxane, a vaso-constrictor.29 The vasodilatory effect of estrogen may be more direct, because estrogen receptors are present throughout the vascular system.30 The binding of estrogen to endothelial estrogen receptors could stimulate the release of nitric oxide, a potent endogenous vasodilator. This was suggested by work in female rabbits, in which endogenous estrogens were found to promote the rates of basal release of nitric oxide.31 Evidence from epidemiologic studies appears to suggest that estrogen use prevents the development of heart disease. However, the results of the first long-term randomized clinical trial of postmenopausal estrogen replacement did not prove this beneficial effect of estrogen.32 The HERS study (Heart and Estrogen/Progestin Replacement Study) enrolled 2763 post-menopausal women with preexisting coronary disease and randomly assigned them to daily treatment with either a placebo or 0.625 mg conjugated equine estrogens and 2.5 mg medroxy-progesterone acetate. They noted a statistically significant 52% increase in myocardial infarction or cardiac death during the first year of treatment. By the third year, the women assigned hormone treatment began to have a decrease in the incidence of CVD. Overall, the incidence of CVD was nearly identical between the two groups when the entire 4.1-year study was considered. The explanation for these unexpected findings is unknown. One possibility is that estrogen treatment has a prothrombotic tendency, which increases the risk of a cardiac event in these high-risk women. Conceivably, after 2 years, the beneficial changes in the lipid profile induced by hormone-replacement therapy (HRT) may predominate over this prothrombotic effect, ultimately lowering the incidence of CVD. An alternative explanation is that the concomitant daily administration of the progestin, medroxyprogesterone acetate, detracts from the cardioprotective action of estrogen by adversely altering lipid levels or vasomotor function. The results of the Women's Health Initiative, which are not expected until 2008, will identify the effects of treatment with estrogen alone (and with progestin) in women not at high risk for CVD. OSTEOPOROSIS Osteoporosis, the reduction in the mass of structural bone per unit volume, is a major affliction of older women. Osteoporosis can cause loss of height and an increased anterior-posterior diameter of the chest. Women may also develop a typical “dowager hump.” It is estimated that 20% of women will experience a hip fracture by the time they reach age 90 years, nearly always because of osteoporosis. Moreover, up to 15% of women die within 3 months because of complications arising from the fracture, including pulmonary edema, myocardial infarction, and pulmonary embolism.33 Spinal compression fractures are also associated with morbidity; this may affect up to 25% of women by the age of 60 years. The number of fractures of the radius also increases in older women, but the consequences are less serious. In all, it is estimated that in the United States, osteoporosis and its complications cost ~$14 billion annually.34

Osteoporosis occurs when the rate of bone resorption (by osteoclasts) exceeds the rate of bone formation (by osteoblasts). Thus, any successful therapy for the condition must either reduce bone resorption or promote bone production. What is the evidence that osteoporosis is essentially a disease of the menopause? There is a dramatic increase in fractures among women after age 40 to 50 years,35 a time when most women are passing through the menopause. Moreover, the bone mass of women tends to fall rapidly after age 50. When a woman's bone density falls below a “fracture threshold,” minor trauma may cause a fracture. The critical event causing this accelerated bone loss is the depletion of estrogens. Indeed, it is now known that postmenopausal women lose less height if they take exogenous estrogens. A double-blind prospective study conducted over a 5-year period demonstrated that mestranol decreased the rate of bone loss compared with a placebo; there was no appreciable increase in bone density. A 5-year follow-up to the same study found that women who stopped taking estrogens began to lose bone mass very rapidly, whereas women who remained on the drug maintained their bone density.36 Most important, however, is not whether estrogens prevent osteoporosis but whether estrogens indeed decrease the risk of fractures. Some studies have clearly shown this beneficial effect. For example, conjugated equine estrogens (0.625 mg) or ethinyl estradiol (20 µg), clearly reduced the risk of fracture in retrospective case-control studies. These studies indicate that to reduce bone fracture incidence by at least 50%, the estrogens must be started within 3 years of the menopause and must be continued for more than 6 years. It is not known how long estrogens must be continued for lasting benefits.37,38 and 39 Because progestins have antiestrogenic properties, it was believed formerly that addition of a progestin to an estrogen in postmenopausal women might negate the beneficial effects of estrogen on bone. However, it appears that combined estrogen and progestin therapy is at least as effective as estrogen alone in reducing bone loss and the risk of fracture in postmenopausal women.40 Both estrogen and progestin receptors and messenger RNA transcripts have been found in human bone cells.41,42 and 43 Estrogen has also been found to stimulate the proliferation and differentiation of cultured osteoblast-like cells derived from an osteogenic sarcoma in rats.44 Thus, estrogens and progestins appear to have both direct and indirect actions on bone metabolism. Osteoporosis may be prevented by therapies other than estrogen replacement. It appears that women who ingest more calcium than others are less at risk for the development of osteoporosis-related fractures. In addition, clinical trials of calcium supplementation combined with vitamin D have found a reduction in the incidence of hip fractures.45 Premenopausal as well as postmenopausal women taking exogenous estrogen require ~1000 mg per day of calcium, whereas postmenopausal and castrate women not on estrogen require 1500 mg per day to be in calcium balance.33 Most women do not reach this quota. Thus, dietary intake may be supplemented by calcium carbonate tablets, 500 mg of elemental calcium twice daily. Importantly, women who exercise moderately are found to have increased bone density. Alendronate is a bisphosphonate that reduces osteoclastic activity, thereby reducing bone resorption. Ten milligrams given daily for 3 years increased the density of the spine by 8.8% and the femoral neck by 5.9%.46 Vertebral fracture rates were reduced by 47%; there were similar decreases in hip and wrist fractures.47 Because alendronate can cause esophagitis,48 the patient should take it with a glass of water and remain upright for 30 minutes. The bisphosphonates may offer promise as (a) an alternative to estrogen in the prevention of osteoporosis; (b) an adjunctive treatment for women who nevertheless demonstrate bone loss while taking estrogen; and (c) treatment for established osteoporosis. Although it might seem reasonable to give all women who pass through menopause some form of estrogen, this must be balanced with the risks of estrogen therapy. It may be prudent to target therapy to women at risk for osteoporosis. In addition to the risk factors listed in Chapter 64, women also at increased risk include those with premature menopause or early surgical castration, individuals with a strong family history of osteoporosis, and those taking corticosteroids. Sedentary women are also at increased risk, as are women who smoke49 or have high caffeine intake.50 Measurement of bone density by bone biopsy is not practical. Of the different radiologic modalities available, quantitative digital radiography is the preferred technique. It is at least as accurate as other radiographic methods, is less costly, and can be performed rapidly with little radiation exposure. Measurement of serum and urine markers for osteoporosis is usually not helpful.51 For example, serum calcium levels are similar in osteoporotic and nonosteoporotic women. The 24-hour urinary excretion of hydroxyproline, which is a breakdown product of bone, is elevated in persons with osteoporosis and tends to decline when treatment is begun. The ratio of urinary calcium to creatinine also tends to decrease with estrogen administration. However, in individual patients, these values vary and are not specific. VASOMOTOR FLUSHES Vasomotor flushes are a common menopausal symptom52 experienced by 75% of postmenopausal women. In ~20%, the flushes will be so severe that the woman will seek medical care. When they occur at night, they may awaken the patient. This may lead to chronic fatigue, poor concentration, emotional lability, and irritability. Many women experience a premonition that they are about to have a flush. This is followed by a reddening of the face and upper body. The apparent vasodilation results in a rise in peripheral temperature and a subjective sensation of warmth; in fact, the temperature at the finger may increase by as much as 2° to 3°C and remain elevated for up to 20 minutes53 and be associated with endocrine-metabolic alterations.54 Because of the heat loss resulting from peripheral vasodilation, core temperature then falls and the woman will feel cold. These sensations of warmth followed by cold can be very disturbing. The pathophysiology of vasomotor flushes is poorly understood. It is known, however, that the source of flushes is the thermoregulatory center in the hypothalamus (Fig. 100-3). Central catecholamines (norepinephrine and dopamine) may alter release of gonadotropin-releasing hormone (GnRH) and, because the GnRH neurons are in close proximity to the temperature control center, a vasomotor flush ensues (Fig. 100-4).55 Thus, it is postulated that a change in neurotransmitters resulting from estrogen withdrawal stimulates GnRH release and alters the temperature control center to produce a flush. However, GnRH itself is not the primary source, because women with a deficiency of this hormone nevertheless can be symptomatic.56 Based on studies with animal models, it has also been suggested that a decrease in gonadal steroids may cause a fall in endogenous opioid activity within the hypothalamus, thus inducing the symptoms of menopause that are similar to those of opiate withdrawal.57

FIGURE 100-3. Serial measurements of finger temperature and serum follicle-stimulating hormone (FSH), luteinizing hormone (LH), estrone (E1), and estradiol (E2) levels in postmenopausal women during hot flushes. Arrows mark the hot flush episodes recorded by increases in finger temperature. (From Meldrum DR, et al. Gonadotropins, estrogens, and adrenal steroids during the menopausal hot flash. J Clin Endocrinol Metab 1980; 50:685.)

FIGURE 100-4. Peripheral levels of luteinizing hormone–releasing hormone (LHRH) and luteinizing hormone (LH) during 3 hours of monitoring in seven women with vasomotor flushes. (From Ravnikar V, Elkind-Hirsch K, Schiff I, et al. Vasomotor flushes and the release of peripheral immunoreactive luteinizing hormone–releasing hormone in postmenopausal women. Fertil Steril 1984; 41:881.)

Estrogen is effective therapy for vasomotor flushes; indeed, it is the standard to which all other therapies must be compared. Many double-blind, prospective, crossover, randomized studies have conclusively found that estrogens are more effective than placebo. Interestingly, the flushes return with increased frequency and severity on crossover from estrogens to placebo.58 This suggests that the probable cause of the flushes is estrogen withdrawal rather than simply estrogen insufficiency; it further suggests that when estrogens are discontinued, they should be discontinued slowly over time. Importantly, when no therapy is used, the vasomotor flushes eventually disappear in most women. Other forms of therapy with some reported effectiveness include medroxyprogesterone acetate at a dose of 10 to 20 mg per day, a-adrenergic agonists such as clonidine (0.1 mg, twice daily), and Bellergal (SANDOZ) (a nonspecific therapy containing ergotamine and belladonna, which are specific inhibitors of the sympathetic and parasympathetic nervous systems, respectively, reinforced by the synergistic action of phenobarbital in dampening cortical brain centers).59 Currently the latter drug is seldom used. INSOMNIA The possible relationship between hormone deprivation and sleep disturbance prompted a number of studies that examined estrogen's effect on insomnia. In one double-blind parallel study using estrone sulfate and placebo, an increase in rapid eye movement (REM) sleep was found but total sleep time was not affected.60 In a second study with a double-blind crossover arrangement using conjugated estrogens (0.625 mg), women taking estrogens experienced fewer hot flushes, a shorter sleep latency period, and more REM sleep than women taking a placebo. The sleep latency decreased most in the patients ranked highest in psychological well-adjustment by the attending physician and nurse.61 These results suggest that women experiencing the greatest number of hot flushes respond to estrogens by a decrease in flushing, with a resulting improvement in their insomnia. Women experiencing insomnia together with severe vasomotor flushes seem to respond best to estrogens. This has been confirmed in a study that found that women are indeed awakened by hot flushes.62 The concept of an association between vasomotor flushes and emotional state is not new. For example, one study that used a graphic scale to measure a subject's emotional state demonstrated that after beginning estrogen therapy, postmenopausal women who had experienced severe vasomotor flushes showed improvement in urinary frequency, vaginal dryness, insomnia, headaches, irritability, and other emotional variables such as decreased memory, anxiety, and worry; alternatively, among women who had not experienced flushes, estrogen therapy was followed by improvement only in memory, anxiety, and worry.63 Thus, the amelioration noted in some psychological complaints may result from a domino-like effect initiated by the reduction in vasomotor flushes. Estrogens may also affect the emotional state by a biochemical effect on the brain. For example, estrogens may inhibit monoamine oxidase, an enzyme found in increased levels in some depressed women.64 The alleged validity of a beneficial effect of estrogen replacement therapy on Alzheimer disease lacks adequate confirmation.64a

ESTROGEN-REPLACEMENT THERAPY COMPLICATIONS Hormone-replacement therapy may produce a number of undesirable side effects and complications. For example, it raises the risk of developing thromboembolic disease nearly three-fold, and increases the risk of gallbladder disease by 38%.32 Complications such as coronary artery disease18 and myocardial infarction have not been shown to occur with increased frequency in postmenopausal women as they have in younger women taking oral contraceptives, probably because the estrogen dose used for postmenopausal women is much lower than the one used in the standard oral contraceptive. However, the major concern of exogenous estrogen use is the possible added risk of developing cancer (see Chap. 222 and Chap. 223). ENDOMETRIAL CANCER In 1975, two published case-control studies showed an increased occurrence of endometrial cancer in postmenopausal women taking estrogens65,66 (see Chap. 223). The occurrence of endometrial cancer appeared to be not only dose related but also duration related. It was possible to identify the estrogen-endometrial cancer relationship because the latency period between initial estrogen ingestion and the onset of the endometrial cancer was as short as 2 years. Some studies have suggested that after the estrogens are discontinued, the added risk disappears within 6 months. Estrogens seem to be associated with low-grade endometrial cancers; that is, although cancer is a frightening disease, perhaps prescribing estrogens does not increase the endometrial cancer mortality rate in women. Successful treatment, however, may require hysterectomy and possibly radiation therapy.67 It remained a paradox for a long time why older women given low-dose estrogens should develop endometrial cancer while younger women who are producing greater endogenous estrogen during pregnancy or taking high-estrogen birth control pills do not develop the disease. The explanation appears to be that the estrogens initially were given to postmenopausal women in an unopposed fashion, that is, without progesterone. This speculation led investigators to add a progestin to the prescribed estrogens given to older women, which lowered the incidence of endometrial cancer and of endometrial hyperplasia (considered a precursor of endometrial cancer). The progestins, used for at least 10 days a month within the estrogen regimen, may act by decreasing the estrogen receptors as well as by converting estradiol to estrone, which is a less potent estrogen.68 BREAST CANCER Breast cancer, which affects >10% of women in the United States, is a much more serious disease than endometrial cancer. Besides being common, breast cancer has a high mortality rate. This disease is frequently disfiguring and emotionally very disturbing. Whether estrogens actually cause breast cancer is presently unknown. Although some studies have found no increased risk of breast cancer.69 Other large studies found excess risk among long-term users.70 This observation was confirmed by the Nurses' Health Study, which analyzed 1935 cases of breast cancer prospectively seen during 725,000 person-years of observation.71 They found that the risk of death due to breast cancer in women who had taken estrogen for five or more years was increased 45%. Other studies are in agreement.71a This conclusion is consistent with available animal data suggesting that breast cancer can be induced with high-dose estrogens (see Chap. 222). Also, estrogens can maintain breast tumor growth in tissue culture. They also found that the addition of a progestin to estrogen treatment did not influence the increased risk of breast cancer seen with long-term estrogen use.71 Interestingly, hormone replacement therapy reduces the sensitivity of mammography.71b THERAPEUTIC ASPECTS Theoretically, the ideal estrogen to administer should be the one the woman's own ovaries produced in the premenopausal years, namely, estradiol. Estradiol taken orally is converted to estrone in the gut and liver. However, estradiol given vaginally, by injection, or transdermally is absorbed rapidly; because it bypasses the liver, it appears in the plasma predominantly as estradiol. Estradiol remains biologically potent because it can suppress gonadotropins when given by any of the above routes.72 The transdermal approach has the advantage of delivering constant physiologic levels of estradiol.73 Because the liver is bypassed, it may be considered for women at risk for phlebitis or hypertension. The most common form of estrogen-replacement therapy uses conjugated equine estrogens prescribed orally. The dose that is effective for osteoporosis and flushes is 0.3 to 0.625 mg daily. Estropipate (piperazine estrone sulfate) (1.25 mg per day) or micronized estradiol (Estrace) (0.5 mg) may also be used. Oral or injectable

estrogens with prolonged half-lives generally should not be used. Transdermal estradiol is applied to the skin twice weekly. It is designed to deliver 0.05 to 0.10 mg per day of estradiol, which achieves a blood level in the range of the normal early follicular phase of the menstrual cycle. The drug avoids the first-pass hepatic metabolism of oral preparations; there is no stimulation of renin substrate, and no increase in sex hormone–binding globulin, corticosteroid-binding globulin, or thyroxine-binding globulin.74 Oral estrogens do increase the levels of these globulins by their effect on the liver, but any long-term adverse reactions of these increases are unknown. On the other hand, with oral estrogen administration the liver produces more HDL-cholesterol24 and clears more LDL-cholesterol23 from the circulation, which is presumably a benefit. The standard regimen adds a progestin75 such as medroxy-progesterone acetate, 5 mg daily, from the 1st to the 13th days of the month to reduce the risk of endometrial cancer. A woman with a uterus will have a 90% chance of experiencing withdrawal bleeding. It has been shown that with continuous use of estrogens and progestins, this annoying side effect can be minimized.76 However, irregular and unpredictable bleeding can occur in the first several months of continuous combined therapy and results in high dropout rates. The long-term safety of this regimen needs to be established. There are reports of endometrial cancer developing years later in women treated in this fashion.77 Endometrial biopsies need not be performed before estrogen therapy is begun unless irregular bleeding has occurred. Biopsies need only be performed during hormone treatment if withdrawal bleeding occurs before day 10 or after day 20 of monthly cyclic progestin therapy78 or after 6 months of continuous progestin therapy. Vaginal probe ultrasonography may reduce the number of biopsies required; endometrial cancer is highly unlikely if endometrial thickness is 70% of these lesions are diagnosed at an advanced stage. Patients with functional tumors commonly present with altered secondary sexual characteristics (e.g., hirsutism or virilization) or reproductive dysfunction; these lesions thus tend to be diagnosed at an earlier stage (Fig. 102-1).

FIGURE 102-1. A 55-year-old woman with a 30-year history of amenorrhea, hirsutism, and clitoromegaly associated with a benign cystic teratoma. Note coarsening of the facial skin, seborrhea, temporal hair recession, and mustache.

HISTORY AND PHYSICAL EXAMINATION The clinical history in children should include growth rate and the time of onset and progression of puberty. In adults, the menstrual pattern, change in libido, body

habitus, growth of hair, and pitch of the voice should be recorded. With tumor growth, abdominal distention occurs, and pressure on the bladder or rectum may cause a sensation of pelvic fullness and discomfort. Abdominal and pelvic pain is associated with torsion, hemorrhage, or rupture of a tumor. A family history of ovarian cancer is the factor most strongly associated with ovarian cancer risk. Studies of inherited ovarian cancer have identified at least three syndromes: the site-specific ovarian cancer syndrome, the breast/ovarian cancer syndrome, and the hereditary nonpolyposis colon cancer/ovarian cancer syndrome.6 Phenotypically, familial ovarian carcinomas are more frequently papillary serous tumors, and evidence exists that women with a genetic predisposition to ovarian cancer may be differentially affected by hormonal levels and metabolism. The occurrence of functioning ovarian tumors is not known to have any particular association with inherited ovarian risk. The general physical examination should include measurement of height and weight, examination of the skin for excess production of sebum, assessment of distribution and quantity of hair, assessment of breast development, inspection of the external genitalia, and palpation of the gonads. Small tumors are difficult to detect on pelvic examination and, therefore, rectovaginal examination may be helpful. Evaluation of the breasts, gastrointestinal tract, and uterus is an important aspect of the examination, because ~10% of ovarian tumors are metastases from tumors in these sites. Ascites is typically associated with intraabdominal metastases. LABORATORY EXAMINATION Laboratory examination may include the measurement of serum levels of ovarian cancer–associated antigens such as CA125. CA125 measurement has also been used as a screening test for ovarian carcinoma, not only for women at high risk but also for the general population of women older than age 50 years. Although the serum levels of CA125 and carcinoembry-onic antigen may be useful in monitoring the course of disease, their lack of specificity limits their diagnostic utility.7 Alpha-fetoprotein, human chorionic gonadotropin (hCG), and human placental lactogen are present in germ-cell tumors, and measurement of their serum levels is valuable in monitoring tumor activity and directing treatment.8,9 The measurement of steroid hormones and their metabolites peripherally, as well as locally (by retrograde catheterization), is particularly valuable in the differential diagnosis of virilizing ovarian versus adrenal neoplasms.10 High serum levels of testosterone and low urine levels of 17-ketosteroids often are associated with ovarian tumors. Adrenal tumors usually produce high levels of dehydroepiandrosterone and its sulfate and low levels of testosterone. Consequently, urinary 17-ketosteroid concentrations are elevated, whereas serum testosterone levels may be only slightly elevated or normal. Vaginal smear cytology, cervical mucus examination, and endometrial biopsy may demonstrate target organ effects, particularly those related to estrogen. Long-standing unopposed elevated estrogen levels cause an abnormal maturation of the vagina, which shows an increased proportion of superficial cells on cytologic examination. Stimulation of the endometrium may lead to various degrees of hyperplasia evident in endometrial biopsy specimens. Pelvic and abdominal study by ultra-sonography, intravenous urography, computed tomography (Fig. 102-2), or magnetic resonance imaging may aid in localizing disease to the ovaries or to the adrenals. Laparotomy is necessary for the diagnosis and appropriate clinical staging and treatment of ovarian neoplasms.

FIGURE 102-2. Ovarian carcinoma demonstrated in computed tomogram of pelvis. Large multiloculated tumor (T) with focal solid areas displaces other pelvic structures.

THERAPY Although therapy is dependent in part on the type of tumor and its stage, the primary treatment of ovarian neoplasms is surgical, ranging from unilateral salpingo-oophorectomy to hysterectomy and bilateral salpingo-oophorectomy in conjunction with cytoreductive surgery. Postoperative adjuvant chemotherapy or radiation is often used.11

COMMON EPITHELIAL TUMORS Epithelial neoplasms account for >60% of all tumors arising in the ovary. These tumors are classified according to the predominant pattern of differentiation as serous (resembling fallopian tube epithelium), mucinous (resembling intestinal or endocervical glands), endometrioid (resembling endometrial glands), clear cell, or Brenner (urothelial). They are subdivided into benign, low malignant potential, and malignant types. Tumors of low malignant potential are distinguished from their benign counterparts by the presence of epithelial stratification, papillary tuft formation, nuclear atypia, and increased mitotic activity, and from the malignant tumors by the absence of stromal invasion. Tumors of low malignant potential have an excellent prognosis (as do stage IA or IB, well-differentiated tumors) with a >90% five-year survival rate; in contrast, advanced (stage III and IV) carcinomas have only a 20% five-year survival rate, even with current chemotherapeutic regimens. Most epithelial tumors are nonfunctional and present as a pelvic mass with vague constitutional symptoms. Endocrinologically active stroma that produces steroid hormones, usually androgens, may be found in approximately one-third of epithelial neoplasms; however, only a few of the patients have symptoms. The surface epithelial tumors most often associated with functional stroma are mucinous or endometrioid in differentiation. Typically, the stroma surrounding the neoplasm is luteinized, with the stromal cells tending to be uniform, small, and polygonal with abundant eosinophilic cytoplasm and round nuclei with prominent nucleoli. These cells have the typical light-microscopic and ultrastructural features of steroid hormone– secreting cells. Functioning stroma is more frequent in pregnancy and in postmenopausal women.12 (See also refs. 12a and 12b.)

SEX CORD STROMAL TUMORS Tumors given this designation are thought to be derived from the specialized ovarian stroma. They are the most common type of hormonally active neoplasm. The components include granulosa, theca, Sertoli, Leydig, and gonadal stromal cells.13,14 (See also ref.14a.) GRANULOSA TUMORS The granulosa tumor is the most common malignant functioning tumor of the ovary. When functional, granulosa tumors almost always are estrinizing, although rare tumors are virilizing. These tumors occur in all age groups but are most frequent in women between 30 and 50 years of age. Approximately 5% of granulosa tumors occur in young girls. The symptoms differ with the age of the patient. Most young girls present with isosexual precocious pseudopuberty, whereas women in their reproductive years usually experience disturbances of the menstrual cycle that range from amenorrhea to menometrorrhagia. Occasionally, breast enlargement occurs. Postmenopausal women typically present with vaginal bleeding secondary to endometrial hyperplasia. The frequency of endometrial hyperplasia and carcinoma ranges from 10% to 20%, depending on the criteria used for the diagnosis of these lesions. Androgenic granulosa cell tumors are associated with recent onset of hirsutism. Granulosa tumors are bilateral in 5% of patients and range in size from microscopic to >30 cm in diameter (median, 9 cm). Small neoplasms are solid, encapsulated by a thin fibrous layer, and located in the ovarian cortex, whereas larger tumors may replace the ovary and become focally cystic. Virilizing tumors tend to be thin-walled, cystic tumors. Blood is often present within the cysts and explains the presentation of hemoperitoneum secondary to rupture that occurs in 10% of cases. Granulosa tumors display many histologic patterns, including microfollicular (the most characteristic, showing many Call-Exner bodies; Fig. 102-3), macrofollicular, trabecular, insular, solid or diffuse, and juvenile. The pathologist must be aware of these patterns so that poorly differentiated carcinomas are not misdiagnosed as diffuse granulosa tumors. This relatively common error accounts for many of the examples of poor-prognosis granulosa tumors reported in the literature. Alpha-inhibin is expressed in granulosa cell tumors and other sex cord stromal tumors, but not in most carcinomas.15 Immunohistochemical staining for this peptide may aid in the

differential diagnosis. The histologic appearance of granulosa cell tumors does not correlate with the prognosis in most studies. The neoplastic cells resemble normal granulosa cells, being small, round to ovoid, and uniform. The nuclei have a characteristic longitudinal groove (a “coffee bean” appearance), and the cytoplasm is pale with poorly defined borders. Occasional cells are luteinized. Reactive thecal cells occur in ~25% of granulosa tumors. Anti-mullerian hormont (AMH) is a useful serum marker for these tumors.15a

FIGURE 102-3. Granulosa tumor, microfollicular pattern.×200

The juvenile granulosa tumor, which accounts for 85% of granulosa cell tumors occurring before puberty, has distinctive features: the cells are larger, more pleomorphic, and often luteinized. Nuclear grooves are rare, and mitoses may be numerous.16 Despite these ominous-appearing features, the prognosis is excellent. Ultrastructural, immunocytochemical, and in vitro incubation studies have confirmed that the granulosa cells within these neoplasms can produce progesterone and testosterone in addition to estrogens, and receptors for these hormones may be detected as well.17,18 These findings are in accord with the occasional virilizing and progestational activity displayed by these tumors. Granulosa tumors are regarded as having low malignant potential, with a 5-year survival rate of >90%. However, these tumors recur in 10% to 33% of patients, and late recurrences (as much as 25 years after primary tumor resection) are characteristic. The prognosis is worse in patients with tumors >5 cm in diameter, bilateral tumors, rupture, and spread beyond the ovary. 19 Unilateral salpingo-oophorectomy is adequate treatment for premenopausal women with stage IA neoplasms, whereas postmenopausal women are treated by total abdominal hysterectomy and bilateral salpingo-oophorectomy. THECOMAS Thecomas comprise 2% to 3% of all ovarian neoplasms and are found primarily in perimenopausal and postmenopausal women. Patients generally present with abnormal uterine bleeding and an abdominal mass. Most functional thecomas are estrinizing, but a few are virilizing.20 Endometrial hyperplasia and carcinoma occur in association with these tumors but not as often as with granulosa tumors. Grossly, these are solid, smooth, white to yellow tumors with occasional cysts and calcified areas. The tumors are bilateral in 5% of patients. Microscopically, thecomas have interlacing whorls of spindle cells, many containing abundant lipid (Fig. 102-4). Fibrocollagenous tissue makes up various proportions of these tumors, and if significant numbers of fibroblasts are present, the tumor may be designated a fibrothecoma.

FIGURE 102-4. Thecoma. Spindle cells are in interlacing fascicles. Clear cytoplasm contains lipid. ×200

These tumors are regarded as benign, and excision results in cure. A few cases of “malignant thecomas” have been said to show both clinical and pathologic features of malignancy, namely, invasion or metastases, atypical nuclei, and an increased mitotic rate. SERTOLI-LEYDIG TUMORS Sertoli-Leydig cell tumors, also known as androblastomas, usually occur in women in the third or fourth decade of life. Less than 50% of the tumors are associated with androgenizing signs, including hirsutism and virilization (temporal balding, deepening of the voice, development of male body configuration, and clitoral enlargement) secondary to secretion of testosterone.20,21 A few patients have estrinizing signs attributable to peripheral aromatization of androgens. The usual presentation is oligomenorrhea or amenorrhea. Defeminization, defined as regression of female secondary sex characteristics (amenorrhea, atrophy of endometrium and vaginal mucosa, decreased breast size), occurs initially and is followed by masculinization. The serum testosterone level is elevated, but because these tumors produce little or no androstenedione and dehydroepiandrosterone, urinary 17-ketosteroids are in the normal range. Masculinizing adrenal tumors produce high levels of androstenedione and dehydroepiandrosterone and small amounts of testosterone. Consequently, the urinary 17-ketosteroids are elevated in patients with masculinizing adrenal tumors. Sertoli-Leydig cell tumors are unilateral in >95% of patients. The cut surface is homogeneous and gray-pink to yellow-orange, with occasional cysts, areas of hemorrhage, or necrosis. Well-differentiated tumors are usually 3 hours requires the use of a backup method for 2 days while continuing pill-taking. The efficacy of minipills is excellent among women older than age 40 and lactating women. In women older than 40 years, decreased fecundity contributes to the efficacy of the minipill, whereas in lactating women, the prolactin-induced suppression of ovulation contributes to its efficacy. Another reason that the minipill is an excellent method of birth control in lactating women is that it does not decrease milk volume and has no negative impact on infant growth or development.36,37 and 38 Side Effects. The most common side effect associated with the minipill is irregular uterine bleeding. Women using the minipill may have irregular bleeding, spotting, or amenorrhea. Other side effects include acne and the development of functional ovarian follicular cysts. PROGESTIN-ONLY IMPLANTABLE CONTRACEPTIVES Norplant, a progestin-only implantable contraceptive, was first introduced in Chile in 1972 and in the United States in 1990. The Norplant system is comprised of six silastic rods, each 34-mm long, filled with levonorgestrel. The semipermeable silastic rods allow for a slow release of levonorgestrel at an initial rate of 80 µg per day (equivalent to the amount in a progestin-only pill). After ~9 to 12 months of use, the rods release levonor-gestrel at ~30 µg per day. They maintain excellent contraceptive efficacy (99.7% per year) and are FDA approved for 5 years of continuous use. In the near future, Implanon, a one-rod/3-year implant system containing 3-keto-desogestrel, and Norplant II, a two-rod/3-year implant system containing levonorgestrel, will be released. Mechanism of Action. The principal mechanism by which progestin-only implants exert their contraceptive efficacy is by altering the cervical mucus and making it impenetrable to sperm. The continuous low levels of progestin also serve to suppress LH and prevent ovulation. Compared to cervical mucus changes, the prevention of ovulation is less reliable, and, as progestin levels decline over time, more ovulatory cycles are noted. With Norplant, in the first 2 years of use only 10% to 20% of cycles appear to be ovulatory, whereas in the fifth year of use, ~45% of cycles appear to be ovulatory. However, even cycles that appear to be ovulatory are often not completely normal cycles. Women who have regular menstrual cycles on Norplant have been shown to have subnormal levels of LH, FSH, and progesterone.39,40 Progestin-only implants also induce changes in the endometrial lining. These changes are likely responsible for the irregular uterine bleeding associated with implant use. With Norplant, the endometrial lining has been found to be hypotrophic with an increased microvascular density of capillaries that appear to be particularly fragile.41 The exact mechanism of these changes is unclear. Study has shown that postfertilization prevention of implantation on an unfavorable endometrial lining is not a mechanism for the contraceptive action of Norplant.42 Use of Implants. Contraceptive implants are inserted sub-dermally, in the upper inner arm, under local anesthesia using the provided trocar. Insertion is a simple procedure that usually takes 5 to 10 minutes. Special care must be taken to place the implants in the correct plane. Placement of all six implants in the same subdermal plane allows easy removal. Because the silastic rods are not biodegradable, after 5 years of use, or at the woman's request, the implants must be removed. Under local anesthesia, a small incision is made, and the implants can be removed either with finger pressure or with a pair of small hemostatic clamps. Removal of the implants can take from 5 to 60 minutes. The major advantage of one- and two-rod implant systems is faster and easier insertion and removal. Implant systems using biodegradable capsules are in development. Absolute contraindications to progestin-only implants include undiagnosed vaginal bleeding, suspected or confirmed pregnancy, active liver disease or tumors, active thromboembolic disorders, and known or suspected breast cancer. Relative contraindications include severe acne, depression, vascular migraine headaches, and the concomitant use of medications that increase the hepatic metabolism of progestins—and, hence, decrease the efficacy of implants—such as phenytoin, carbamazepine, phenobarbital, and rifampin. Progestin-only implants may be very well suited for women with hypertension, diabetes, a history of cardiovascular disease (such as stroke, myocardial infarction, or prior deep venous thrombosis), gallbladder disease, hypercholesterolemia, or hypertriglyceridemia. Implants are also appropriate contraception for heavy smokers, including women older than age 35 years. Side Effects. The most common side effect of progestin-only implants, and the most commonly cited reason for removal of Norplant, is irregular menstrual bleeding. In the first year of use, 68% of women using Norplant report menstrual problems.43 Of the women with these problems, 23% report increased bleeding, 16% report decreased bleeding or amenorrhea, and 29% report irregular bleeding or spotting.43 Over time, bleeding patterns tend to improve, and many women report regular menstrual cycles. The Norplant 5-year cumulative removal rate for bleeding problems is 17.5%.43 The cause of the irregular menstrual bleeding is not entirely clear. Under the influence of progestin-only contraception, the endometrial lining becomes hypotrophic, with an increased microvascular density of fragile capillaries.41 These fragile capillaries may be especially prone to bleeding. Varying levels of estrogen, produced by partially stimulated follicles, may also contribute to the irregular bleeding associated with progestin-only implants.39 Other side effects occurring in 10% to 16% of Norplant users in the first year of use include headache, acne, weight gain, leukorrhea, pelvic pain, vaginal fungal infections, genital pruritus, and reaction at the implantation site.43 The incidence of these side effects, with the exception of genital pruritus, decreases in later years. Of these side effects, only weight gain, mood changes, and headache lead to removal rates of >1%.43 After 5 years of use, 59% of U.S. women with implants gained weight. The mean 5-year weight gain for these women was 5.2 kg.43 Another complication of implant use is difficulty in removing the rods. In a large U.S. study, 8% of Norplant removals were classified as difficult, and 3% were associated with adverse effects (including multiple incisions and a reaction to the local anesthetic).43 The amount of difficulty encountered and the time required for removal are related to the provider's skill and experience. New systems with fewer implant rods should minimize difficulty with insertion and removal. No adverse effects on fertility are seen after removal of the rods. Within 24 hours after removal, fertility returns to baseline.40 No consistent effects on blood pressure, lipoproteins, or coagulation have been noted. Also, because hypoestrogenemia does not occur, bone density is not affected. PROGESTIN-ONLY INJECTABLE CONTRACEPTIVES DMPA (Depo-Provera) is an injectable progestin-only contraceptive. It was first introduced in the mid-1960s and, since then, has been used extensively around the world. In 1992 it was approved by the FDA for use in the United States. It is an easy-to-use, private, and very effective (99.7%) method of birth control. DMPA is given as an intramuscular injection every 3 months. Peak serum hormone levels occur shortly after injection and then progressively fall, yet remain in the effective contraceptive range, over the next 3 months. Mechanism of Action. As do other progestin-only contraceptives, DMPA exerts its contraceptive effects by thickening the cervical mucus, making it impenetrable to sperm. However, compared with Norplant, which provides continuous low levels of levonorgestrel, DMPA provides much higher levels of progestin. These high levels of progestin are effective at inhibiting the LH surge, and as a result DMPA is effective at inhibiting ovulation. Despite the fact that DMPA is more effective at inhibiting ovulation than Norplant, both have similar contraceptive efficacy. Although DMPA effectively inhibits the LH surge, it does not completely suppress FSH and, thus, stimulated follicles continue to make estrogen. Estrogen levels in DMPA users have been found to be approximately at the level found in the early follicular phase in a normal menstrual cycle.44 Use of Depot Medroxyprogesterone Acetate. DMPA is an aqueous solution of suspended crystals, given in a dose of 150 mg intramuscularly, either in the deltoid or gluteus maximus muscle. The first injection should be given within 5 days of the menstrual period to ensure the patient is not pregnant at that time. The next injection should be given 12 to 13 weeks after the first injection. If these guidelines are followed, ovulation is inhibited from the onset of use. If the patient is not within 5 days of the onset of her menstrual period, or is beyond 13 weeks after her last injection, a pregnancy test is indicated. If a sensitive pregnancy test is negative and no episodes of unprotected intercourse have occurred in the prior 2 weeks, the injection can be given. If any doubt exists as to whether unprotected intercourse has occurred within the 2 weeks before the negative pregnancy test, a backup method of birth control should be used, and a second pregnancy test given 2 weeks later. If the sensitive pregnancy test is still negative, the injection may be given. In these situations, a backup method of birth control

should be used for the first 2 weeks of DMPA use, as ovulation may not be inhibited during this time. DMPA is contraindicated in women who are pregnant and those who have undiagnosed vaginal bleeding. Relative contraindications include active liver disease, breast cancer, severe depression, severe cardiovascular disease, and a desire to conceive within 1 year. SIDE EFFECTS. The most common side effect associated with DMPA use is irregular menstrual bleeding.44,45 and 46 Within the first year of use, ~70% of women using DMPA have irregular bleeding.44 With continued use, a majority of women become amenorrheic. After 5 years of use, ~80% of women are amenorrheic.47 The irregular bleeding associated with DMPA use is usually not excessive in quantity, but can be of increased frequency or duration. Although average hemoglobin levels rise in women using DMPA, frequent and prolonged bleeding are the most common reasons for discontinuation of DMPA.47 As many as 25% of new DMPA users discontinue this contraceptive in the first year of use due to frequent or prolonged bleeding.48 Other side effects reported with DMPA use include weight gain, depression, decreased libido, headaches, dizziness, abdominal pain, anxiety, and delayed return to fertility. Systemic levels of the drug may persist for 9 months after injection, and long infertile periods of up to 18 months may occur.49 The increased risk of breast cancer found in beagle dogs treated with DMPA has not been observed in humans using DMPA. A multinational study of breast cancer risk in women using DMPA has shown no increased risk50 (see Chap. 105). In cross-sectional studies, DMPA use has been associated with decreased bone mineral density.51,52 and 53 This loss in bone mineral density has been shown to be reversible with discontinuation of DMPA.54 The long-term effect of a temporary loss of bone mineral density on osteoporosis and fractures later in life is unknown (see Chap. 105). NONCONTRACEPTIVE BENEFITS OF DEPOT MEDROXYPROGES-TERONE ACETATE. DMPA has been shown to have many non-contraceptive benefits (Table 104-6). Although DMPA is FDA approved only for use as a contraceptive and in the treatment of metastatic endometrial cancer, it has been shown to raise the seizure threshold and to improve seizure control in some women with seizure disorders, to decrease the incidence of sickling events in women with sickle cell disease, and to decrease the incidence of endometrial cancer. DMPA is also safe for use in lactating women. In contrast to COCs, DMPA causes an increased volume of breast milk in lactating women.

TABLE 104-6. Known and Potential Noncontraceptive Benefits of Depot Medroxyprogesterone Acetate

POSTCOITAL “EMERGENCY” CONTRACEPTIVES Postcoital contraception is an “emergency” aid that can be provided for women who have experienced a single unprotected or inadequately protected act of intercourse within the previous 72 hours. The mechanism of action of postcoital contraception is unclear. Studies have shown that emergency contraceptive pills (ECs) both alter the endometrium and delay ovulation.55,56,57 and 58 Postcoital contraception also may prevent fertilization. Various regimens containing either an estrogen-progestin combination (combined ECs) or a progestin alone (progestin ECs) have been used with considerable success (Table 104-7). The first dose of a combined EC regimen should be given within 72 hours of unprotected intercourse, and the second dose should be given 12 hours later. The reported failure (pregnancy) rate is ~2%. Women having unprotected intercourse during the second or third week of their menstrual cycle have an 8% chance of conceiving that cycle. Thus, postcoital contraception may decrease the risk of conception from 8% to 2%, a 75% decrease in risk.59,60

TABLE 104-7. Oral Contraceptives Used for Emergency Contraception*

The principal side effects of combined ECs are nausea and vomiting. The prophylactic use of an antiemetic 1 hour before each dose can significantly reduce these symptoms.60 Progestin ECs cause less nausea and vomiting. The only absolute contraindication to the use of postcoital contraception is confirmed or suspected ongoing pregnancy. Postcoital contraception with combined or progestin ECs does not terminate an ongoing pregnancy. No data are available on the safety of combined ECs in women with contraindications to estrogen. The short duration of use make significant complications unlikely; however, the progestin-only regimen has similar efficacy and should be considered for these women. In a woman planning to use an IUD for contraception, a copper IUD inserted within 5 days of unprotected intercourse is also a very effective method of emergency contraception (failure rates of 2000 years. In addition to providing contraception, most barrier methods offer some protection against STIs. CONDOM The use of a male or female condom, if made of latex, protects against the transmission of bacteria and viruses, including human immunodeficiency virus (HIV).74 Between 1982 and 1995, condom use rose from 12% to 20% in the United States.1 This rise likely reflects efforts to prevent STIs, specifically HIV. Many couples use a condom both for contraception and to prevent STIs. Others use condoms for disease prevention in addition to another, more effective contraception method. DIAPHRAGM The diaphragm is a round latex barrier that is placed in the vagina before intercourse. The spring-like edges of the diaphragm allow it to collapse to enable placement in the vagina. Once it is properly positioned in front of the cervix, the spring opens and keeps the diaphragm in place. The three basic types of diaphragms are the arcing spring, the coil spring, and the flat spring. Diaphragms come in sizes between 50 and 105 mm in diameter and must be individually fitted for each woman. They must be refitted after childbirth. Various clinical studies indicate typical use effectiveness rates ranging from 2 to 25 pregnancies per 100 woman-years. This broad range of contraceptive effectiveness is attributable to differences in the degree of motivation of the woman and to experience with the method. The principal contraindications to diaphragm use are anatomic factors causing poor fit and allergies to latex or spermicide. The use of a diaphragm plus spermicide provides prophylaxis against many STIs and has been associated with a decrease in cervical dysplasia (likely due to decreased spread of human papilloma virus). The principal complication of diaphragm use is recurrent cystitis, probably due to partial urethral blockage. If, despite adequate antibiotic treatment, the frequency of cystitis increases, another method of contraception should be considered. A serious cause of concern among diaphragm users is the reporting of several nonfatal cases of toxic shock syndrome. In all of these cases, however, the patients had left the diaphragm in place for long periods of time. Women should be carefully instructed never to leave the diaphragm in the vagina longer than 24 hours. SPERMICIDAL PREPARATIONS A great variety of spermicidal preparations are available as foams, creams, jellies, films, or suppositories. All spermicides contain a surfactant, which is responsible for the contraceptive effect. Surfactants have long-chain alkyl groups that easily penetrate the lipoprotein membrane of spermatozoa, increasing the permeability of the cell and leading to irreversible loss of motility. The vagina absorbs some of the spermicidal chemicals. No human studies have reported deleterious effects resulting from the absorption of surfactants. A double-blind, placebo-controlled trial has colposcopically evaluated the local effects of the spermicide nonoxynol-9 on the vagina and cervix and found no increase in epithelial disruption.75 Two studies have suggested a greater risk of congenital birth defects in the offspring of women using vaginal spermicides,76,77 but other studies have not shown any association between spermicide exposure and congenital malformations. Not only do spermicidal preparations provide a contraceptive benefit, but the incidence of cervical gonorrhea, vaginal candidiasis, trichomoniasis, and genital infection with herpes simplex virus are all decreased by these chemical agents. Clinical trials to assess the efficacy of nonoxynol-9 in preventing HIV transmission are in progress.75

FEMALE STERILIZATION Sterilization has become the most common method of fertility regulation in the United States. It is an elective procedure that offers women permanent, nonreversible contraception. Female sterilization is performed by ligating, excising, cauterizing, banding, or clipping portions of both fallopian tubes. The majority of these procedures are performed either laparoscopically or via laparotomy. Failure rates for the most commonly used techniques of tubal sterilization are listed in Table 104-8.78 The larger the amount of tube destroyed, the poorer the potential for a surgical reversal, should the woman desire to have this latter procedure later.

TABLE 104-8. Failure Rates of Different Techniques of Tubal Sterilization

Although pregnancy after tubal sterilization is not common, when pregnancies do occur after sterilization, they are almost as likely to be ectopic as intrauterine pregnancies. The U.S. Collaborative Review of Sterilization found a 10-year cumulative probability of pregnancy of 18.5 per 1000 procedures (for all types of sterilization procedures combined). They found a 10-year cumulative probability of ectopic pregnancy of 7.3 per 1000 procedures. Women at highest risk for ectopic pregnancy after tubal sterilization are those sterilized under age 30 using the bipolar electrocautery technique.79 In the United States, deaths attributable to female sterilization are rare, with a case fatality rate of ~1.5 per 100,000 procedures.80 This is markedly lower than the maternal mortality rate associated with childbearing, which is ~10 per 100,000 live births. Deaths associated with female sterilization have resulted from complications of general anesthesia as well as from infection and hemorrhage. Although major complications are infrequent, one study found that 1.7% of laparoscopic sterilizations are complicated by penetrating injuries, injuries to major abdominal and pelvic vessels, and bowel burns.81 Complication rates are probably lower among clinicians with more laparoscopy experience. CHAPTER REFERENCES 1. Piccinino LJ, Mosher WD. Trends in contraceptive use in the United States: 1982–1995. Fam Plann Perspect 1998; 30:4. 1a. Westhoff C, Davis A. Tubal sterilization: focus on the U.S. experience. Fertil Steril 2000; 73:913. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53.

Trussell J, Kowal D. The essentials of contraception. In: Hatcher RA, et al. Contraceptive technology, 17th ed. New York: Ardent Media, 1998:216. Collins DC. Sex hormone receptor binding, progestin selectivity, and the new oral contraceptives. Am J Obstet Gynecol 1994; 170:1508. Goldzieher JW. Selected aspects of the pharmacokinetics and metabolism of ethinyl estrogens and their clinical implications. Am J Obstet Gynecol 1990; 163:318. Mishell DR, Darney PD, Burkman RT, Sulak PJ. Practice guidelines for OC selection: update. Dialogues Contracept 1997; 5(4):7. Bagwell MA, Coker AL, Thompson SJ, et al. Primary infertility and oral contraceptive steroid use. Fertil Steril 1995; 63:1161. Carpenter S, Neinstein LS. Weight gain in adolescent and young adult oral contraceptive users. J Adolesc Health Care 1986; 7:342. Gerstman BB, Piper JM, Tomita DK, et al. Oral contraceptive estrogen dose and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991; 133:32. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Ischaemic stroke and combined oral contraceptives: results of an international, multicentre, case-control study. Lancet 1996; 348:498. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Venous thromboembolic disease and combined oral contraceptives: results of international multicentre case-control study. Lancet 1995; 346:1575. Pettiti DB, Sidney S, Bernstein A, et al. Stroke in users of low-dose oral contraceptives. N Engl J Med 1996; 335:8. Sidney S, Pettiti DM, Quesenberry CP Jr, et al. Myocardial infarction in users of low-dose oral contraceptives. Obstet Gynecol 1996; 88:939. Beral V, Hermon C, Kay C, et al. Mortality associated with oral contraceptive use: 25 year follow up of cohort of 46,000 women from Royal College of General Practitioners' oral contraception study. BMJ 1999; 318:96. Croft P, Hannaford PC. Risk factors for acute myocardial infarction in women: evidence from the Royal College of General Practitioners' oral contraception study. BMJ 1989; 298:165. Vandenbroucke JP, Koster T, Briet E, et al. Increased risk of venous thrombosis in oral-contraceptive users who are carriers of factor V Leiden mutation. Lancet 1994; 344:1453. Schwingl PJ, Ory HW, Visness CM. Estimates of the risk of cardiovascular death attributable to low-dose oral contraceptives in the United States. Am J Obstet Gynecol 1999; 180:241. Burkman RT Jr. Benefits and risks of oral contraceptives: a reassessment. J Reprod Med 1991; 36(Suppl):217. Larsson G, Milsom I, Lindstedt G, Rybo G. The influence of a low-dose combined oral contraceptive on menstrual blood loss and iron status. Contraception 1992; 46:327. Wolner-Hanssen P, Eschenbach DA, Paavonen J, Kiviat N, et al. Decreased risk of symptomatic chlamydial pelvic inflammatory disease associated with oral contraceptive use. JAMA 1990; 263:54. Barbone F, Austin H, Louv WC, Alexander WJ. A follow-up study of methods of contraception, sexual activity, and rates of trichomoniasis, candidiasis, and bacterial vaginosis. Am J Obstet Gynecol 1990; 163:510. Brinton LA, Vessey MP, Flavel R, et al. Risk factors for benign breast disease. Am J Epidemiol 1981; 113:203. Schlesselman JJ. Risk of endometrial cancer in relation to use of combined oral contraceptives. A practitioner's guide to meta-analysis. Hum Reprod 1997; 12:1851. Sherman ME, Sturgeon S, Brinton LA, et al. Risk factors and hormone levels in patients with serous and endometriod uterine carcinomas. Mod Pathol 1997; 10:963. Mol BW, Ankum WM, Bossuyt PM, Van der Veen F. Contraception and the risk of ectopic pregnancy: a meta-analysis. Contraception 1995; 52:337. Hankinson SE, Colditz GA, Hunter DJ, et al. A quantitative assessment of oral contraceptive use and ovarian cancer. Obstet Gynecol 1992; 80:708. Lanes SF, Birmann B, Walker AM, Singer S. Oral contraceptive type and functional ovarian cysts. Am J Obstet Gynecol 1992; 166:956. Grimes DA, Godwin AJ, Rubin A, et al. Ovulation and follicular development associated with three low-dose oral contraceptives: a randomized controlled trial. Obstet Gynecol 1994; 83:29. Redmond GP, Olson WH, Lippman JS, et al. Norgestimate and ethinyl estradiol in the treatment of acne vulgaris: a randomized, placebo-controlled trial. Obstet Gynecol 1997; 89:615. Kleerekoper M, Brienza RS, Schultz LR, Johnson CC. Oral contraceptive use may protect against low bone mass. Arch Intern Med 1991; 151:1971. Ness RB, Keder LM, Soper DE, et al. Oral contraception and the recognition of endometritis. Am J Obstet Gynecol 1997; 176:580. Cramer DW, Goldman M, Schiff I, et al. The relationship of tubal infertility to barrier method and oral contraceptive use. JAMA 1987; 257:2446. Speroff L, Glass RH, Kase NG. Clinical gynecologic endocrinology and infertility, 5th ed. Baltimore: Williams & Wilkins, 1994:60. Speroff L, Darney PD. Oral contraception. In: A clinical guide for contraception, 2nd ed. Baltimore: Williams & Wilkins, 1996. Darney PD. The androgenicity of progestins. Am J Med 1995; 98(Suppl 1A):104. World Health Organization Committee on Contraceptive Research for Human Reproduction. A multicentered phase III comparative clinical trial of Mesigna, Cyclofem and Injectable No. 1 given monthly by intramuscular injection to Chinese women. Contraception 1995; 51:167. World Health Organization, Special Programme of Research, Development, and Research Training in Human Reproduction, Task Force on Oral Contraceptives. Effects of hormonal contraceptives on milk volume and infant growth. Contraception 1984; 30:505. World Health Organization, Special Programme of Research, Development, and Research Training in Human Reproduction, Task Force for Epidemiological Research on Reproductive Health. Progestogen-only contraceptives during lactation. I. Infant growth. Contraception 1994; 50:35. World Health Organization, Special Programme of Research, Development, and Research Training in Human Reproduction, Task Force for Epidemiological Research on Reproductive Health. Progestogen-only contraceptives during lactation. II. Infant development. Contraception 1994; 50:55. Faundes A, Brache V, Tejada AS, et al. Ovulatory dysfunction during continuous administration of low-dose levonorgestrel by subdermal implants. Fertil Steril 1991; 56:27. Walker DM, Darney PD. Implantable contraception. In: Sciarra JJ, ed. Gynecology and obstetrics. Philadelphia: Lippincott–Raven Publishers, 1998. Fraser IS, Hickey M, Song J. A comparison of mechanisms underlying disturbances of bleeding caused by spontaneous dysfunctional uterine bleeding or hormonal contraception. Hum Reprod 1996; 11(Suppl 2):165. Segal SJ, Alvarez-Sanchez F, Brache V, et al. Norplant implants: the mechanism of contraceptive action. Fertil Steril 1991; 56:273. Sivin I, Mishell DR, Darney PD, et al. Levonorgestrel capsule implants in the United States: a 5-year study. Obstet Gynecol 1998; 92:337. Speroff L, Darney PD. Injectable contraception. In: A clinical guide for contraception, 2nd ed. Baltimore: Williams & Wilkins, 1996. Darney PD, Klaisle CM. Contraception-associated menstrual problems: etiology and management. Dialogues Contracept 1998; 5(5):1. Grimes DA, Wallach M. Injectable contraception. In: Modern contraception: updates from the contraception report. Totowa, NJ: Emron, 1997. Gardener JM, Mishell DR Jr. Analysis of bleeding patterns and resumption of fertility following discontinuation of a long-acting injectable contraceptive. Fertil Steril 1970; 21:286. World Health Organization. Clinical evaluation of the therapeutic effectiveness of ethinyl oestradiol and oestrone sulphate on prolonged bleeding in women using depot medroxyprogesterone acetate for contraception. Hum Reprod 1996; 11(Suppl 2):1. Kaunitz AM. Injectable depot medroxyprogesterone acetate contraception: an update for clinicians. Int J Fertil 1998; 43(2):73. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Breast cancer and depot–medroxyprogesterone acetate: a multinational study. Lancet 1991; 338:833. Cundy T, Evans M, Roberts H, et al. Bone density in women receiving depot medroxyprogesterone acetate for contraception. BMJ 1991; 303:13. Cromer BA, Blair JM, Mahan JD, et al. A prospective comparison of bone density in adolescent girls receiving depot medroxyprogesterone acetate (Depo-Provera), levonorgestrel (Norplant) or oral contraceptives. J Pediatr 1996; 129:671. Scholes D, Lacroix AZ, Ott SM, et al. Bone mineral density in women using depot–medroxyprogesterone acetate for contraception. Obstet Gynecol 1999; 93:233.

54. 55. 56. 57. 58. 59. 60. 61. 62.

Cundy T, Cornish J, Evans MC, et al. Recovery of bone density in women who stop using medroxyprogesterone acetate. BMJ 1994; 308:247. Swahn ML, Westlund P, Johannisson E, Bygdeman M. Effect of postcoital contraceptive methods on the endometrium and the menstrual cycle. Acta Obstet Gynecol Scand 1996; 75:738. Ling WY, Robichaud A, Zayid I, et al. Mode of action of DL-norgestrel and ethinyl estradiol combination in postcoital contraception. Fertil Steril 1979; 32:297. Rowlands S, Kubba AA, Guillebaud J, Bounds W. A possible mechanism of action of danazol and an ethinyl estradiol/norgestrel combination used as postcoital contraceptive agents. Contraception 1986; 33:539. Ling WY, Wrixon W, Acorn T, et al. Mode of action of dlnorgestrel and ethinyl estradiol combination in postcoital contraception. III. Effect of pre-ovulatory administration following the luteinizing hormone surge on ovarian steroidogenesis. Fertil Steril 1983; 40:631. Trussell J, Ellertson C, Stewart F. The effectiveness of the Yuzpe regimen of postcoital contraception. Fam Plann Perspect 1996; 28(2):58. American College of Obstetricians and Gynecologists. Practice patterns. Emergency oral contraception. Washington: American College of Obstetricians and Gynecologists, December 1996 (no. 3). Trussell J, Ellertson C. Efficacy of emergency contraception. Fertil Control Rev 1995; 4(2):8. Lockwood CJ, Krikin G, Papp C, et al. Biological mechanisms underlying RU486 clinical effects: inhibition. J Clin Endocrinol Metab 1994; 79:786.

62a. Baird DT. Clinical uses of antiprogestagens. J Soc Gynecol Investig 2000; 7(1 Suppl):S49. 63. Glasier A, Thong KJ, Dewar M, et al. Mifepristone (RU486) compared with high-dose estrogen and progestogen for emergency postcoital contraception. N Engl J Med 1992; 237:1041. 63a. Harvey SM, Beckman LJ, Sherman C, Petitti D. Women's experience and satisfaction with emergency contraception. Fam Plann Perspective 1999; 31:237. 63b. Hewitt G, Cromer B. Update on adolescent contraception. Obstet Gynecol Clin North Am 2000; 27:143. 63b. Hewitt G, Cromer B. Update on adolescent contraception. Obstet Gynecol Clin North Am 2000; 27:143. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81.

Braken MB. Oral contraceptives and congenital malformations in offspring; a review and meta-analysis of the prospective studies. Obstet Gynecol 1990; 76:552. Kjos SL, Ballagh SA, La Cour M, et al. The copper T380A intrauterine device in women with type II diabetes mellitus. Obstet Gynecol 1994; 84:1006. Kimmerle R, Weiss R, Berger M, Kurz K. Effectiveness, safety and acceptability of a copper intrauterine device (Cu Safe 300) in type I diabetic women. Diabetes Care 1993; 16:1227. Speroff L, Darney PD. Intrauterine contraception. In: A clinical guide for contraception, 2nd ed. Baltimore: Williams & Wilkins, 1996. Querido L, Ketting E, Haspels AA. IUD insertion following induced abortion. Contraception 1985; 31:603. Chi I-C, Farr G. Postpartum IUD contraception—a review of an international experience. Adv Contracept 1989; 5:127. Lee NC. The intrauterine device and pelvic inflammatory disease revisited: new results from the Women's Health Study. Obstet Gynecol 1988; 72:1. Kessel E. Pelvic inflammatory disease with intrauterine device use: a reassessment. Fertil Steril 1989; 51:1. Cramer DW. Tubal infertility and intrauterine device. N Engl J Med 1985; 312:941. Daling JR. Primary tubal infertility in relation to use of intrauterine device. N Engl J Med 1985; 312:937. Laga M, Alary M, Nzila N, et al. Condom promotion, sexually transmitted disease treatment, and declining incidence of HIV-1 infection in female Zairian sex workers. Lancet 1994; 1:246. Jick H. Vaginal spermicides and congenital disorders. JAMA 1981; 1245:1329. Cordero JF, Layde PM. Vaginal spermicides, chromosomal abnormalities and limb reduction defects. Fam Plann Perspect 1983; 15:16. Van Damme L, Niruthisard S, Atisook R, et al. Safety evaluation of nonoxynol-9 gel in women at low risk of HIV infection. AIDS 1998; 12:433. Petersen HB, Xia Z, Hughes JM, et al. The risk of pregnancy after tubal sterilization: findings from the U.S. Collaborative Review of Sterilization. Am J Obstet Gynecol 1996; 174:1161. Petersen HB, Xia Z, Hughes JM, et al. The risk of ectopic pregnancy after tubal sterilization. N Engl J Med 1997; 336:762. Escobedo LG, Petersen HB, Grubb GS, Franks AL. Case-fatality rates for tubal sterilization in U.S. hospitals, 1979 to 1980. Am J Obstet Gynecol 1989; 160:147. DeStefano F. Complications of interval laparoscopic tubal sterilization. Obstet Gynecol 1983; 61:163.

CHAPTER 105 COMPLICATIONS AND SIDE EFFECTS OF STEROIDAL CONTRACEPTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 105 COMPLICATIONS AND SIDE EFFECTS OF STEROIDAL CONTRACEPTION ALISA B. GOLDBERG AND PHILIP DARNEY Pharmacology Biologic Potencies of Contraceptive Steroids Metabolic Effects Carbohydrate Metabolism Lipid Metabolism Cardiovascular Effects Thromboembolic Disease Myocardial Infarction Cerebrovascular Accidents Hypertension Effects on the Breasts Benign Breast Changes Breast Cancer Effects on the Reproductive Tract Ovarian Effects Endometrial Effects Cervical Effects Effects on the Gastrointestinal Tract Liver and Biliary Tree Effects Liver Tumors Effects on Bone Mineral Density Common Minor Side Effects Chapter References

All steroidal contraceptives are composed either of a progestin alone or a combination of an estrogen and a progestin. The side effects or complications observed with the use of hormonal contraception can be attributed either to the estrogen or progestin component. Understanding which effects are estrogen related and which are progestin related can help clinicians individualize hormonal contraceptive use for their patients.

PHARMACOLOGY Hormonal contraceptives are formulated from synthetic steroid 19-nortestosterone derivatives, which include norethindrone, norethindrone acetate, norethynodrel, ethynodiol diacetate, norgestrel, levonorgestrel (LN), desogestrel, gestodene, norgestimate, and dienogest. Exceptions are the injectable contraceptives (Depo-Provera, Lunelle), which use the progesterone derivative, depot medroxyprogesterone acetate (MPA). One of these progestins is combined with various dosages of an estrogen, either ethinyl estradiol or ethinyl estradiol-3-methyl ether (mestranol) in oral tablets and in some long-acting methods (injectables, vaginal rings, patches). The presence of a C17 ethinyl group on all synthetic estrogens and progestins enhances the oral activities of these agents by slowing their rapid hydroxylation and conjugation in the hepatic portal system.1,2 Ethinyl estradiol and mestranol are fairly well absorbed; ~60% of the oral dose is recovered in the urine. In the liver, mestranol is demethylated to ethinyl estradiol. After oral administration, concentrations of both estrogens peak at 1 to 2 hours, with the areas under the plasma concentration time curves being equal. The metabolic degradation path of the two estrogens is identical, with the principal urinary metabolite being ethinyl estradiol glucuronide. The progestins used in hormonal contraceptives are absorbed rapidly, with peak concentrations reached in ~1 hour after a pill is taken. The acetate compounds attain peak concentrations somewhat later because they must be deacetylated in the gastrointestinal tract before the progestin can be absorbed. All progestins are hydroxylated and conjugated in the liver before excretion primarily in the urine. Drugs that accelerate hepatic metabolism of steroids (rifampin, barbiturates, phenytoin, carbamazepine, fluconazole) can decrease the serum concentrations of low-dose hormonal contraceptives, such as the minipills or Norplant, and decrease efficacy. BIOLOGIC POTENCIES OF CONTRACEPTIVE STEROIDS Various animal tissue responses (e.g., rat ventral prostate) have been used to assess the biopotency of contraceptive steroids. Much scientific criticism has been directed at these test systems, particularly at the extrapolation of dog and rodent data to humans. Specific steroid-receptor binding assays are now used. These allow in vitro comparisons of the androgenic, progestogenic, and estrogenic properties of sex steroids. In vivo, however, these effects are modulated by endogenous sex steroids and their binding globulins, notably sex hormone–binding globulin (SHBG). Progestins differ in their bioactivities. This variation in bio-activity among different progestins is in large part due to modifications in the steroid structure that result in different receptor-binding affinities and different rates of metabolism. These factors require that different doses of each different progestin be used to achieve contraceptive efficacy. The administration of estrogen together with a progestin in combined contraceptives allows for use of a lower dose of progestin. The dose of progestin required for contraceptive efficacy in combined contraceptives is affected by the amount of estrogen administered. Because differences among progestin potencies are compensated for by dose adjustment, scales that attempt to correlate steroid dose with clinical effect are not useful.3

METABOLIC EFFECTS Pharmacologic doses of contraceptive hormones have widespread metabolic effects, but many of these are merely alterations in laboratory values without clinical significance. Nevertheless, some laboratory test alterations may reflect clinically significant metabolic changes. For example, changes in coagulation factors may predispose certain women to intravascular clotting, and changes in renin and angiotensin may affect blood pressure in a few users. Many of the metabolic alterations associated with steroid contraceptive use are attributable to the estrogenic component of the combination pill. These effects would not be expected with the use of progestin-only contraception. In contrast, the metabolic alterations caused by progestins, which would be expected in progestin-only contraceptive users, may be altered by the concomitant use of estrogen.4 For example, the estrogen in combined oral contraceptives (COCs) raises triglycerides and high-density lipoprotein (HDL), whereas most progestins have the opposite effects. CARBOHYDRATE METABOLISM Combined Oral Contraception. Early studies with high-dose COCs showed impairment of glucose tolerance and increased insulin resistance. However, multiple studies of low-dose COCs have not shown a clinically significant impact of COCs on carbohydrate metabolism.5,6 Even women with a history of gestational diabetes have not been found to be at additional risk of developing diabetes due to COC use.7,8 The small increase in insulin resistance seen with low-dose COC use may alter glucose metabolism in some women with overt diabetes mellitus; however, this effect has not been consistent among individual patients. Also, the use of COCs has not been found to increase the risk of development of nephropathy or retinopathy in patients with type 1 diabetes mellitus.9 Current consensus opinion is that healthy diabetic women with no end-organ complications of diabetes mellitus may safely use low-dose COC. The changes observed in carbohydrate metabolism with oral contraceptive use (increased insulin resistance and decreased glucose tolerance) are believed to be attributable almost entirely to the progestin component of combination pills, and are dose related. Ethinyl estradiol administered alone, even in high doses, does not cause glucose tolerance deterioration or abnormal insulin responses. Progestin-Only Contraceptives. Progestin-only oral contraceptives, minipills or “POPs,” may decrease carbohydrate tolerance in healthy women, but this effect is generally not clinically significant.10 Because they do not adversely affect breast milk volume or quality, POPs are often used by lactating women. Those containing norethindrone have been found to increase the risk of developing diabetes among high-risk breast-feeding women (obese Latinas with a history of gestational diabetes; relative risk [RR] = 2.87, 95% confidence interval [CI] = 1.57– 5.27).8 The same study found no increase in the rate of development of diabetes on the basis of breast-feeding alone.8 The predominantly progestogenic state of lactation combined with POPs may be enough to cause significant glucose intolerance in women at high risk. It remains unknown whether POPs will have a similar diabetogenic effect on high-risk, nonlactating women. One study of non–breast-feeding, type 1 diabetic

women using POPs containing lynestrenol found no change in insulin requirements.11 LN implant contraceptives (e.g., Norplant, Jadelle) have been shown to have no clinically significant effect on carbohydrate metabolism in healthy women. In one study, 100 healthy women had a glucose-loading test before Norplant insertion, and then annually over 5 years of its use. The investigators found no significant changes in mean fasting serum glucose levels or in 2-hour postprandial serum glucose levels. The 1-hour postprandial glucose levels were elevated from baseline in years 1 and 2 of Norplant use, but not in years 3 through 5. These elevations were not above the normal range.12 There is no evidence that LN implants increase the risk of developing diabetes among lactating or nonlactating women at high risk, as POPs containing norethindrone may. Since depot MPA acetate injection (DMPA, Depo-Provera) results in higher serum concentrations of progestin at the beginning of each 3-month injection cycle, its effects on carbohydrate metabolism might differ from those of lower-dose POPs or continuous-release progestin implant systems. Studies have shown no deterioration in glucose tolerance in nondiabetic women using DMPA for contraception for short durations; however, after 4 to 5 years of continuous use, some studies show an increase in glucose intolerance.13 Whether this is due to DMPA itself, or to associated weight gain, is unclear. DMPA is not contraindicated in diabetic women, and often is an excellent method of contraception for women with vascular disease; however, changes in glucose metabolism may occur. LIPID METABOLISM Combined Oral Contraceptives. All estrogen-containing contraceptive pills increase serum triglyceride levels by an average of ~50%. The progestin-only pill has not been associated with such changes. The increased triglyceride levels are attributable mainly to an increase in very low-density lipoproteins (LDLs; see Chap. 162). Most low-dose COCs do not cause significant increases in mean serum cholesterol levels; however, high-dose estrogen-progestin formulations can decrease HDL while increasing LDL.10 Although estrogens increase HDL levels and progestins decrease HDL levels, COCs may have various effects because of endogenous factors that modulate these effects. COCs containing the less androgenic progestins (desogestrel, dienogest, gestodene, and norgestimate) moderately elevate triglycerides (as does the estrogen component of all COCs), as well as total cholesterol, but the increase in total cholesterol is all in the HDL fraction. LDL levels fall, so that except for women with very high triglycerides (>450 mg/dL), less androgenic COCs can improve the lipoprotein profile.14,15 and 16 Whether this effect has consequences for cardiovascular health is undetermined. High-dose COCs can decrease HDL and increase LDL; however, they have not been associated with arteriosclerotic disease. In fact, the estrogen component of high-dose COCs protects against plaque deposition despite the adverse lipid effects of high doses of progestins. Progestin-Only Contraceptives. Low-dose, sustained-release contraceptives (e.g., Norplant) do not perturb lipoprotein metabolism. A study of more than 20 LN implant users followed for 5 years showed that modest changes in the cholesterol/HDL ratio were accounted for by weight gain and aging. The low serum concentrations of LN (0.3–0.5 ng/mL) had no persistent effect on lipoprotein metabolism (Population Council, data on file). Other smaller, shorter-term studies also failed to show significant lipoprotein effects of LN implants.12,17 With LN implant use, triglyceride levels fall slightly because no estrogen is administered, and endogenous estradiol production is modestly suppressed. The effect of DMPA on lipid metabolism is not clear. Some studies have suggested that DMPA has a negative impact on lipids, because it has been associated with decreased HDL cholesterol and increased total and LDL cholesterol levels.18 Other studies have not found DMPA to be associated with these negative changes in lipids.19,20 Epidemiologic studies have not associated DMPA with cardiovascular disease.21

CARDIOVASCULAR EFFECTS THROMBOEMBOLIC DISEASE Combined Oral Contraceptives. Early epidemiologic studies22,23 and 24 indicated a four-fold to eight-fold increase in the risk of venous thromboembolism (VTE) among oral contraceptive users. As the estrogen content of COCs declined, reported risks of VTE fell to approximately three-fold, but the increased risk of VTE remains the greatest health threat that COCs pose. 25,26,26a A large World Health Organization (WHO) international multicenter hospital-based case-control study found an increased risk of VTE with low-dose COC use (Europe: odds ratio [OR] = 4.24, 95% CI 3.07–5.87; developing countries: OR = 3.02, 95% CI 2.28–4.00).27 Further analysis of the data from this study, 28 as well as others,26 has suggested an additional two-fold increase in VTE risk among users of COCs containing desogestrel and gestodene, compared to users of COCs containing LN. The additional risk of VTE observed with the use of desogestrel- or gestodene-containing COCs is best explained by selection and prescribing bias.26 Women at highest risk of VTE, new starters, and women who have had complications on COCs in the past are most likely to have received COCs containing desogestrel or gestodene. In contrast, women who had been using COCs without complication for years comprise a population at low risk of complications, and were more likely to remain on older formulations. Whether selection and prescribing bias completely account for this observed effect is unknown. Genetic predisposition also plays a role in modifying the risk of VTE associated with oral contraceptive use. The factor V Leiden mutation is a point mutation that results in resistance to the anticoagulant effects of activated protein C. The factor V Leiden mutation is estimated to affect 4.4% of Europeans and 0.6% of Asians, and is extremely rare (or nonexistent) in populations from Africa and Southeast Asia.29 Women with the factor V Leiden mutation have an eight-fold increased risk of VTE (RR = 7.9, 95% CI 3.2–19.4), compared to women without the mutation.30 When a woman with the factor V Leiden mutation uses COCs, her baseline risk of VTE increases four-fold, and her overall risk of VTE is 35 times greater (RR = 34.7, 95% CI 7.8–154) than for women without the mutation who are not using COCs.30 Screening the general population for the factor V Leiden mutation is currently not recommended; however, women who are known to have the mutation or a strong family history of thromboembolic disease should avoid COCs. Other thrombophilic disorders, such as protein C or S deficiency, also increase the risk of VTE with COC use. The increased risk of VTE with COCs is largely due to the estrogen component. Progestin-only contraceptives do not appear to confer an increased risk of VTE.21 Therefore, POPs, implants, and injectables are often reasonable contraceptive choices for women at high risk of VTE. MYOCARDIAL INFARCTION Combined Oral Contraceptives. Several early epidemiologic studies of COCs containing 50 µg estrogen showed an increased risk of myocardial infarction (MI).31 Much of the observed increased risk of MI among COC users was actually due to the increased incidence of smoking and hypertension among COC users in the past. More recent studies of low-dose COCs have not revealed an increased risk of MI among nonhypertensive COC users who do not smoke.32,33 Similarly, current or past use of COCs has not been associated with an increased risk of mortality from MI.34 Among women who smoke and use COCs, the RR of MI has been reported to range from 3.5 (1.3–9.5) for those who smoke 53,000 women with breast cancer and >100,000 controls. They found that women who are currently using COCs have a small increase (RR = 1.24, CI 1.15–1.33) in the risk of breast cancer as compared to nonusers. This risk remains slightly elevated within the first 10 years after discontinuation of use (1–4 years after stopping: RR = 1.16, CI 1.08– 1.23; 5–9 years after stopping: RR = 1.07, CI 1.02–1.13). Beyond 10 years after discontinuation of use, there is no increased risk of breast cancer (10 or more years after stopping: RR = 1.01, CI 0.96–1.05). They found no significant effect of duration of use, age at first use, or the type of hormone contained in the COC on breast cancer risk. In this study, although women currently using COCs were at increased risk of a new diagnosis of breast cancer, their tumors were significantly more likely to be confined to the breast at diagnosis than were those of nonusers (RR = 0.89 for spread to lymph nodes at diagnosis; RR = 0.70 for distant spread at diagnosis). Whether these effects are due to earlier diagnosis in current or recent COC users or due to a pathophysiologic effect of COCs on breast cancer remains unclear. The risk of breast cancer may be greater for younger than for older women.54a (See also Chap. 222 and Chap. 223.) Progestin-Only Contraceptives. DMPA was initially found to cause malignant mammary tumors in beagle dogs. This finding created great fear as to whether DMPA would have the same effect in humans. Two large case-controlled studies, and one pooled analysis of these studies, have addressed this concern and have found no increased risk of breast cancer in ever-users of DMPA as compared to never-users.55,56 and 57 The pooled analysis did show a slightly increased risk of breast cancer in DMPA users within the first 5 years of use (pooled analysis: RR = 2.0, CI 1.5–2.8).57 Whether this is due to increased surveillance in DMPA users or stimulation of preexisting tumors is unclear. Two of the studies showed no increase in breast cancer risk with increasing duration of use.55,57 The third study showed an increase in breast cancer risk with >6 years of DMPA use only in those women who began using DMPA before age 25 years (RR = 4.2, CI 1.1–16.2).56 Based on a few studies, POPs do not appear to increase breast cancer risk.50,58 To date, there are no epidemiologic studies evaluating the effect of LN implants on breast cancer risk.

EFFECTS ON THE REPRODUCTIVE TRACT OVARIAN EFFECTS OVARIAN CYSTS Combined Oral Contraceptives. High-dose COCs have been shown to decrease the incidence of functional ovarian cysts.59 Low-dose COCs have not demonstrated a similar effect, although an attenuated effect may exist.60 Progestin-Only Contraceptives. An increased incidence of follicular cysts has been noted with LN implant (Norplant) use. The continuous low level of LN allows follicles to develop, and follicular cysts often result.60a These cysts only need to be evaluated sonographically or laparoscopically if they become large and painful. Similar to LN implants, the low dose of progestin in POPs allows follicles to develop and follicular cysts to form. However, one large cohort study evaluating ovarian cysts in COC users also looked at progestin-only pill users, and found no cases of follicular cysts in 219 person-months of observation.59 In contrast to POPs and implants, DMPA effectively suppresses follicular development and ovulation; thus, follicular cysts are rare among DMPA users. OVARIAN CANCER Combined Oral Contraceptives. Women using COCs have a markedly reduced risk of ovarian cancer, with increasing duration of use increasing the protective effect.61,62 and 63 A 10% decrease in risk is noted after 1 year of COC use, and a ~50% decrease in risk is achieved after 5 years of use.62 This protective effect has been found to extend to women at highest risk of ovarian cancer, including women with BRCA-1 and BRCA-2 mutations64 (see also Chap. 223). Progestin-Only Contraception. DMPA is probably associated with a slightly decreased risk of epithelial ovarian cancer, given that it effectively inhibits ovulation; however, studies have been unable to detect a difference in ovarian cancer risk largely because of the high parity of DMPA users.65 There are no epidemiologic data

evaluating the effects of POPs or implants on ovarian cancer risk. ENDOMETRIAL EFFECTS MENSTRUAL CHANGES In the majority of COC users, menses are regular, and become shorter in duration, lighter in flow, and associated with less dysmenorrhea. Progestin-only contraceptives are usually associated with disruption of the menstrual cycle, but no overall increase in menstrual blood loss. Menstrual irregularity is one of the leading causes of dissatisfaction with progestin-only methods. (See Chap. 104 for a complete discussion of menstrual changes with hormonal contraceptives.) ENDOMETRIAL CANCER Combined Oral Contraceptives. Multiple studies indicate that the risk of endometrial cancer is reduced among users of COCs.66,67 Longer duration of use is associated with increased protection against endometrial cancer. A metaanalysis found that women using COCs for 4 years had a 56% decreased risk of endometrial cancer; after 8 years of use, a woman's risk was reduced by 67%, and after 12 years of use, risk was reduced by 72%.68 Although the greatest risk reduction is observed for women who are currently using COCs, even 20 years after discontinuation, the risk of endometrial cancer is ~50% lower in ever-users than never-users (see also Chap. 223). Progestin-Only Contraceptives. DMPA use has been associated with a decreased risk of endometrial cancer.69 Although epidemiologic studies have not yet shown a decreased risk of endometrial cancer with POPs or LN implant use, given the protective effect of progestin on the endometrium, these agents may decrease endometrial cancer risk as well. CERVICAL EFFECTS CERVICAL CANCER Combined Oral Contraceptives. The relation of oral contraceptive use to cervical dysplasia, carcinoma in situ, and invasive squamous cell cervical cancer is unclear, since some studies show positive relations between cervical dysplasia and continuing oral contraceptive use, whereas other studies report no relation.70,71 Cervical dysplasia, carcinoma in situ, and squamous cell carcinoma are thought to be at least partly of viral origin (human papilloma virus). Consequently, the confounding factors of coitus at early age, multiple sexual partners, sexually transmittable diseases, and oral contraceptive use are difficult to decipher. Similarly, women using COCs are subject to increased surveillance as compared to nonusers. Increased surveillance results in increased detection of cervical squamous cell disease, and although the association between COC use and cervical squamous cell disease may be real, it may reflect screening bias or confounding (see Chap. 223). Nonetheless, annual Papanicolaou (Pap) smear screening should be recommended for all women taking COCs, and women at highest risk because of multiple sexual partners or a history of sexually transmitted diseases should be screened twice a year. There is convincing evidence suggesting a relationship between COC use and adenocarcinoma of the cervix.72 Data from the Surveillance, Epidemiology and End Results (SEER) tumor registry for Los Angeles found ever-users of COCs to have an RR of 2.1 of adenocarcinoma of the cervix, compared to never-users (CI 1.1–3.8). With COC use of >12 years' duration, this rose to a RR of 4.4 for ever-users (CI 1.8–10.8). This association may be mediated directly via estrogen or progestin receptors on endocervical cells, or may be explained by the increased incidence of ectropion in oral contraceptive users. Ectropion results in exposure of endocervical cells to the vagina, which may result in increased exposure to carcinogens. Progestin-Only Contraceptives. DMPA does not appear to have an independent effect on cervical cancer or dysplasia. One study found an increased risk of dysplasia among DMPA users; however, this increased risk was attributable to known risk factors for cervical dysplasia among DMPA users.73 A WHO case-control study found a slightly higher risk of carcinoma in situ in DMPA users.74 Although this may be a real finding, it may reflect confounding or screening biases. No epidemiologic data yet exist evaluating the effects of POPs and implants on cervical cancer risk. When these data become available, it is likely that similar issues of confounding and screening bias will make interpretation difficult.

EFFECTS ON THE GASTROINTESTINAL TRACT LIVER AND BILIARY TREE EFFECTS Combined Oral Contraception. COCs have a variety of effects on the liver. Estrogen influences the hepatic synthesis of DNA, RNA, enzymes, plasma proteins, lipids, and lipoproteins. It also influences the hepatic metabolism of carbohydrates and intracellular enzyme activity. Progestins have less, if any, effect on the liver. Although the liver is affected in a variety of ways by COCs, many of these changes have proved to be of no clinical significance, and others remain incompletely understood. Virtually all phase II and phase III studies of COCs have evaluated the effects of COCs on liver function tests, and have shown no effect.75 Some studies have shown an increased incidence of gallbladder disease and gallstones among current COC users.76 Other studies77 and a metaanalysis78 of the effect of COCs on biliary disease showed that COCs initially did increase the risk of biliary disease (RR = 1.36, 95% CI 1.15–1.62),78 but over time the risk returns to baseline. This suggests that, under the influence of COCs, susceptible (or asymptomatic) women become symptomatic from biliary disease shortly after starting the pill, whereas nonsusceptible women do not develop biliary disease over time with continued use. The mechanism by which estrogen has its cholestatic effects79 is incompletely understood. Under the influence of estrogen, bile becomes increasingly saturated with cholesterol.80 This effect is probably secondary to elevated cholesterol concentrations in the bowel caused by altered cholesterol and lipid metabolism dependent on estrogen dose. Given these effects, women with active hepatitis, jaundice, or cholestasis should avoid COCs. Women with a past history of hepatitis can safely be given COCs. Progestin-Only Contraceptives. POPs, implants, and injectables do not alter liver function tests.81,82 Women with a past history of liver disease may safely use these methods. Whether women with active hepatitis or cirrhosis should use progestin-only methods is controversial, and the decision should be strongly influenced by the likelihood of pregnancy with other methods. Clearly, nonhormonal methods are safest for these women; however, progestin-only methods are less likely to have an adverse effect on their liver disease than either COCs or pregnancy.83 It is unlikely that progestin-only methods increase the risk of gallbladder disease.83 LIVER TUMORS Combined Oral Contraceptives. The use of COCs has been associated with an increased risk of hepatocellular adenoma, and risk may increase with longer duration of use.84,85 Although hepatocellular adenomas are benign tumors, they may rupture, causing hemorrhage and death. Several case-control studies have shown an increased incidence of hepatocellular carcinoma among COC users,86 whereas other similar studies have not demonstrated the same effect.85 A large population-based study evaluated trends in the incidence of primary liver cancer and concomitant oral contraceptive use in three countries, and found no association of oral contraceptive use with primary liver cancer.87 Progestin-Only Contraceptives. The WHO found no association between DMPA use and liver cancer.69 There are no data linking POPs or implants to benign or malignant liver tumors.

EFFECTS ON BONE MINERAL DENSITY Combined Oral Contraceptives. Estrogen-replacement therapy has been proven to prevent bone loss and fractures in postmenopausal women. Similarly, women with a history of COC use are less likely to have low bone mineral density later in life (RR = 0.35, 95% CI 0.2–0.5).88 The effect of COCs on increasing bone mineral density appears to be related to duration of use.88 As the first large population of COC users pass into menopause, future epidemiologic studies will determine whether women with a history of COC use have fewer fractures than women who never used COCs. It also remains to be seen if a past history of COC use is protective against osteoporosis in the presence and absence of postmenopausal estrogen-replacement therapy. Progestin-Only Contraceptives. In cross-sectional studies, DMPA use has been associated with a decrease in bone mineral density.89,90,91 and 91a This effect appears

to be greater with a longer duration of use and among young women (ages 18–21).91 This decrease in bone mineral density is likely the result of decreased endogenous estrogen secretion due to the suppression of follicular development by DMPA. With discontinuation of DMPA, bone mineral density recovers92; however, the long-term effects of a temporary loss in bone mineral density remain unknown. Since POPs and implants do not completely suppress follicular development, endogenous estrogen secretion remains within the normal premenopausal range93; it is unlikely that progestin-only pills or implants adversely affect bone mineral density. Different progestins also might directly affect bone mineral density differently. In vitro, osteoblasts have been found to have both estrogen and progesterone receptors.94 In other cell lines, the nortestosterone derivatives (i.e., norethindrone, norgestrel) have been shown to stimulate the growth of estrogen-receptor–positive cells, whereas MPA did not.95 Perhaps the nortestosterone group of progestins has an estrogenic effect on bone that MPA does not.

COMMON MINOR SIDE EFFECTS Other common side effects of hormonal contraception include weight gain,96 nausea, headaches, skin changes, and changes in mood and libido. Although these side effects are usually not dangerous, they often limit the acceptability of a contraceptive method. These side effects are discussed in Chapter 104. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Goldzieher JW. Selected aspects of the pharmacokinetics and metabolism of ethinyl estrogens and their clinical implications. Am J Obstet Gynecol 1990; 163:318. Collins DC. Sex hormone receptor binding, progestin selectivity, and the new oral contraceptives. Am J Obstet Gynecol 1994; 170:1508. Upmalis D, Phillip A. Receptor binding and in vivo activities of the new progestins. J Soc Obstet Gynecol Can 1991; 13(Suppl):35. Krauss RM, Burkman RT Jr. The metabolic impact of oral contraceptives. Am J Obstet Gynecol 1992; 167:1177. Rimm EB, Manson JE, Stampfer MJ, et al. Oral contraceptive use and the risk of type 2 (non-insulin dependent) diabetes mellitus in a large prospective study of women. Diabetologia 1992; 35:967. Duffy TJ, Ray R. Oral contraceptive use: prospective follow-up of women with suspected glucose intolerance. Contraception 1984; 30:197. Kjos SL, Shoupe D, Douyan S, et al. Effect of low-dose oral contraceptives on carbohydrate and lipid metabolism in women with recent gestational diabetes: results of a controlled, randomized, prospective study. Am J Obstet Gynecol 1990; 163:1822. Kjos SL, Peters RK, Xiang A, et al. Contraception and the risk of type 2 diabetes mellitus in Latina women with prior gestational diabetes mellitus. JAMA 1998; 280:533. Garg SK, Chase HP, Marshall G, et al. Oral contraceptives and renal and retinal complications in young women with insulin-dependent diabetes mellitus. JAMA 1994; 271:1099. Godsland IF, Crook D, Simpson R, et al. The effects of different formulations of oral contraceptive agents on lipid and carbohydrate metabolism. N Engl J Med 1990; 323:1375. Radberg T, Gustafson A, Skryten A, Karlsson K. Oral contraception in diabetic women: a cross-over study on serum and high density lipoprotein (HDL) lipids and diabetes control during progestogen and combined estrogen/progestogen contraception. Horm Metab Res 1982; 14:61. Singh K, Viegas OAC, Loke D, Ratnam SS. Effect of Norplant implants on liver, lipid and carbohydrate metabolism. Contraception 1992; 45:141. Liew DF, Ng CS, Yong YM, Ratnam SS. Long-term effects of Depo-Provera on carbohydrate and lipid metabolism. Contraception 1985; 31:51. Speroff L, De Cherney A. Evaluation of a new generation of oral contraceptives. Obstet Gynecol 1993; 81:1034. Kloosterboer HJ, Vonk-Noordegraaf CA, Turpijn EW. Selectivity in progesterone and androgen receptor binding of progestins in oral contraceptives. Contraception 1988; 38:325. Petersen KR, Skouby SO, Pedersen RG. Desogestrel and gestadene in oral contraceptives: 12 months assessment of carbohydrate and lipoprotein metabolism. Obstet Gynecol 1991; 78:666. Viegas OAC, Singh K, Liew D, et al. The effects of Norplant on clinical chemistry in Singaporean acceptors after 1 year of use: metabolic changes. Contraception 1988; 38:79. World Health Organization. A multicentre comparative study of serum lipids and apolipoproteins in long-term users of DMPA and a control group of IUD users. Contraception 1993; 47:177. Mainwaring R, Hales HA, Stevenson K, et al. Metabolic parameter, bleeding and weight changes in U.S. women using progestin only contraceptives. Contraception 1995; 51:149. Garza-Flores J, De la Cruz DL, Valles de Bourges V, et al. Long-term effects of depot medroxyprogesterone acetate on lipoprotein metabolism. Contraception 1991; 44:61. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Cardiovascular disease and use of oral and injectable progestogen-only contraceptives and combined injectable contraceptives. Contraception 1998; 57:315. Royal College of General Practitioners. Oral contraceptives and health: interim report. New York: Pitman, 1974. Collaborative Group for the Study of Stroke in Young Women. Oral contraception and increased risk of cerebral ischemia or thrombosis. N Engl J Med 1973; 288:871. Collaborative Group for the Study of Stroke in Young Women. Oral contraceptives and stroke in young women: associated risk factors. JAMA 1975; 231:718. Gerstman BB, Piper JM, Tomita DK, et al. Oral contraceptive estrogen dose and the risk of deep venous thromboembolic disease. Am J Epidemiol 1991; 133:32. Lewis MA, Heinemann LAJ, MacRae KD, et al. The increased risk of venous thromboembolism and the use of third generation progestagens: role of bias in observational research. Contraception 1996; 54:5.

26a. Editorial. Oral contraceptives and cardiovascular risk. Drug Ther Bull 2000; 38:1. 27. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Venous thromboembolic disease and combined oral contraceptives: results of international multicentre case-control study. Lancet 1995; 346:1575. 28. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Effect of different progestagens in low oestrogen oral contraceptives on venous thromboembolic disease. Lancet 1995; 346:1582. 29. Rees DC, Cox M, Clegg JB. World distribution of factor V Leiden. Lancet 1995; 346:1133. 30. Vandenbrouke JP, Koster T, Briet E, et al. Increased risk of venous thrombosis in oral-contraceptive users who are carriers of factor V Leiden mutation. Lancet 1994; 344:1453. 31. Pettiti DB, Sidney S, Quesenberry CP. Oral contraceptive use and myocardial infarction. Contraception 1998; 57:143. 32. Croft P, Hannaford PC. Risk factors for acute myocardial infarction in women: evidence from the Royal College of General Practitioners' oral contraception study. Br Med J 1989; 298:165. 33. Sidney S, Pettiti DB, Quesenberry CP, et al. Myocardial infarction in users of low-dose oral contraceptives. Obstet Gynecol 1996; 88:939. 34. Beral V, Hermon C, Kay C, et al. Mortality associated with oral contraceptive use: 25 year follow up of cohort of 46,000 women from Royal College of General Practitioners' oral contraception study. Br Med J 1999; 318:96. 35. Schwingl PJ, Ory HW, Visness CM. Estimates of the risk of cardiovascular death attributable to low-dose oral contraceptives in the United States. Am J Obstet Gynecol 1999; 180:241. 36. Pettiti DB, Sidney S, Bernstein A, et al. Stroke in users of low-dose oral contraceptives. N Engl J Med 1996; 335:8. 37. Stampfer MJ, Willett WC, Colditz GA. A prospective study of past use of oral contraceptive agents and risk of cardiovascular diseases. N Engl J Med 1988; 319:1313. 38. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Ischaemic stroke and combined oral contraceptives: results of an international, multicentre, case-control study. Lancet 1996; 348:498. 39. World Health Organization Collaborative Study of Cardiovascular Disease and Steroid Hormone Contraception. Haemorrhagic stroke, overall stroke risk, and combined oral contraceptives: results of an international, multicentre, case-control study. Lancet 1996; 348:505. 40. Chasan-Taber L, Willett WC, Manson JE, et al. Prospective study of oral contraceptives and hypertension among women in the United States. Circulation 1996; 94:483. 41. Pritchard JA, Pritchard SA. Blood pressure response to estrogen-progestin oral contraception after pregnancy-induced hypertension. Am J Obstet Gynecol 1977; 733. 42. Speroff L, Darney PD. Implant contraception: Norplant. In: A clinical guide for contraception, 2nd ed. Baltimore: Williams & Wilkins, 1996. 43. Brinton LA, Vessey MP, Flavel R, Yeates D. Risk factors for benign breast disease. Am J Epidemiol 1981; 113:203. 44. World Health Organization Task Force on Oral Contraceptives. Effects of hormonal contraceptives on milk volume and infant growth. Contraception 1984; 30:505. 45. Nillson S, Mellbin T, Hofvander Y, et al. Long-term follow-up of children breast-fed by mothers using oral contraceptives. Contraception 1986; 34:443. 46. Sivin I, Mishell DR, Darney P, et al. Levonorgestrel capsule implants in the United States: a 5-year study. Obstet Gynecol 1998; 92:337. 47. Centers for Disease Control Cancer and Steroid Hormone Study. Long-term oral contraceptive use and risk of breast cancer. JAMA 1983; 249:1591. 48. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Breast cancer and combined oral contraceptives: results from a multinational study. Br J Cancer 1990; 61:110. 49. Huggins GR, Zucker PK. Oral contraceptives and neoplasia: 1987 update. Fertil Steril 1987; 47:733. 50. United Kingdom National Case-Control Study Group. Oral contraceptive use and breast cancer risk in young women. Lancet 1989; 1:1973. 51. Stadel BV, Schlesselman JJ, Murray PA. Oral contraceptives and breast cancer. Lancet 1989; 1:1257. 52. Vessey MP, McPherson K, Villard-Mackintosh L. Oral contraceptives and breast cancer: latest findings in a large cohort study. Br J Cancer 1989; 59:613. 53. White E, Malone KE, Weiss NS, Daling JR. Breast cancer among U.S. women in relation to oral contraceptive use. J NCI 1994; 86:505. 54. Collaborative Group on Hormonal Factors in Breast Cancer. Breast cancer and hormonal contraceptives: collaborative reanalysis of individual data on 53,297 women with breast cancer and 100,239 women without breast cancer from 54 epidemiological studies. Lancet 1996; 347:1713. 54a. Pathak DR, Osuch JR, He J. Breast carcinoma etiology: current knowledge and new insights into the effects of reproductive and hormonal risk factors in black and white populations. Cancer 2000; 88(5 Suppl):1230. 55. 56. 57. 58.

World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Breast cancer and depot-medroxyprogesterone acetate: a multinational study. Lancet 1991; 338:833. Paul C, Skegg DCG, Spears GFC. Depot medroxyprogesterone (Depo-Provera) and risk of breast cancer. Br Med J 1989; 299: 759. Skegg DCG, Noonan EA, Paul C, et al. Depot medroxyprogesterone acetate and breast cancer. JAMA 1995; 273:799. The Cancer and Steroid Hormone Study of the Centers for Disease Control and the National Institute of Child Health and Human Development. Oral contraceptive use and the risk of breast cancer. N Engl J Med 1986; 315:405. 59. Lanes SF, Birmann B, Walker AM, Singer S. Oral contraceptive type and functional ovarian cysts. Am J Obstet Gynecol 1992; 166:956. 60. Grimes DA, Godwin AJ, Rubin A, et al. Ovulation and follicular development associated with three low-dose oral contraceptives: a randomized controlled trial. Obstet Gynecol 1994; 83:29. 60a. Alvarez-Sanchez F, Brache V, Montes de Oca V, et al. Prevalence of enlarged ovarian follicles among users of levonorgestrel subdermal contraceptive implants (Norplant). Am J Obstet Gynecol 2000; 182:535. 61. Centers for Disease Control Cancer and Steroid Hormone Study. Oral contraceptive use and risk of ovarian cancer. JAMA 1983; 249:1596. 62. Hankinson SE, Colditz GA, Hunter DJ. A quantitative assessment of oral contraceptive use and risk of ovarian cancer. Obstet Gynecol 1992; 80:708. 63. Cancer and Steroid Hormone Study of the Centers for Disease Control and the National Institutes of Child Health and Human Development. The reduction in risk of ovarian cancer associated with oral contraceptive use. N Engl J Med 1987; 316:650. 64. Narod SA, Risch H, Moslehi R, et al. Oral contraceptives and the risk of hereditary ovarian cancer. N Engl J Med 1998; 339:424. 65. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Depot-medroxyprogesterone acetate (DMPA) and risk of epithelial ovarian cancer. Int J Cancer 1991; 49:191.

66. Centers for Disease Control Cancer and Steroid Hormone Study. Oral contraceptive use and risk of endometrial cancer. JAMA 1983; 249:1600. 67. Cancer and Steroid Hormone Study of the Centers for Disease Control and the National Institutes of Child Health and Human Development. Combination oral contraceptive use and the risk of endometrial cancer. JAMA 1987; 257:796. 68. Schlesselman JJ. Risk of endometrial cancer in relation to use of combined oral contraceptives. A practitioner's guide to meta-analysis. Hum Reprod 1997; 12:1851. 69. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Depot-medroxyprogesterone acetate (DMPA) and risk of endometrial cancer. Int J Cancer 1991; 49:186. 70. Brinton LA. Oral contraceptives and cervical neoplasia. Contraception 1991; 43:581. 71. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Combined oral contraceptives and risk of cervical carcinoma in situ. Int J Epidemiol 1995; 24:19. 72. Ursin G, Peters RK, Henderson BE, et al. Oral contraceptive use and adenocarcinoma of the cervix. Lancet 1994; 344:1390. 73. The New Zealand Contraception and Health Study Group. History of long-term use of depot-medroxyprogesterone acetate in patients with cervical dysplasia; case-control analysis nested in a cohort study. Contraception 1994; 50:443. 74. Thomas DB, Ye Z, Ray RM, and the World Health Organization Collaborative Study of Neoplasia and Steroid Contraception. Cervical carcinoma in situ and use of depot-medroxyprogesterone acetate (DMPA). Contraception 1995; 51:25. 75. Goldzieher JW. Effects on other tissues. In: Fraser IS, ed. Estrogens and progestogens in clinical practice. London: Churchill Livingstone, 1998. 76. Grodstein F, Colditz GA, Hunter DJ, et al. A prospective study of symptomatic gallstones in women: relation with oral contraceptives and other risk factors. Obstet Gynecol 1994; 84:207. 77. Royal College of General Practitioners' Oral Contraception Study. Oral contraceptives and gallbladder disease. Lancet 1982; 2:957. 78. Thijs C, Knipschild P. Oral contraceptives and the risk of gallbladder disease: a meta-analysis. Am J Pub Health 1993; 83:113. 79. Sillem MH, Teichmann AT. The liver. In: Goldheizer J, Fotherby K (eds). Pharmacology of the contraceptive steroids. New York: Raven Press, 1994:247. 80. Bennion LJ, Ginsberg RL, Garnick MB, Bennett PH. Effects of oral contraceptives on the gallbladder bile of normal women. N Engl J Med 1976; 294:189. 81. Korba VD, Paulson SR. Five years of fertility control with microdose norgestrel: an updated clinical review. J Reprod Med 1974; 13:71. 82. Population Council. Norplant levonorgestrel implants: a summary of scientific data. New York: The Population Council, 1990. 83. McCann MF, Potter LS. Progestin-only oral contraception: a comprehensive review. Contraception 1994; 50(Suppl 1):S96. 84. Palmer JR, Rosenberg L, Kaufman DW, et al. Oral contraceptive use and liver cancer. Am J Epidemiol 1989; 130:878. 85. World Health Organization Collaborative Study of Neoplasia and Steroid Contraceptives. Combined oral contraceptives and liver cancer. Int J Cancer 1989; 43:254. 86. Prentice RL. Epidemiologic data on exogenous hormones and hepatocellular carcinoma and selected other cancers. Prev Med 1991; 20:38. 87. Waetjen LE, Grimes DA. Oral contraceptives and primary liver cancer: temporal trends in three countries. Obstet Gynecol 1996; 88:945. 88. Kleerekoper M, Brienza RS, Schultz LR, Johnson CC. Oral contraceptive use may protect against low bone mass. Arch Intern Med 1991; 151:1971. 89. Cundy T, Evans M, Roberts H, et al. Bone density in women receiving depot medroxyprogesterone acetate for contraception. Br Med J 1991; 303:13. 90. Cromer BA, Blair JM, Mahan JD, et al. A prospective comparison of bone density in adolescent girls receiving depot medroxyprogesterone acetate (Depo-Provera), levonorgestrel (Norplant ®) or oral contraceptives. J Pediatr 1996; 129:671. 91. Scholes D, Lacroix AZ, Ott SM, et al. Bone mineral density in women using depot-medroxyprogesterone acetate for contraception. Obstet Gynecol 1999; 93:233. 91a. Petiti DB, Piaggio G, Mehta S, et al. for the WHO Study of Hormonal Contraception and Bone Health. Steroid hormone contraception and bone mineral density: a cross-sectional study in an international population. Obstet Gynecol 2000; 95:736. 92. 93. 94. 95. 96.

Cundy T, Cornish J, Evans MC, et al. Recovery of bone density in women who stop using medroxyprogesterone acetate. BMJ 1994; 308:247. Faundes A, Brache V, Tejada AS, et al. Ovulatory dysfunction during continuous administration of low-dose levonorgestrel by subdermal implants. Fertil Steril 1991; 56:273. Eriksen EF, Colvard DS, Berg NJ, et al. Evidence of estrogen receptors in normal human osteoblast-like cells. Science 1988; 241:84. Jordan VC, Jeng MH, Catherino WH, Parker CJ. The estrogenic activity of synthetic progestins used in oral contraceptives. Cancer 1993; 71(Suppl):1501. Gluntz S, Gluntz JC, Campbell-Heider N, Schaff E. Norplant use among urban minority women in the United States. Contraception 2000; 61:83.

CHAPTER 106 MORPHOLOGY OF THE NORMAL BREAST, ITS HORMONAL CONTROL, AND PATHOPHYSIOLOGY Principles and Practice of Endocrinology and Metabolism

CHAPTER 106 MORPHOLOGY OF THE NORMAL BREAST, ITS HORMONAL CONTROL, AND PATHOPHYSIOLOGY RICHARD E. BLACKWELL Morphology and Hormonal Control Comparative Anatomy of Lactation History of the Hormonal Control of the Breast Anatomy of the Mammary Gland Embryology and Histology of the Mammary Gland Nonpregnant (Inactive) Mammary Gland Mammary Development in Pregnancy Control of Lactation Maintenance of Lactation Breast Function and Aging Pathophysiology Developmental Anomalies Galactorrhea Mastodynia Breast Infections Mammary Dysplasias Tumors Often Confused with Breast Carcinoma Assessment of Breast Disease Importance of Early Diagnosis of Breast Anomalies and Diseases Complications of Breast Augmentation Breast Cancer Chapter References

MORPHOLOGY AND HORMONAL CONTROL COMPARATIVE ANATOMY OF LACTATION The constituents of milk products differ widely among species, undoubtedly reflecting differences in the nutritional requirements of the neonate and the environmental restrictions on the mother. The mammary gland is unique in the animal kingdom in that only 4200 species of mammals possess this organ. Most of these mammals (95%) belong to the subclass Eutheria; the remainder belong either to the subclass Monotremata, which contains the primitive egg-laying mammals such as the duckbill platypus, or to the Metatheria, which contains the single-order Marsupilia (i.e., kangaroos).1 HISTORY OF THE HORMONAL CONTROL OF THE BREAST Haller, in 1765, was the first to conclude that milk was derived from blood. The relation of blood and milk production was investigated by Sir Astley Cooper, who first described the early physiologic occurrence of milk letdown and lactogenesis. In the 1930s, it was shown by means of pressure monitors that milk secretion and ejection are separate events.2 In 1928, prolactin was extracted and demonstrated to be different from other known pituitary hormones.3 In the 1940s, it was proposed that during pregnancy, estrogen and progesterone promote full mammary growth while progesterone inhibits estrogen stimulation of prolactin secretion, and that at parturition, an increase in circulating prolactin and cortisol accompanied by a fall in estrogen and progesterone trigger lactation.4 Although incorrect in some aspects, this hypothesis endured for more than 20 years. It began to be challenged with the discovery that mammary growth occurs in the absence of steroid hormones in adrenalectomized and gona-dectomized rats that are recipients of pituitary mammotrophic tumor xenografts secreting prolactin, growth hormone, and adrenocorticotropic hormone.5 Subsequent studies showed that estrogen stimulates the secretion of prolactin.6 Partially inhibiting the response, progesterone suppresses prolactin secretion below baseline. It has been proposed that elevated progesterone levels during pregnancy prevent the secretion of milk and that the withdrawal of this hormone after parturition is in part responsible for lactogenesis.7 ANATOMY OF THE MAMMARY GLAND The mammary gland lies on the pectoralis fascia and musculature of the chest wall over the upper anterior rib cage (Fig. 106-1). It is surrounded by a layer of fat and encased in skin. The tissue extends into the axilla, forming the tail of Spence. The mammary gland consists of 12 to 20 glandular lobes or lobules that are connected by a ductal system. The ducts are surrounded by connective and periductal tissues, which are under hormonal control. The lactiferous ducts enlarge as they approach the nipple, which is pigmented and surrounded by the areola. The ductal tissue is lined by epithelial cells. The individual functional unit of the breast is the alveolar cell, which is surrounded by the hormonally responsive myoepithelial cells. Milk is produced at the surface of the alveolar cells and is ejected by the contraction of the myoepithelial cells under the influence of oxytocin. Fibrous septa run from the lobules into the superficial fascia. The suspensory ligaments of Cooper permit mobility of the breast.

FIGURE 106-1. Anatomy of the breast (sagittal view).

The principal blood supply of the breast comes from the lateral thoracic and internal thoracic arteries, although components have been identified from the anterior intercostal vessels. The breast is innervated chiefly by the intercostal nerves carrying both sensory and autonomic fibers. The nipple and areola are innervated by the interior ramus of the fourth intercostal nerve. Seventy-five percent of the lymphatic drainage involves axillary pathways through the pectoral and apical axillary nodes. Drainage also occurs through parasternal routes. EMBRYOLOGY AND HISTOLOGY OF THE MAMMARY GLAND The mammary gland can be identified 6 weeks after fertilization; it is derived from ectoderm. At 20 weeks' gestation, the 16 to 24 primitive lactiferous ducts invade the mesoderm. These ectodermal projections continue to branch and grow deeper into the tissue. Canalization occurs near term. Importantly, although the central lactiferous duct is present at birth, the gland does not differentiate until it receives the appropriate hormonal signals. By the time an embryo is 7 mm in length, the mammary tissue has thickened to form a ridge (known as the mammary crest or milk line) extending along the ventrolateral body wall from the axillary to the inguinal region on each side. The caudal epithelium regresses, and the crest in the thoracic region thickens further to form a primordial mammary bud by the time the embryo is 10 to 12 mm in length. These embryologic origins account for the occasional development of supernumerary

nipples and accessory breast tissue. Although mammary tissue remains relatively unresponsive until pregnancy, it is responsive to systemic hormone administration during fetal life. In the third trimester, when fetal prolactin levels increase, terminal differentiation of ductal cells occurs. This hormonal milieu accounts for the witch's milk expressible from the nipples of some normal newborn girls. After birth, these cells revert slowly to a more primitive state.8 The glands remain quiescent until the establishment of ovulatory menstrual cycles, at which time breast development proceeds in the manner described by Marshall and Tanner9 (see Chap. 91). Although the hormonal regulation of mammogenesis is unclear, estrogen in vivo appears to bring about ductal proliferation, although it has little ability to stimulate lobuloalveolar development.10 In vitro, however, estrogens do not promote mammary growth. It has been suggested that various epidermal growth factors participate in this process.11 When progesterone is administered in vivo, lobuloalveolar development occurs.12 However, the administration of estrogen and progesterone to hypophysectomized animals fails to promote mammogenesis.13 These data strongly suggest that hormones other than estrogens and progestogens play a role in mammogenesis. For instance, if the pituitary and adrenal glands are removed from oophorectomized rats, the addition of estrogen plus corticoids and growth hormone restores duct growth similar to that seen in puberty.14 NONPREGNANT (INACTIVE) MAMMARY GLAND Before pregnancy, breast lobules consist of ducts lined with epithelium and embedded in connective tissue. The preponderance of the tissues in the gland are of the connective and adipose types. There is a scant contribution from glandular parenchyma, and a few bud-like sacculations arise from the ducts. The entire gland consists predominantly of the lactiferous ducts. The breast does undergo cyclic changes associated with normal ovulation, and the premenstrual breast engorgement noted by most women is probably secondary to tissue edema and hyperemia. Epithelial proliferation is also detectable during the menstrual cycle.15 MAMMARY DEVELOPMENT IN PREGNANCY After conception, the mammary gland undergoes remarkable development. Lobuloalveolar elements differentiate during the first trimester. Both in vitro and in vivo, it is possible to induce mammary development with either placental lactogen or prolactin in the absence of steroid hormones.16 Although both placental lactogen and prolactin increase throughout pregnancy, data suggest that either of these hormones can stimulate complete mammogenesis. The role of estrogen in mammogenesis appears to be secondary, since lactation has been reported in pregnancies of women with placental sulfatase deficiency.17 Progesterone, although stimulating lobuloalveolar development, also appears to antagonize the terminal effects induced by prolactin, at least in vitro. Cortisol, which potentiates the action of prolactin on mammary differentiation, apparently is unnecessary for either ductal or alveolar proliferation.18 Insulin and other growth factors also stimulate mammogenesis.19 For example, insulin is required for the survival of postnatal mammary tissue in vitro. It is also possible that insulin-like molecules such as the insulin growth factors participate in this process. However, studies suggest that epidermal growth factor is involved in mammogenesis and, together with glucocorticoids, facilitates the accumulation of type IV collagen, a component of the basal lamina, on which epithelial cells are supported (Table 106-1).

TABLE 106-1. Hormone Regulation of the Breast

CONTROL OF LACTATION Although milk letdown occurs fairly abruptly between the second and fourth postpartum days in the human, the transition from colostrum production to mature milk secretion is gradual. This process may take up to a month and seems to coincide with a fall in plasma progesterone and a rise in prolactin levels. Twelve weeks before parturition, changes in milk composition begin20: increased production of lactose, proteins, and immunoglobulins and decreased sodium and chloride content. There is an increase in blood flow and in oxygen and glucose uptake in the breasts. There is also a marked increase in the amount of citrate at about the time of parturition. The composition of milk remains fairly stable until term, which is best exemplified by the stable production of a-lactalbumin, a milk-specific protein. At parturition, there is a marked fall in placental lactogen production, and progesterone levels reach nonpregnant levels within several days.21 Plasma estrogen falls to basal levels in 5 days, whereas prolactin decreases over 14 days.22 A fall in the progesterone level seems to be the most important event in the establishment of lactogenesis. Exogenous progesterone prevents lactose and lipid synthesis after ovariectomy in pregnant rats and in ewes.23 Furthermore, progesterone administration inhibits casein and a-lactalbumin synthesis in vitro.24 The major proteins of human milk are -lactalbumin (30% of the total protein content), lactoferrin (10–20%), casein (40%), and immunoglobulin A (IgA; 10%). Milk also contains many substances that are potentially capable of exerting biologic effects. Their physiologic role is, as yet, largely unexplored25 (Table 106-2).

TABLE 106-2. Some Bioactive Substances in Milk of Humans and Other Mammals

MAINTENANCE OF LACTATION In the human, lactation is maintained by the interaction of numerous hormones. After removal of either the pituitary or the adrenal glands from a number of animal species, milk production is terminated rapidly.26 The species dictates the type of replacement therapy required to reinstitute milk production. For instance, in rabbits and sheep, prolactin is effective alone, whereas in ruminants, milk secretion is restored by the addition of corticosteroids, thyroxine, growth hormone, and prolactin. In humans, prolactin appears to be a key hormone in the maintenance of lactation, since the administration of bromocriptine blocks lactogenesis.27 The role of thyroid hormones in lactation is unclear. Thyroidectomy inhibits lactation, and replacement therapy with thyroxine increases milk yield. It has been suggested that growth hormone and thyroxine synergize to alter milk yield, and triiodothyronine acts directly on mouse mammary tissue in vitro to increase its sensitivity to prolactin.28

Despite species differences, in humans, prolactin levels reach a peak before delivery and subsequently rebound after the initiation of lactation. This phenomenon can be inhibited by progesterone. Despite the importance of declining progesterone levels in initiating this event, lactation fails to occur with inadequate prolactin production. Prolactin production becomes attenuated over time, with the most dynamic period being 8 to 41 days postpartum. By the 63rd day, the prolactin response will be attenuated by a factor of 5, and this is maintained to ~194 days postpartum.29,30,31 and 32 SUCKLING AND MILK EJECTION The integrated baseline level of prolactin is elevated in lactating women.33 Suckling or manipulation of the breasts leads to elevation in prolactin within 40 minutes.34 In rats, this response can be mimicked by electrical stimulation of the mammary nerves. Both growth hormone and cortisol are also increased. The response appears to be greatest in the immediate postpartum period and is attenuated over 6 months. If lidocaine is applied to the nipple, thus blocking nerve conduction, the rise in prolactin levels is abolished.35 If two infants suckle simultaneously, the rise in prolactin is amplified. Along with prolactin release, suckling increases the secretion of oxytocin. After the application of a stimulus, there is an 8- to 12-hour delay before milk secretion is fully stimulated. This response seems to be correlated with the frequency and duration of vigorous suckling. There is no correlation between the amount of prolactin released and the milk yield. Suckling of the breasts increases intramammary pressure bilaterally, secondary to contraction of the myoepithelial cells in response to the octapeptide oxytocin. This contraction follows the application of stimulation to the nipple, which activates sensory receptors transmitting impulses to the spinal cord and hypothalamus. Oxytocin-producing neurons are located both in the paraventricular and supraoptic nuclei (see Chap. 25). It is estimated that ~2 ng oxytocin is released per 2- to 4-second pulse interval.36 The synthesis and release of oxytocin are rapid, because 90 minutes after injection of a radioactive amino acid into cerebrospinal fluid, radiolabeled oxytocin is released by exocytosis, and electrical pulse activity has been measured in oxytocic neurons 5 to 15 seconds before milk ejection. The response may be conditioned, since the cry of an infant or various other perceptions associated with nursing can trigger activity in the central pathways. Thus, both oxytocin and prolactin are released in response to suckling, but the patterns of release clearly are different.37 When nursing women are allowed to hold their infants but not to breast-feed, serum prolactin concentrations do not increase, despite the occurrence of the milk letdown reflex; prolactin levels rise only with nursing. The increase in prolactin with nursing is apparently sufficient to maintain lactogenesis and an adequate milk supply for the next feeding. This accounts for the ability of “wet nurses” to continue to breast-feed infants for years—even after the menopause—once lactation is established. RESOLUTION OF LACTATION Postpartum lactation can be maintained over an extended period of time by discontinuing suckling. Nevertheless, prolactin levels decrease progressively over a number of weeks despite breast-feeding. The physiologic hyperprolactinemia is achieved by altering the endogenous secretory rate of each prolactin pulse. No alteration occurs in the number of bursts of prolactin or its half-life. A large group of Australian women breast-feeding for extended periods of time demonstrated a mean of 322 days of anovulation and 289 days of amenorrhea. Fewer than 20% of the women ovulated or had menstruated by 6 months postpartum. Ovulation was delayed to a maximum of 750 days and menstruation to 698 days.38 During pregnancy, luteinizing hormone (LH) and follicle-stimulating hormone (FSH) secretion are inhibited through hypothalamic mechanisms. The exogenous opioid tone is increased during the postpartum period, and the administration of exogenous gonadotropin-releasing hormone (GnRH) pulses restores gonadotropin secretion. All of this suggests a central blockade of folliculogenesis secondary to hyperprolactinemia.39 BREAST FUNCTION AND AGING In the reproductive-aged woman, glandular tissue makes up ~20% of the breast volume. The remainder of the breast is composed of connective and adipose tissue. Breast volume changes throughout the menstrual cycle by ~20% secondarily to vascular and lymphatic congestion. Adding to the increased volume of the breast is increased mitotic activity in nonglandular tissue. Breast engorgement and change in volume result in some element of mastalgia in most women, and this combined with an increase in tactile sensitivity of the breast results in the premenstrual tenderness found in most women.40,41 With the advent of menopause and a decrease in secretion of the gonadotropins estrogen and progesterone, involution of both glandular and ductal components of the breast occurs. Without replacement estrogen therapy there is a decrease in the number and size of glandular elements and both ducts and lobules become atrophic. Over time, the volume of the breast is primarily replaced with both adipose and stromal tissues, and as with most tissues there is a loss of both contour and structure, which makes the aging breast more amenable to surveillance with mammography.42,43

PATHOPHYSIOLOGY Any disorder of the breast is viewed by the patient with alarm. Although, with the exception of carcinoma, disorders of the breast are not life threatening, any deviation from normal size and appearance must be thoroughly evaluated. Because the development of the breasts is hormone dependent and breast disorders either may have a hormonal etiology or may be misconstrued as having a hormonal cause, the endocrinologist should be familiar with the pathophysiology of these organs. DEVELOPMENTAL ANOMALIES It was not until 1969 that a system for classifying breast development was established by Marshall and Tanner9 (see Chap. 91). In addition to its obvious use in evaluating the adequacy of breast development, this classification can be used to determine the presence of pathology. CONGENITAL ANOMALIES Congenital anomalies of the breast itself are uncommon; however, one frequently sees anomalies of development. Even so, amastia, congenital absence of the breast; athelia, congenital absence of the nipple; polymastia, multiple breasts; polythelia, multiple nipples; or some combination, occur in 1% to 2% of the population and may have a familial tendency.44 If treatment is deemed necessary, surgical augmentation or excision is recommended (Fig. 106-2).

FIGURE 106-2. Patient with Poland syndrome (aplasia of the pectoralis muscles, rib deformities, webbed fingers, and radial nerve aplasia, often associated with unilateral amastia) after partial reconstruction. The areola and nipple remain to be reconstructed.

Young patients may present with the problem of premature thelarche (see Chap. 92). Many would define this condition as breast development beginning before age 8. Affected individuals may have either bilateral or unilateral development. The disorder may be differentiated from precocious puberty by the finding of prepubertal serum levels of gonadotropin and estrogen. Precocious thelarche is self-limited and demands no therapy other than assurance, once complete or incomplete isosexual precocity has been ruled out. BREAST ASYMMETRY

Breast asymmetry is fairly common (Fig. 106-3) and presumably is secondary to a difference in end-organ sensitivity to estrogen and progesterone. Occasionally, full symmetry is obtained in adolescents by the administration of an oral contraceptive agent, although either augmentation or reduction mammoplasty may be required if severe asymmetry does not resolve. Most patients with breast asymmetry do not require therapy other than reassurance that this is simply a variation of normal.

FIGURE 106-3. Normal mammogram showing breast asymmetry. Contemporary mammography uses a low kilovoltage–high milliamperage technique; the dose generally ranges from 20 to 30 kVp. Limitation of breast motion with a compression device allows decreased milliamperage and increased magnification to produce a more uniform image density. The breast image varies with the age of the patient. Virginal breasts are small and have near-consistent fibroglandular tissue. Breasts of a reproductive-aged individual, as shown in this figure, vary between well-developed fibroglandular and adipose tissues. Ducts and fibrous tissue are difficult to differentiate and are often found together. A wide variation is noted in the postmenopausal period, but there generally is increased fat content, making trabeculae, subareolar ducts, and veins easily visible. The atrophic breast shows a ground-glass homogeneity with prominent residual trabeculae.

HYPOPLASIA OF THE BREASTS Perhaps one of the most common disorders of the breast involves hypoplasia. These individuals may simply have small breasts secondary to a transient delay in puberty or may have a genetic tendency toward hypoplasia, with other siblings having similar problems. Such breasts show a physiologic response to pregnancy, and lactation can follow. Not uncommonly, because of social pressure to have “normal-sized” breasts, augmentation is often sought by affected individuals (Fig. 106-4). Such augmentation will not interfere with lactation or breast-feeding but does increase the difficulty of self-examination and surveillance for breast malignancies. Breast hypoplasia is sometimes found in patients with severe anorexia nervosa and other variants of psychogenic amenorrhea associated with decreased body weight or extremes of exercise (see Chap. 128); such individuals have an altered fat/lean mass body ratio, which generally renders them hypogonadotropic and hypoestrogenic. The removal of estrogen-progesterone stimulation leads to breast atrophy. Reconstructive therapy is contraindicated in this group; correction of nutritional requirements is the therapy of choice, although this must often be accompanied by psychotherapy in patients with emotionally related weight loss.

FIGURE 106-4. Breast hypoplasia (top) and same patient after augmentation mammoplasty (bottom).

Breast hypoplasia also occurs in the female pseudohermaphroditism of congenital adrenal hyperplasia (see Chap. 77) and in Turner syndrome (see Chap. 90). The early institution of corticosteroid therapy will greatly benefit the former patient; the latter should be treated at the appropriate time with cyclical estrogen and progesterone. BREAST HYPERTROPHY Breast hypertrophy or macromastia is encountered commonly in both adolescents and adults. The breasts may be either symmetric or asymmetric. The patient frequently presents seeking advice on reduction mammoplasty (Fig. 106-5), perhaps because of chest wall pain secondary to the weight of the breasts, difficulty in finding clothes that fit the upper body, and difficulty with her self-image. Frequently, young women are under intense sexual pressure and are often embarrassed by peers during gymnasium classes or when wearing swimming suits. As an alternative to surgical correction, danazol has been tried.45 Unfortunately, this drug has many side effects and definitely is not acceptable for long-term therapy.

FIGURE 106-5. Patient before (top left and right) and after (bottom left and right) reduction mammoplasty.

NIPPLE INVERSION Nipple inversion is common but rarely presents as a complaint to the clinician. Cosmetic repair can be performed, but breast-feeding is difficult after such procedures. GALACTORRHEA Galactorrhea, the inappropriate production and secretion of milk, may be intermittent or continuous, bilateral or unilateral, free flowing or expressible. By definition, fat droplets must be present on microscopic examination for a breast secretion to be considered milk and as evidence of galactorrhea. Galactorrhea is frequently associated with hyperprolactinemia (see Chap. 13),46 which should be sought by repeatedly measuring serum prolactin levels, remembering that prolactin is a

stress-related hormone whose secretion may be increased by breast examination and stimulation, acute exercise, food intake (particularly protein), and sleep.47 Although the differential diagnosis of hyperprolactinemia is extensive, the common causes of this condition are prolactinoma, primary hypothyroidism, and drug intake. Galactorrhea should be evaluated by the measurement of multiple serum prolactin levels, thyroxine, and thyroid-stimulating hormone and by radiographic or magnetic resonance imaging studies of the pituitary. The prolactin level at which radiographic surveillance is begun is debated; however, computed tomography or magnetic resonance imaging should be done if basal prolactin levels exceed 100 ng/mL. Galactorrhea and its treatment are considered in more detail in Chapter 13, Chapter 21, Chapter 22 and Chapter 23. MASTODYNIA Mastodynia, painful engorgement of the breasts, is usually cyclic, becoming worse before menstruation.48 Although most women describe mastodynia at some times, they require no therapy. However, some patients require cyclic analgesics or nonsteroidal antiinflammatory drugs. Occasionally mastodynia is a complaint of women experiencing the premenstrual syndrome; some affected patients will sporadically obtain some relief with nonspecific therapy, as discussed in Chapter 99. Mastodynia may also be treated effectively with danazol, but the side effects of the drug mandate its use only in severe cases. In addition, a second generation of drugs, the GnRH analogs, have been used to induce hypogonadotropism and hypoestrogenism, thus treating disorders such as endometriosis, fibroids, hirsutism, and premenstrual syndrome. Treatment with these agents, either on a daily or monthly basis, will result in profound hypogonadism, breast atrophy, and relief of mastodynia. These drugs are not approved by the Food and Drug Administration for this purpose, and therapy beyond 6 months results in reversible bone demineralization. To compensate for this loss in other disorders, estrogen “add-back” therapy, cotreatment with progestogens, and the use of variable-dose estrogen-progestogen overlapping protocols have been used to counter this and other side effects. BREAST INFECTIONS Breast infections are often confused with galactorrhea but require therapy with appropriate systemic antibiotics. Patients present with unilateral or bilateral breast drainage, which, when examined by microscopy, fails to show fat globules. Gram stain frequently will reveal Staphylococcus, Streptococcus, or Escherichia coli. If the discharge has a greenish tint, Pseudomonas should be suspected. If the discharge is accompanied by abscess formation, drainage as well as antibiotics should be used. The galactocele, or retention cyst, which usually occurs after cessation of lactation, is caused by duct obstruction and can masquerade as mastitis. These lesions usually lie below the areola and are often tender to palpation. Such cysts occasionally can be emptied by properly placed pressure; however, drainage frequently must be carried out. Untreated galactoceles may be sites of future sepsis and can calcify and become confused with malignant lesions radiologically. The drainage from a galactocele may range from milky to clear to yellow-green purulent-appearing material; however, these lesions are usually sterile. MAMMARY DYSPLASIAS Mammary dysplasia is perhaps the most common lesion of the female breast (Fig. 106-6). Historically, mammary dysplasias have carried the label fibrocystic disease, chronic lobular hyperplasia, cystic hyperplasia, or chronic cystic mastitis.49 The term cystic mastitis should be discarded, since inflammation is not present in this disorder.

FIGURE 106-6. Composite mammogram showing bilateral fibrocystic disease. Note multiple large cysts in body of breast (arrows).

Mammary dysplasia may be unilateral or bilateral and most frequently occurs in the upper outer quadrants. The disorder tends to be exacerbated in the premenstrual period. Patients usually complain of pain or lumps in the breasts. The breasts may be tender in many locations; axillary adenopathy is generally not found. Palpable breast lumps are usually cystic and tense; they shrink after menstruation. The natural history of the disease varies; however, it tends to resolve at menopause. Mammary dysplasia may be accompanied by a nipple discharge, which may be clear or bloody, in up to 15% of patients. The disorder may be confused with carcinoma, and a Papanicolaou smear of the discharge, mammography, and perhaps needle aspiration may be necessary to rule out a malignancy. At the time of aspiration, one may evacuate a cyst filled with a dirty gray-green fluid. The management of this problem includes frequent breast examination, periodic mammography, use of a brassiere with good support, perhaps avoiding methylxan-thines and chocolate, and, in extreme cases, the use of danazol therapy in daily doses of 100 to 800 mg in two divided doses. Although bromocriptine suppresses prolactin secretion, it has not been effective in the management of fibrocystic disease. It does decrease cyclic mastodynia, however. Likewise, GnRH analogs can be used in extreme cases. The dose of leuprolide acetate used to induce hypogonadism is 3.75 mg given intra-muscularly every month. TUMORS OFTEN CONFUSED WITH BREAST CARCINOMA A few lesions, such as fat necrosis, adenosis (especially the sclerosing type), intraductal papilloma (including juvenile papillomatosis), and fibroadenoma (including cystosarcoma phylloides), are often confused with carcinoma of the breast.50 FAT NECROSIS Fat necrosis may present as a hard lump that may be tender; it rarely enlarges. Skin retraction may be seen, along with irregularity of the edges; fine, stippled calcifications may be present on mammography. Approximately half the patients have a history of trauma. Excisional biopsy is the treatment of choice. At biopsy, one may note hemorrhage into the fatty tissue. SCLEROSING ADENOSIS Breast adenosis may be confused with carcinoma, particularly if sclerosing adenosis is present. This latter condition is characterized by the proliferation of ductal tissue, producing a palpable lesion. These lesions are common in younger women, especially in the third and fourth decades of life; they are rarely seen postmenopausally. Grossly, breast carcinoma is often firm and gritty to palpation, whereas adenosis is usually rubbery. Microscopically, one sees a nodule or whorl pattern. In sclerosing adenomatous adenosis, one also sees circumscription of the lesion, central attenuation of ductal caliber, an organoid arrangement of the ducts, mild epithelial tissue surrounding the ducts, and absence of intraductal epithelial bridging (Fig. 106-7).

FIGURE 106-7. Light mammograph of tissue sample showing sclerosing adenosis.

INTRADUCTAL PAPILLOMA Intraductal papilloma is a benign lesion of the lactiferous duct walls that occurs centrally beneath the areola in 75% of cases.51 Such lesions present as pain or bloody discharge. They are soft, small masses that are difficult to palpate. Indeed, if the patient presents with a small palpable mass associated with a bloody nipple discharge, there is a 75% chance that an intraductal papilloma will be found. If no mass can be palpated, Paget disease of the nipple or a carcinoma must be considered. Intraductal papilloma is not premalignant and is best managed by excision of the duct by wedge resection. Although intraductal papillomas generally occur in women in the late childbearing years, they may present in the adolescent. In these younger patients, such lesions are generally found at the periphery of the breast; multiple ducts may be involved, and cystic dilation is noted. These lesions have been called “Swiss cheese disease” or juvenile papillomatosis. The treatment of juvenile papillomatosis involves excisional biopsy (Fig. 106-8).

FIGURE 106-8. Light micrograph of tissue sample showing intraductal papilloma.

FIBROADENOMA One of the most common benign neoplasms of the adolescent and adult breast is the fibroadenoma52,53 (Fig. 106-9). These tumors may be small, firm nodules or large, rapidly growing masses that are multiple 20% of the time. They are more common in black than in white women. They may be painful. The fibroadenoma may be hormonally responsive; rapid growth occurs during pregnancy and lactation. These tumors are best treated by excisional biopsy. Rare variants of the fibroadenoma (cystosarcoma phylloides, also known as the giant fibroadenoma) have been described. Although these tumors are generally benign, a few have true sarcomatous potential.

FIGURE 106-9. Mammogram showing fibroadenoma (arrow).

ASSESSMENT OF BREAST DISEASE SELF-EXAMINATION Breast self-examination is still one of the most important methods for the diagnosis of diseases of the breast, either benign or malignant. A poll conducted for the American Cancer Society found that the physician plays a pivotal role in encouraging patients to practice breast self-examination.53 It was noted that when patients received personal instruction from their physicians, 92% continued to practice breast self-examination regularly. Once taught self-examination, many patients can detect nodules in their own breasts before they are palpable by a skilled physician. Care must be taken to convince each patient that her breasts are not homogeneous but rather contain various structures and different degrees of nodularity, thickening, and small lumps. The texture of the breast changes throughout a woman's life and during the menstrual cycle (Table 106-3).

TABLE 106-3. Breast Self-Examination

It is suggested that breast self-examination be practiced each month, preferably just after the menstrual period.54 The examination should consist of inspection of the size, shape, and skin color and for puckering, dimpling, retraction of any of the surface, and any nipple discharge. The patient should look at her breasts by placing her hands on her hips and flexing the shoulders forward, then raising the hands behind her head. Breasts may be asymmetric in size, but asymmetry in movement is an indication of pathology. Next, palpation of the breast should be carried out in both the sitting and supine positions. Each quadrant should be palpated systematically, including the nipple and areolar area. Special attention should be paid to the upper outer quadrant and the axilla, because this is the most frequent site of breast carcinoma. Examination can be carried out with the fingertips or with a rotary motion as suggested by the American Cancer Society. An annual physical examination also should be performed by a physician skilled in the diagnosis of breast disease. There are different variations and techniques of breast examination, but a consistent systematic examination is central to all. MAMMOGRAPHY Breast imaging dates back to 1913. Mammography has been refined subsequently such that improved image, clarity, and contrast have increased its accuracy to well more than 90%.55,56,57,58 and 59 Despite widespread publicity urging the use of mammography and mass screening, breast cancer still strikes many women and is a common cause of death.60 Although most recommendations relating to breast screening are aimed at women aged 35 or older, the incidence of breast cancer in women younger than 35 is not zero. An interdisciplinary task force in the United States has made the following recommendations61: First, mammography should be a part of clinical examination for breast disease and does not substitute for any part. Second, the mainstay of detection of breast disease remains self-examination and physician consultation. Third, women are candidates for mammography at any age if they have masses or nipple discharge, masses felt by the patient but not confirmed by a physician,62 previous surgical alteration of the breast by augmentation procedures or implants, contralateral disease, previous breast cancer, history of breast cancer in a mother or sister, first pregnancy after age 30, and abnormal patterns in baseline mammography suggestive of increased risk63 (Table 106-4). Women older than age 50 should receive regular breast examinations, including mammography, as determined by the physician. Baseline mammography should be performed on all women at some time between the ages of 35 and 50.

TABLE 106-4. Principal Breast Cancer Risk Factors

BREAST IMAGING WITHOUT RADIATION Because of concerns about the hazards of multiple radiation exposures for mammography, several other modes have been introduced in the field of breast imaging. Thermography. Thermography maps focal variations in skin temperature by various techniques.64 Invasive breast cancers produce higher skin temperature, and thermography is accurate for detecting advanced disease. However, it is ineffective in the diagnosis of nonpalpable cancers, detecting only approximately one-half the cases that can be discovered by mammography. Thus it is not an acceptable modality for population screening. Ultrasound Mammography. Breast ultrasound mammography can produce images in conjunction with immersion of the glands in a water bath65 (Fig. 106-10). However, it has poor resolution; it will not image structures smaller than 1 mm or identify microcalcification. Ultrasonography seems to be most successfully applied to the diagnosis of breast disease in younger patients and is thought to be complementary to mammography.

FIGURE 106-10. Breast sonogram showing fibrocystic disease. Multiple cysts are outlined by the small white dots.

IMPORTANCE OF EARLY DIAGNOSIS OF BREAST ANOMALIES AND DISEASES The endocrinologist should encourage early diagnosis of congenital and acquired breast disorders. Fear of breast disease and subsequent surgical mutilation often causes the patient to defer evaluation, often worsening the outcome. In particular, the mortality from breast cancer remains high, in part because of such delays.66 It thus should be emphasized that both benign and malignant diseases of the breast are diagnosable at the most treatable stage by self-examination, early physician consultation, radiographic study, and, sometimes, other methods, and that the treatment of developmental anomalies and of benign and malignant breast disease may be hormonal as well as surgical (see Chap. 224). Moreover, even when surgery is mandated, early diagnosis plus available plastic surgery procedures can produce both cure and aesthetically satisfactory results. COMPLICATIONS OF BREAST AUGMENTATION Major developments in breast augmentation such as silicone implants, mucocutaneous flaps, and autogenous tissue transfers have occurred within the past 20 years. When silicone implants were first introduced in 1964,67 they were thought to be biocompatible products. However, it has been postulated that biomaterials such as silicone might behave like other immunogenic substances. Antibodies to silicone were described in sera of two patients who had severe chronic inflammatory reactions around implanted Silastic ventriculoperitoneal shunt tubing.68 A study evaluating the sera of 79 women with breast implants who experienced a wide variety of problems found that half had antibody levels >2 standard deviations above the control group without implants.69 Subsequent studies have been contradictory. Access to silicone implants has been restricted, and the Food and Drug Administration approved a protocol to evaluate silicone implants in women whose saline-filled implants are

considered medically unsatisfactory.70 Augmentation mammoplasty creates a second problem, that is, the effect of capsular contracture on the quality of mammography. Moderate contracture has been predicted to result in a 50% reduction in the quality of visualization.71,72 These factors need to be considered in advising patients about augmentation mammoplasty. BREAST CANCER ETIOPATHOLOGY Despite the investments that have been made in the diagnosis and treatment of breast cancer over the last two decades, only modest headway has been made in managing this disease. Currently, women in the United States have a 1 in 8 risk, which is twice that found in 1940. In one study, at age 25, a woman had a 1 in 19,608 risk of developing breast cancer; by age 40 this had increased to 1 in 217; by age 70, 1 in 14; and by age 85, 1 in 9.73 Family history seems to play a major role in the development of breast cancer, with a two- to three-fold increased risk in the incidence of the disease being found in women who have female relatives with the disease. For instance, the patient with an affected mother or sister has a 2.3 relative risk and an affected aunt 1.5 relative risk, and a 14% incidence when both mother and sister are affected. Hereditary forms of breast cancer make up ~8% of the disease population, and those women who have a strong family history tend to develop the disease at a younger age.74 In this respect, perhaps the most exciting event to have occurred in breast cancer research is the identification of genes predisposing to breast cancer. BRCA-1 and BRCA-2 together account for approximately two-thirds of familial breast cancer or roughly 5% of all cases.75,76,77 and 78 It also appears that BRCA-1 is associated with the predisposition of ovarian cancer. BRCA-1 is located on a locus on chromosome 17Q, and an analysis of 200 families has shown that BRCA-1 is responsible for multiple cases of breast cancer in ~33% of families but more than 80% of families in which there is both breast cancer and epithelial ovarian cancer. Women who inherit the BRCA genes have a 60% risk of acquiring breast cancer by age 50, and a 90% overall lifetime risk. BRCA-2 lies within a 6-centimorgan interval on chromosome 13Q12.13 centered on D13S260. The loss of this gene may also result in elimination of suppressor function. The discovery of these genes presents the possibility for genetic testing, which remains controversial at present. Cigarettes, coffee, alcohol, and diet may play a role in the development of breast cancer. Tobacco-related cancers appear in the lung, esophagus, oral cavity, pancreas, kidney, bladder, and breast. Therefore, smoking remains the chief preventable cause of death and illness in the United States. It is responsible for ~70% of all deaths; however, although smoking decreased from 40% in 1965 to 29% in 1987, more than 5 million Americans continue to smoke. The incidence of smoking in women has risen at an alarming rate, and this parallels the increase in lung cancer found in women. Further, there has been an abrupt increase in smoking in girls aged 11 through 17.79 Methylxanthine-containing compounds have been implicated as a causative factor in the development of fibrocystic disease of the breast and cancer. The Boston Collaborative Drug Surveillance Program showed an increased risk in women who drank between one and three cups of coffee or tea per day.80 Several studies have evaluated the role of alcohol and its association with an increased risk of breast cancer. Women who consume more than three drinks per day have been reported to have a 40% increase in the risk of breast cancer.81,82 Dietary fat intake has been thought to be linked to breast cancer, but this relationship is controversial.82a It was noted that postmenopausal women in the United States are at a much higher risk for breast cancer than are Asian women. This does not appear to be a geographic phenomenon as movement of Asian women to either the Hawaiian Islands or Pacific Coast seems to eradicate the difference in incidence. The suppression, however, usually requires one to two generations to demonstrate significance. Other populations with high fat intake but relatively low risk of cancer such as seen in Greece or Spain use monounsaturated fats composed primarily of oleic acid. Likewise, fish oil which is rich in omega-3 fatty acids has been associated with a lower incidence of breast cancer in countries ranging from Greenland to Japan.83 Steroid hormones are thought to affect the expression of breast cancer. For instance, a woman who has a child at the age of 18 has approximately one-third the risk of a woman who delivers after age 35. However, pregnancy must occur before age 30 to be protective, but, in fact, a woman who gives birth after age 35 appears to be at greater risk than a woman who has never been pregnant. There is also a 70% reduction of risk in the incidence of breast cancer in women who undergo oophorectomy before the age of 35. There also appears to be a small increased risk in patients who experience early menarche as well as late menopause. It has been suggested that the endocrine milieu influences the susceptibility of the breast to environmental carcinogens.84 This is the so-called estrogen window hypothesis, which suggests that an unopposed estrogen stimulation at certain periods of life favors tumor induction. The longer the unopposed estrogen stimulation acts on the breast, the greater is the risk factor. Perhaps pregnancy, a high progesterone state, closes the window, because progesterone is known to down-regulate the estrogen receptors in the endometrium and is protective against the development of endometrial cancer. Although the data appear to be inconclusive at present, one might speculate that a similar mechanism may be achieved at the level of the breast. Various chemical agents have been implicated in a decrease or increase in breast cancer. Estrogens of all types and their analogs may stimulate tumorigenesis. Progestogens, while regulating estrogen expression, can induce significant mitosis of both epithelial and stromal components.85 Historically, birth control pills have been evaluated using a variety of different study designs, and many, but not all reports have shown no increased risk of breast cancer.86,87 and 88 GnRH analogs decrease estrogen production and therefore are thought to be protective against breast cancer. Tamoxifen is a weak estrogen agonist that antagonizes the biologic effect of 17b-estradiol. It is now used for the treatment of breast cancer in both menopausal and perimenopausal women, and current data suggest that this drug may in fact retard the expression of breast cancer. Raloxifene, a selective estrogen receptor modulator (SERM), used in hormonal replacement therapy, has also been shown to reduce the incidence of breast cancer and is given as a hormonal replacement therapy in menopause.88a POSTOPERATIVE REHABILITATION OF THE PATIENT WITH BREAST CANCER Chapter 224 discusses the current therapy of breast cancer. Thirty years ago, radical mastectomy was considered by many surgeons to be the treatment of choice for resectable breast cancer. Reconstructive options were few, and required 3 to 4 stage procedures to create an adequate breast replacement. Usually, women were required to wear external breast prostheses. This resulted in surgical patients feeling disfigured, having a poor body image, lower self-esteem, and diminished feelings of sexual attractiveness and of femininity. Now, reconstructive techniques can be carried out immediately, or can be delayed. There has been a trend toward immediate reconstruction, as this tends to reduce the degree of psychological morbidity experienced by the patient, and the reconstructed breast is integrated into the body image. Further, the integrity of the soft tissue that envelops the breast is intact at the time of surgery; there is no fibrosis or contraction of the tissue, and a plastic surgeon can be involved in the surgery and the reconstruction to give the best cosmetic result, whether implant or autologous tissue is used.89,90,91 and 92 Following reconstruction, breast cancer patients are usually examined every 3 months for the first 5 years, at every 6 months for the next 5 years, and yearly thereafter. A metastatic survey including a complete blood cell count, blood chemistry, chest x-ray, and mammogram should be performed routinely in patients with stage I or II disease. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Mepham T. Physiological aspects of lactation. In: McPhan T, ed. Biochemistry of lactation. New York: Elsevier, 1983:3. Peterson LV. Lactation. Physiol Rev 1944; 24:340. Riddle O, Bates R, Dykshorn S. The preparation, identification and assay of prolactin-A hormone of the anterior pituitary. Am J Physiol 1933; 105:191. Meites J, Turner C. Studies concerning the mechanism controlling the initiation of lactation at parturition: II. Why lactation is not initiated during pregnancy. Endocrinology 1942; 30:719. Clifton K, Furth J. Ductoalveolar growth in mammary glands of adrenogonadectomized male rats bearing mammotropic pituitary tumors. Endocrinology 1960; 66:893. Chen C, Meites J. Effects of estrogen and progesterone on serum and pituitary levels in ovariectomized rats. Endocrinology 1970; 86:503. Kuhn N. Progesterone withdrawal as the lactogenic trigger in the rat. J Endocrinol 1969; 44:39. McKiernan J, Coyne J, Canglone S. Histology of breast development in early life. Arch Dis Child 1988; 63:136. Marshall W, Tanner J. Variations in pattern of pubertal changes in girls. Arch Dis Child 1969; 44:291. Cowie AT. Backward glances. In: Yokoyama A, Mizuno H, Nagasawa II, eds. Physiology of mammary glands. Baltimore: University Park Press, 1978:43. Tonelli G, Sorof S. Epidermal growth factor: requirement for development of cultured mammary glands. Nature 1980; 285:250. Ichinose R, Nandi S. Influence of hormones on lobuloalveolar differentiation of mouse mammary glands in vitro. J Endocrinol 1966; 35:331. Cowie A, Tindal J, Yokoyama A. The induction of mammary growth in the hypophysectomized goat. J Endocrinol 1966; 34:184. Lyons WR. Hormonal synergism in mammary growth. Proc R Soc Biol 1958; 149:303. Going JJ, Anderson TJ, Battersby S, et al. Proliferative and secretory activity in human breast during natural and artificial menstrual cycles. Am J Pathol 1988; 130:193. Talwalker P, Meites T. Mammary lobuloalveolar growth induced by anterior pituitary hormones in adreno-ovariectomized-hypophysectomized rats. Proc Soc Exp Biol Med 1961; 107:880. France J, Seddon R, Liggins G. A study of a pregnancy with low estrogen production due to placental sulfatase deficiency. J Clin Endocrinol Metab 1973; 36:19. Topper Y, Freeman C. Multiple hormone interactions in the developmental biology of the mammary gland. Physiol Rev 1980; 60:1049. Elias J. Effect of insulin and cortisol on organ cultures of adult mouse mammary gland. Proc Soc Exp Biol Med 1959; 101:500.

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.

Fleet I, Goode J, Hamon M, et al. Secretory activity of goat mammary glands during pregnancy and the onset of lactation. J Physiol 1975; 251:763. Weiss G, Facog E, O'Byrne E, et al. Secretion of progesterone and relaxin by the human corpeus luteum at midpregnancy and at term. Obstet Gynecol 1977; 50:679. Martin R, Glass M, Wilson G, Woods K. Human a-lactalbumin and hormonal factors in pregnancy and lactation. Clin Endocrinol (Oxf) 1980; 13:223. Hartmann P, Trevethan P, Shelton J. Progesterone and oestrogen and the initiation of lactation in ewes. J Endocrinol 1973; 59:249. Bruce J, Ramirez V. Site of action of the inhibitory effect of estrogen upon lactation. Neuroendocrinology 1978; 6:19. Grosvenor CE, Picciano MF, Baumrucker CR. Hormones and growth factors in milk. Endocr Rev 1993; 14:710. Hearn J. Pituitary inhibition of pregnancy. Nature 1973; 241:207. Brun del Re R, del Pozo E, deGrandi P, et al. Prolactin inhibition and suppression of puerperal lactation by a Brergocriptine (CB 154): a comparison with estrogen. Obstet Gynecol 1973; 41:884. Vonderhaar BK. Studies on the mechanism by which thyroid hormones enhance a-lactalbumin activity in explants from mouse mammary glands. Endocrinology 1977; 150:1423. Riggs LA, Yen SSC. Multiphasic prolactin secretion during parturition in human subjects. Am J Obstet Gynecol 1977; 128:215. Kuhn NJ. Lactogenesis. The search for trigger mechanisms in different species. Symp Zool Soc Lond 1977; 41:165. Noel GL, Suh HK, Frantz AG. Prolactin release during nursing and breast stimulation in postpartum and non-postpartum subjects. J Clin Encocrinol Metab 1974; 38:413. Nunley WL, Urban RT, Kitchin JD, et al. Dynamics of pulsatile prolactin release during the postpartum lactational period. J Clin Endocrinol Metab 1991; 72:287. Gross B, Eastman C, Bowen C, McEldruff A. Integrated concentration of prolactin in breast-feeding mothers. Aust NZ J Obstet Gynaecol 1979; 19:150. Howie P, McNeilly A, McArdle T, et al. The relationship between suckling-induced prolactin response and lactogenesis. J Clin Endocrinol Metab 1980; 50:670. Tyson J. Nursing and prolactin secretion: principal determinants in the mediation of puerperal infertility. In: Crosignani P, Robyn C, eds. Prolactin and human reproduction. New York: Academic Press, 1977:97. Lincoln O, Wakerley J. Electrophysiological evidence for the activation of supraoptic neuronics during the release of oxytocin. J Physiol (Lond) 1974; 242:533. Brandts C, Rozenberg S, Meuris S. Advances in physiology of human lactation. In: Angeli A, Bradlow H, Dogliotti L, eds. Endocrinology of the breast. Ann NY Acad Sci 1986; 464:66. Lewis PR, Brown JB, Renfree MB, et al. The resumption of ovulation and menstruation in a well-nourished population of women breast feeding for an extended period of time. Fertil Steril 1991; 55:529. Matsuzaki T, Azuma K, Irabara M, et al. Mechanism of anovulation in hyperprolactinemic amenorrhea determined by pulsatile gonadotropin-releasing hormone injection combined with human chorionic gonadotropin. Fertil Steril 1994; 62:2254. Milligan D, Drife JO, Short RV. Changes in breast volume during normal menstrual cycle and after oral contraceptives. Br Med J 1975; 4:494. Robinson JE, Short RV. Changes in breast sensitivity at puberty, during the menstrual cycle, and at parturition. Br Med J 1977; 1:1188. Cowie AT, Forsyth JA, Hart JC. Hormonal control of lactation. Berlin: Springer-Verlag, 1980. Andolina V, Lille S, Wilson KM, eds. Mammographic imaging: a practical guide. Philadelphia: Lippincott, 1992. Pellegrini J, Wagner R. Polythelia and associated conditions. Am Fam Physician 1983; 28:192. Aksu MF, Tzingounis VA, Greenblatt RB. Treatment of benign breast disease with danazol: a follow-up report. J Reprod Med 1978; 31:181. Blackwell RE. Diagnosis and treatment of hyperprolactinemic syndromes. In: Wynn RM, ed. Obstetrics and gynecology annual 1985. Norwalk, CT: Appleton-Century-Crofts, 1985:305. Blackwell RE. Diagnosis and management of prolactinomas. Fertil Steril 1985; 43:5. Pilnik S. Clinical diagnosis of benign breast disease. J Reprod Med 1979; 22:277. Love S, Gelman R, Silen W. Fibrocystic “disease” of the breast: a nondisease? N Engl J Med 1982; 307:1010. Oberman HA. Benign breast lesions confused with carcinoma. In: McDiuitt RW, Oberman HA, Ozzello L, Kaufman N, eds. International Academy of Pathology monograph: the breast. Baltimore: Williams & Wilkins, 1984:1. Haagensen C, Stout A, Phillips J. Neoplasms of the breast: I. Benign intraductal papilloma. Am J Surg 1951; 133:18. Hertel B, Zaloudek C, Kempson R. Breast adenomas. Cancer 1976; 37:2891. Egan R. Breast imaging, 3rd ed. Baltimore: University Park Press, 1984:5. Wilson RW. The breast. In: Sabiston D, ed. Davis-Christopher textbook of surgery, 10th ed. Philadelphia: WB Saunders, 1972:573. Colman M, Mattheiem W. Imaging techniques in breast cancer: workshop report. Eur J Cancer Clin Oncol 1988; 24:69. Maisey MN. Imaging techniques in breast cancer: what is new? What is useful? A review. Eur J Cancer Clin Oncol 1988; 24:61. Bassett LW, Gold RH. The evolution of mammography. AJR Am J Roentgenol 1988; 150:493. Salomon A. Beitrage zur Pathologie und Klinik des Mammarkarzinome. Arch Klin Chir 1913; 101:573. Egan R. Mammography. Springfield, IL: Charles C Thomas Publisher, 1964:1. Pietsch J. Breast disorders. In: Lavery J, Sanfilippo J, eds. Pediatric and adolescent obstetrics and gynecology. New York: Springer-Verlag, 1985:103. Executive Board of the American Academy of Obstetrics and Gynecology. ACOG statement of policy mammography statement. Chicago: American College of Obstetrics and Gynecology, 1979:1. Edeiken S. Mammography and palpable cancer of the breast. Cancer 1988; 61:263. Solin LJ, Legoretta A, Schultz DJ, et al. The importance of mammographic screening relative to the treatment of women with carcinoma of the breast. Arch Intern Med 1994; 154:745. Gauterie M, Gross C. Breast thermography and cancer risk prediction. Cancer 1980; 45:51. Wild J. Review of the ultrasonic examination of the breast. In: Jellins J, Kobayashi T, eds. Ultrasonic examination of the breast. New York: John Wiley and Sons, 1983:21. Carlson RW, Stockdale FE. The clinical biology of breast cancer. Annu Rev Med 1988; 39:453. Cronin TD, Gerow F. Augmentation mammoplasty: a new “natural feel” prosthesis. In: Transactions of the Third International Congress of Plastic Surgeons. Amsterdam: Excerpta Medica, 1964. Goldblum RM, Pelley RP, O'Donell AA, et al. Antibodies to silicone elastomers and reactions to ventriculoperitoneal shunts. Lancet 1992; 340:510. Heggers JP, Goldblum RM, Pyron MT, et al. Immunologic responses to silicone implants: fact or fiction? Plast Surg Forum 1990; 8:13. Randall T. First clinical study of breast implants launched. JAMA 1992; 268:1822. Douglas KP, Bluth EI, Sauter ER, et al. Roentgenographic evaluation of the augmented breast. South Med J 1991; 64:49. Handel N, Silverstein MJ, Gamagami P. Factors affecting mammographic visualization of the breast after augmentation mammaplasty. JAMA 1992; 268:1913. Davis DL, Dinse GE, Hoel DG. Decreasing cardiovascular disease and increasing cancer among whites in the United States from 1973 through 1978. JAMA 1994;271:431. Colton T, Greenberg ER, Noller K, et al. Breast cancer in mothers prescribed diethylstilbestrol in pregnancy. JAMA 1993; 269:2096. Futreal PA, Liu Q, Shattuck-Eldens D, et al. BRCA1 mutations in primary breast and ovarian carcinomas. Science 1994; 266:120. Miki Y, Swensen J, Shattuck-Eldens D, et al. A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science 1994; 266:66. Nowak R. Breast cancer gene offers surprises. Science 1994; 265:1796. Wooster R, Neuhausen SL, Mangion J, et al. Localization of a breast cancer susceptibility gene, BRCA2 to chromosome 13q12–13. Science 1996; 265:2088. Rosenhert L, Schwingi PA. Breast cancer and cigarette smoking. N Engl J Med 1984; 310:92. Welsch CW. Caffeine and the development of a normal and neoplastic mammary gland. Proc Soc Exp Biol 1994; 207:1. Schotzkin A, Jones DY, Hoover RN, et al. Alcohol consumption and breast cancer in the epidemiologic followup study of the first national health and nutrition examination survey. N Engl J Med 1987; 316:1169. Willett WC, Colditz G, Stampler MJ, et al. A prospective study of alcohol intake and risk of breast cancer. Am J Epidemiol 1986; 124:540.

82a. Velie E, Kulldorff M, Schaiver C, et al. Dietary fat, fat sib types, and breast cancer in postmenopausal women: a prospective cohort sudy. J Natl Cancer Inst 2000; 92:833. 83. 84. 85. 86. 87.

Bland KI. Risk factors as an indicator for breast cancer screening in asymptomatic patients. Maturitas 1987; 9:135. Korenmann SC. Estrogen window hypothesis of the etiology of breast cancer. Lancet 1980; 1:700. Li JJ, Li SA. Estrogen carcinogenesis in hamster tissues: a critical review. Endocrinol Rev Monograph 1. Endocrine Aspects of Cancer 1993; 1:86. Ramcharan S, Pellegrin FA, Ray RM, Hau J-P. The Walnut Creek Contraceptive Drug Study: a prospective study of the side-effects of oral contraceptives. J Reprod Med 1980; 25:366. Royal College of General Practitioners Oral Contraceptive Study. Further analysis of mortality in oral contraceptive users. Lancet 1981; 1:541.

88a. Minton SE. Chemoprevention of breast cancer in the older patient. Hematol Oncol Clin North Am 2000; 14:113. 88. 89. 90. 91. 92.

Kay CR, Hannaford PC. Breast cancer and the pill—a further report from the Royal College of General Practitioners Oral Contraceptive Study. Br J Cancer 1988; 58:675. Lewis FM, Bloom JR. Psychosocial adjustment to breast cancer: a review of selected literature. Int J Psychiatry Med 1978; 9:1. Stevens LA, McGrath MH, Druss RG, et al. The psychological impact of immediate breast reconstruction for women with early breast cancer. Plast Reconstr Surg 1984; 73:619. Rowland JH, Holland JC, Chaglassian T, et al. Psychological response to breast reconstruction. Psychosomatics 1993; 34:241. Moran SL, Herceg S, Kurtelawicz K, Serletti JM. TRAM flap breast reconstruction with expanders and implants. AORN J 2000; 71:354.

CHAPTER 107 CONCEPTION, IMPLANTATION, AND EARLY DEVELOPMENT Principles and Practice of Endocrinology and Metabolism

CHAPTER 107 CONCEPTION, IMPLANTATION, AND EARLY DEVELOPMENT PHILIP M. IANNACCONE, DAVID O. WALTERHOUSE, AND KRISTINA C. PFENDLER Oocyte Maturation Sperm Capacitation and Fertilization Preimplantation Development Oviduct Transport X Chromosome Inactivation Implantation Postimplantation Development Spontaneous Abortion Perspectives Chapter References

The hormonal milieu plays an essential role in the production of parental germ cells, the biology of the reproductive process, and the subsequent creation, development, and survival of the offspring. To understand fully the impact of hormones on the adult, the child, the newborn, and the as-yet-unborn, the endocrinologist must be aware of the processes of conception, implantation, and early fetal development. Human loss through fetal wastage is significant. There are ~600,000 clinically apparent spontaneous abortions per year in the United States, and the 3 million live births probably represent 10 million conceptions.1 Fetal loss can occur at any of the major steps in development. The probability of pregnancy in any given menstrual cycle under optimal conditions is ~30%.2,3 and 4 The probability of successful fertilization may be as high as 85%, but 25% to 35% of conceptuses do not implant, and as many as 30% fail shortly after implantation. Undoubtedly, this loss represents a reproductive strategy. The rate-limiting feature of reproduction in mammals is the childbearing period. Therefore, if there is a possibility of loss of the individual during or immediately after the pregnancy, it is in the best interests of the species to eliminate the pregnancy as soon as possible and to make the mother available for another pregnancy. Thus, several critical steps of increasing sophistication of cellular coordination are required for the conceptus to enter midgestation and the organogenesis phase of development.

OOCYTE MATURATION Oocyte maturation in mammals proceeds from the development of a differentiated gonadal ridge in the fetus. For example, at approximately day 11 in the female mouse, primordial germ cells have located within the genital ridge, and ovarian development ensues.5 In humans, primordial germ cells arise in the yolk sac in the fourth week of fetal life and begin migration caudally toward the genital ridge during the fifth week. The first meiotic division occurs in the fetal ovary, and the oocyte becomes arrested at the diplotene stage. A germinal vesicle (nucleus) then forms. The first meiotic division is completed in the adult ovary, and the onset of this process is heralded by germinal vesicle breakdown. After telophase I, the first polar body of the oocyte is formed. The second meiotic division begins before ovulation, and the mature oocyte is fertilizable at the anaphase II stage. Fertilization occurs during anaphase II, and the completion of telophase II finds the zygote with two polar bodies and two pronuclei (Fig. 107-1). The oocyte and follicular maturation are discussed in Chapter 94.

FIGURE 107-1. Diagram of oocyte maturation. Completion of the prophase of first meiotic division occurs in the fetal ovary of most animals. At zygotene (stage 2), homologous maternal and paternal chromosomes commence pairing; at pachytene (stage 3), pairing has occurred throughout their lengths, and they form bivalents. Each homologue separates longitudinally to create two sister chromatids; thus, each bivalent forms a tetrad. It is during this stage that crossing over occurs, causing an interchange of genetic material between the paternal and maternal chromatids. At the diplotene stage (stage 4), the chromosomes commence their separation; they remain connected at their points of interchange (chiasmata). The germinal vesicle appears after the first meiotic arrest after the diplotene stage. The dictyate stage is a quiescent period, which may last for many years. In the adult ovary, the first meiotic division is completed. Ovulation occurs after extrusion of the first polar body (stage 11), and the second meiotic division (stages 12–14) is completed after sperm penetration. The zona pellucida is shown as a stippled ring (see Chap. 87). (From Tsafiri A. Oocyte maturation in mammals. In: Jones RE, ed. The vertebrate ovary. New York: Plenum Publishing, 1978:410.)

Maturation of the oocyte and ovulation are regulated by hormone levels, notably those of follicle-stimulating hormone (FSH). The extruded oocyte and its closely adherent cumulus adherens (follicular cells; corona radiata) are collected by the fimbriated end of the oviduct. The adherent cells communicate with one another through a complex network of intercellular bridges that extends from the innermost cells through the zona pellucida to the perivitelline space and into the oocyte6 (Fig. 107-2). These cells may have important nutritional functions for the oocyte and may control events in maturation or fertilization.7 The cumulus cells can bind tightly to the epithelial cells of the tube and may help initiate tubal transport. Transport of the egg to fertilization sites at the distal end of the oviduct and transport of the fertilized ovum to the uterus appear to be the concerted effort of the ciliary movement of the epithelium and muscular contractions of the myosalpinx. These contractions are not peristaltic. The sperm at this time are moving in the opposite direction and, although the cilia beat in the direction of the uterus, the muscular contractions of the oviduct do not give direction to moving particles within it. Particles can be propelled in either direction in the fallopian tubes of most species.8 The role of tubal secretions in the development of the early embryo has not been elucidated. These secretions do not have a demonstrable effect on the sperm because capacitation, which permits the acrosome reaction, can occur in chemically defined media.9

FIGURE 107-2. A, Photomicrograph of unfertilized, mature ovum with associated corona radiata cells. Coronal cells close to the ovum send processes through the zona pellucida. These processes (e.g., arrow) are evident as granules in the perivitelline space. B, Higher magnification in phase contrast shows these connecting processes of the corona radiata cells more clearly (arrow). (From Shettles LB. Ovum humanum. Munich: Urban und Schwarzenberg, 1960; 42:52.)

Because human oocytes can undergo spontaneous maturation in vitro, there appears to be an active inhibition of oocyte maturation within the follicle. Meiosis is

prevented by a maturation inhibitor produced by the granulosa cells of the follicle.10 Meiosis is resumed within the follicle after a surge of luteinizing hormone. If oocytes are removed from cumulus cells, maturation inhibitors are ineffective. Moreover, receptors for luteinizing hormone have been demonstrated on cumulus cells but not on denuded oocytes. Therefore, the cumulus cells are important mediators of both maturation inhibition and resumption of meiosis, directed by the preovulatory luteinizing hormone surge. Interestingly, in humans, oocyte maturation in vitro, as judged by germinal vesicle breakdown, is not necessary for sperm penetration, because penetration can be demonstrated at the first meiotic division. After sperm penetration of the mature oocyte, the sperm head swells and a pronucleus forms, with the sperm midpiece remaining visible. In immature oocytes, the sperm penetrates and swells but no pronucleus is formed. Thus, fertilization competence in humans is achieved only in fully mature oocytes at the time of the second meiotic division.

SPERM CAPACITATION AND FERTILIZATION The relatively thick and rigid structure that invests the mammalian egg, called the zona pellucida, has necessitated some changes in the physiology of fertilization, particularly with respect to the sperm. Mammalian sperm require the occurrence of two events before they can fertilize an oocyte. The first, known as capacitation, is the process by which sperm become competent for fertilization, an act they are not able to accomplish before an appropriate, species-dependent incubation time within the female reproductive tract milieu or similar in vitro medium.11,12 and 13 During this time, the sperm not only mature but also attain a state of hyperactivated motility that is necessary for them to move through the length of the female reproductive tract and to generate the force necessary to pierce through the cumulus oophorus and the zona pellucida of the oocyte. In addition, certain incompletely defined factors known as decapacitation factors must be removed from the sperm before they become competent for fertilization. Presumably, these factors are macromolecules that are blocking certain receptor sites necessary for this functional change to occur, and there is evidence that removal of these factors increases the response of the sperm to extracellular Ca2+.11 Once the sperm are capacitated, the acrosome reaction can begin, and it is through this process that the sperm can ultimately fuse with the oocyte. The morphology of the sperm head is such that an inner acrosomal membrane is immediately adjacent to the nuclear membrane of the cell, whereas an outer acrosomal membrane and the plasma membrane act as the limiting membrane of the acrosome.14 The acrosome itself contains proteases, such as acrosin, and other enzymes necessary for the sperm to navigate the interstices of the corona radiata. The outer acrosomal membrane possesses specific molecules for attachment to the zona before penetration of the egg, including a receptor that binds to a glycoprotein named ZP3 of the zona pellucida of the oocyte and a galactosyltransferase that recognizes N-acetylglucosamine residues.15,16,17 and 18 This morphology necessitates some interesting adaptations during the fusion of the sperm to the oocyte. Because the surface molecules necessary for attachment to the zona must be retained, the outer membrane must remain intact after the release of enzymes. The spermatozoon joins with the egg by membrane fusion of a mid-portion membrane, the equatorial region of the sperm head. The acrosome reaction, then, seems designed to create the structural alterations required for these various constraints to be overcome. First, the sperm-limiting membrane changes to allow influx of calcium, presumably along an electrochemical gradient. Immediately thereafter, the acrosomal membrane becomes fenestrated, appearing to allow the acrosomal contents to be released while leaving the acrosomal membrane, with its putative zona attachment elements, largely intact. The equatorial portion of the membrane is left intact for fusion with the oolemma, the limiting membrane of the unfertilized egg. Once fusion has occurred, the sperm head swells rapidly and forms the male pronucleus, leaving the sperm midpiece visible within the fertilized egg.19,20,21 and 22 Numerous cations play distinctive roles in these processes of capacitation and acrosomal exocytosis.22a Moreover, it is thought that the female reproductive tract is instrumental in regulating these processes by forming gradients of the cations at different positions along its length, as well as allowing their concentrations to change during certain times of the menstrual cycle.10 Ca2+, one of the most studied of these cations, is necessary for achieving the hyperactivated motility and the fertilizing ability associated with capacitation, in addition to being required for the acrosome reaction. It has been postulated that the binding of sperm to ZP3 of the zona pellucida triggers a G-protein pathway that ultimately leads to the release of bound Ca2+,15 and that this Ca2+ stimulates adenylate cyclase to produce cyclic adenosine monophosphate, which in turn activates cyclic adenosine monophosphate–dependent protein kinases that alter the sperm function during these prefertilization events.11 Na+also has been shown to be critical for capacitation and the acrosome reaction, although much higher concentrations of Na+ are required for the latter process. Finally, K+ plays a crucial role in these events, albeit in a more regulatory capacity. High levels of K+ do not inhibit capacitation, but they do suppress the fertilizing potential of the sperm. Before ovulation, Ca2+ and Na+ concentrations in the female reproductive tract are sufficient for capacitation, but the K+ concentration is too high to permit either the acrosome reaction to proceed or fertilization to occur. Follicular fluids released during ovulation, however, are thought to cause a substantial decrease in K+ concentration, as well as an added increase in Na+ concentration, which result in the fulfillment of fertilizing potential. In addition to the increased potential for fertilization during ovulation that is regulated by the concentrations of these ions, concentrations also seem to vary along the length of the female reproductive tract to help ensure that sperm proceed through capacitation and the acrosome reaction at the proper time and place to optimize fertilization.10,11 Immediately after fertilization, the maternal genome is activated and forms the female pronucleus. The sperm nucleus reforms and is evident morphologically as the male pronucleus. As the cells enter mitosis, the nuclear membranes of the pronuclei break down, and the chromosomes comigrate to the poles of the cell, where they are packaged as a unit in the nuclei of the progeny blastomeres. Thus, at the first cleavage, there is a symmetric division of the fertilized egg, and the two blastomeres have fused nuclei containing the maternal and the paternal genomes. It is clear that genetic information from both the mother and the father is an absolute requirement for normal development. When the maternal or paternal pronucleus is removed from fertilized mouse eggs and the egg is manipulated such that it contains either two maternal or two paternal pronuclei, development cannot proceed past midgestation. Bipaternal conceptions form only placenta, while bimaternal conceptions form disorganized embryonal tissue.23,24 As a result of these types of experiments, it has become evident that the same gene derived from the mother may be functionally different when derived from the father, leading to the concept of imprinting.25 Imprinting refers to a situation in which a gene is “marked” or “imprinted” during either female or male gametogenesis so that it is not expressed and, consequently, either the remaining paternal or maternal allele is exclusively expressed.26 The mechanism of imprinting remains uncertain, and it is not clear if each imprinted gene is imprinted using the same mechanism. Whatever the mechanism, the imprint must be maintained in somatic tissues during specific periods of development but must be reversible in the germline. Most imprinted genes are methylated in a parental-specific manner in the germ-line, and DNA methylation appears to be the most likely mechanism of imprinting.27,28 In fact, mice deficient for DNA methyltransferase, an enzyme that helps maintain methylation stability, lose monoallelic expression of several imprinted genes and die in the early postimplantation period, probably because of instability of primary imprints.29 Methylation may alter chromatin structure and modulate binding of transcriptional regulatory proteins to imprinted genes. The list of imprinted genes is expanding and includes Wilms tumor 1 (WT1), insulin, insulin-like growth factor-II (IGF-II), insulin-like growth factor-II receptor (IGF-IIR), H19, X-inactive specific transcript (Xist), and others.28 The extent of monoallelic expression varies for different imprinted genes during development. These discoveries are of great importance to medicine because aberrant imprinting has been demonstrated in the setting of human syndromes and cancer. Biallelic expression of an imprinted gene results in overexpression of the gene product compared with monoallelic expression as seen with imprinting. Expression from two alleles may occur by loss of the imprint or by deletion of the imprinted allele, with reduplication of the expressed allele resulting in uniparental disomy. Paternal uniparental disomy for the IGF-II locus has been described in Beckwith-Wiedemann syndrome (BWS).30 BWS is an overgrowth disorder characterized by gigantism, macroglossia, and visceromegaly. Because IGF-II is maternally imprinted, reduplication of the paternal allele results in a double dose of IGF-II expression. Since IGF-II functions as a fetal growth factor, this may be in part responsible for the overgrowth phenotype. Loss of the maternal IGF-II imprint or paternal uniparental disomy, again resulting in a double dose of IGF-II expression, has also been described in Wilms tumor cells.31 Here the excess growth factor may contribute to tumorigenesis. Finally, inheritance of a paternal deletion of chromosome region 15q11–13 is associated with Prader-Willi syndrome, characterized by obesity, hypogonadism, and mental retardation, whereas inheritance of this same deletion on the maternal chromosome is associated with Angelman syndrome, characterized by ataxic movements, inappropriate laughter, mental retardation, and hyperactivity.32 In both cases, the allele located in this region on the one normal chromosome 15 is not capable of sustaining normal development, suggesting that imprinting occurs on a chromosome of particular parental descent. Clearly, imprinting plays a crucial role in determining nonequivalence of the maternal and paternal genomes and is necessary for normal development.

PREIMPLANTATION DEVELOPMENT The preimplantation period of development in mammals, the time from conception to nidation (implantation), has variable lengths in the various species. In humans, the preimplantation period lasts for ~7 days; in the mouse, it is 4 days, whereas in the rat, it is 5 days. The fertilized egg (Fig. 107-3 and Fig. 107-4) is morphologically similar to the mature unfertilized egg. The embryo at this stage is 100 µm in diameter, is associated with two polar bodies remaining from meiotic division, and is surrounded by the amorphous zona pellucida. The zona pellucida, which is composed primarily of three complex glycoproteins known as ZP1, ZP2, and ZP3, is important to early development for several, largely mechanical, reasons. First, there is evidence suggesting that certain glycoproteins of the zona pellucida may play a role in the recognition of the egg by the sperm. Competitive inhibition assays have shown that by incubating mouse sperm with ZP3 before fertilization, binding of the sperm to the zona pellucida is inhibited, thus suggesting that this glycoprotein is responsible for the recognition and binding of the sperm to the zona pellucida.15 Second, the zona responds nearly instantaneously to sperm penetration and renders the egg impervious to additional penetration. Third, the zona provides a constraint to cleavage and ensures that as the blastomeres divide, they remain together and in the proper orientation. Finally, the zona prevents the naturally sticky

cleavage-stage embryo from adhering to the wall of the oviduct as it progresses to the uterus.

FIGURE 107-3. Early stages of mammalian development. The preimplantation stages shown are blastocysts with and without their zonae pellucidae. The embryonic end of the embryo contains the inner cell mass, which will form the fetus. Trophectodermal cells are fated to form the extraembryonic tissues, including the placenta. The primitive endoderm forms at the time of implantation and eventually will produce yolk sac structures. The primitive ectoderm will form the definitive ectoderm, endoderm, and mesoderm following the stages shown in the lower half of the diagram (postimplantation). The proamnion forms within the substance of the primitive ectoderm, and the trophoblast begins to differentiate into definitive placental structures (cytotrophoblast and syntrophoblast). By day 13, the primitive ectoderm has formed a single layer of columnar cells, and the craniocaudal groove (primitive streak) begins to form. Mesoderm differentiates from the primitive ectoderm at the point of the primitive streak in most primates.

FIGURE 107-4. Photomicrographs (Hoffmann modulation contrast) of living preimplantation mouse embryos. A, One-cell pronuclearstage egg ~12 hours after fertilization. A prominent pronucleus (arrow) and a polar body are evident. B, Two-cell cleavage stage. One of the two polar bodies is evident. C, Four-cell cleavage stage. One of the two polar bodies is evident. D, Eight-cell cleavage stage. E, Compacted 16-cell cleavage stage. F, Early blastocyst stage. A nascent blastocoele is evident. G, Midblastocyst stage. A well-formed blastocoele is evident in each embryo. Individual trophectodermal cells can be distinguished. The inner cell mass is apparent as an amorphous mass of cells. H, Expanded late blastocyst stage embryo. Inner cell mass is evident at lower right pole of the embryo. Individual trophectodermal cells are also evident. The embryos are surrounded by zonae pellucidae. The outside diameter of the embryos remains ~100 µm until the expanded blastocyst stage.

The mammalian oviduct also plays an important role in these early stages of preimplantation development. Not only does it provide a route through which the embryo is transported from the ovary to its site of implantation in the uterus, but it also provides a crucial timing mechanism for a process known as cleavage division.33,34 and 35 At this stage of development, the embryo divides symmetrically and reductively such that a geometric increase in the number of cells (blastomeres) occurs without an actual increase in the overall size of the embryo. These divisions occur entirely in the oviduct while the embryo is being propelled through its length as a result of ciliary action and muscular contractions in the oviduct wall. The primary function of development at this stage is to provide additional cells and membrane. Beginning after the first division in the mouse embryo and after the second division in the human embryo, a critical transition occurs in the genetic control of development.36 Before this time, the embryo contains a host of maternally derived mRNAs, ribosomes, and macromolecules that are sufficient to drive transcription and translation through the first (or second, as in the human embryo) cleavage division.37 Further development, however, is dependent on the activation of embryonic control of transcription and the subsequent degradation of maternal mRNAs and proteins.37 In the mouse, one population of polypeptides exhibits at least a two-fold decrease in abundance during the two-cell stage, whereas another population of polypeptides exhibits a similar increase in abundance.38 Although this change most likely reflects the degradation of maternal mRNA and the appearance of new embryonic mRNA, it is possible that this transition is not complete and that some maternal products still may persist for a time after this transition.37 Once the embryonic genome has been activated, two important morphogenetic events occur in the embryo during the preimplantation period. The first, known as compaction, occurs late in the eight-cell stage when individual blastomeres condense and their boundaries become less prominent, thus forming a cellular mass known as a morula (see Fig. 107-4E). This process results in several profound changes in the embryo (Fig. 107-5). During this time, several new gene products are expressed that contribute to many of the morphologic manifestations of compaction. Included in this group are E-cadherin, gap junction proteins, tight junction proteins, growth factors, and components of the cytoskeleton.39 E-cadherin (which originally was referred to as uvomorulin in the morula-stage mouse and later was identified as E-cadherin) acts as a Ca2+-dependent cell adhesion molecule that binds adjacent blastomeres together and appears to facilitate the formation of junctional complexes, which include both gap junctions and tight junctions. Gap junctions form between all cells of the compaction-stage embryo and are constructed from a family of proteins known as the connexins, the structure of which creates channels between cells that allow for communication between blastomeres. During compaction, these gap junctions migrate from central regions of intercellular contact to peripheral locations of contact where tight junctions also form, thus creating junctional complexes between lateral surfaces of the outer blastomeres.37,39,40 The tight junctions within these complexes serve a dual purpose. First, tight junctions play a critical role in the second preimplantation morphogenetic process known as cavitation. Second, they contribute to the polarization of the outer blastomeres by separating an apical region, where microvilli will form, from a basolateral region, to which the nuclei will migrate. By the 16-cell stage, these outer polar blastomeres form the trophectoderm, a cell lineage leading to the formation of extraembryonic tissues such as the placenta, whereas the inner apolar blastomeres form the inner cell mass, which ultimately will develop into the embryo proper.41,42 This is the first stage of commitment of cells to a particular fate. Before this time, each blastomere in the two-cell, four-cell, and early eight-cell embryo is totipotent and, therefore, has the potential to develop into a complete organism when it is isolated from the remaining embryo. Once the 16-cell stage of development has been reached, however, the embryo has sufficient cells to form an inside and an outside, and thus establishes the conditions necessary for the first step of embryonic commitment. It is at this point that some of the embryo's cells lose their totipotency.

FIGURE 107-5. Compaction and cavitation of the preimplantation mouse embryo. During the late eight-cell stage, embryos begin a morphogenetic process known as compaction, which ultimately results in the polarization of the outer blastomeres and the establishment of two cell lineages. During this process, individual blastomeres become less evident as the cell adhesion molecule E-cadherin functions to bind adjacent cells to one another. Simultaneously, microvilli form on the apical surfaces of

the outer blastomeres, the nuclei migrate basolaterally, gap junctions form between all adjacent blastomeres, and tight junctions form between outer blastomeres, thus separating apical and basolateral regions. These changes result in the formation of two cell lineages by the 16-cell stage; the outer polar cells will form the trophectoderm, whereas the inner apolar cells will form the inner cell mass. The second morphogenetic event, cavitation, begins as soon as the two cell lineages are established and results in the formation of a blastocoelic cavity. The basolateral location of E-cadherin aids in restricting the distribution of Na+/K+ –adenosine triphosphatases to this region, thus causing a Na+ gradient to form within the embryo. Water flows into the embryo osmotically, and the presence of the tight junctions in the outer blastomeres prevents this fluid from leaking out. Thus, a blastocoelic cavity forms, and the embryo is now known as a blastocyst.

Although the details of the mechanism by which the morphologic change in cellular contact associated with compaction can induce all of these varied events are incompletely deciphered, evidence suggests that protein kinase C and subsequent phosphorylation of proteins may be involved.43,44 and 45 Early activation of protein kinase C not only can trigger premature compaction through its effect on E-cadherin, but also can induce the migration of blastomere nuclei to a basolateral position.43,44 If protein kinase C functions in the signal pathway leading to events of compaction much as it does in other signal pathways, it is possible that some type of surface signal is detected by the blastomeres that causes a G protein to activate phospholipase C, which in turn cleaves phosphatidylinositol 4,5-bisphosphate. This results in the formation of two products: Ins (1,4,5)P3, which causes an increase in intracellular calcium, and 1,2-diacylglycerol, which activates protein kinase C. The activated protein kinase C is then available to phosphorylate proteins involved in nuclear migration and cell adhesion.43 It remains to be seen, however, what triggers this pathway, if indeed the phosphatidylinositol cycle functions to activate protein kinase C during compaction. The second major morphogenetic event to occur in preimplantation development is known as cavitation. This process begins several days after conception (3 days in the mouse, 4 in the rat, and 6 in the human) and culminates in the formation of the blastocyst (see Fig. 107-4F, Fig. 107-4G and Fig. 107-4H. At least two factors are known to be critical to the proper execution of this event. First, the tight junctions that form between plasma membranes of the outer blastomeres not only provide an apical/basolateral polarization of the cells, but also prevent paracellular leakage of fluid from the nascent blastocoele. Second, the cell adhesion properties of E-cadherin, which is located in the basolateral regions of the plasma membrane, are crucial in restricting the distribution of Na+/K+–adenosine triphosphatases to this region as well. With these two factors in place, the polar distribution of Na+/K+–adenosine triphosphatases to this basolateral location causes a Na+ gradient to be established within the interior of the embryo, and subsequently osmotic uptake of water occurs such that it accumulates in the extracellular space of the nascent blastocoele. Because the tight junctions prevent this fluid from leaking out, it accumulates until the blastocoelic cavity is fully expanded.37,39 At this stage, the two cell types are easily distinguishable: The inner cell mass cells are located internally at the embryonic pole of the embryo, whereas the trophectodermal cells, which are extremely large, owing to acytokinetic cell division, surround both the inner cell mass cells and the blastocoele.

OVIDUCT TRANSPORT The role of oviduct transport in the maturation of the mammalian embryo is poorly understood. It is reasonable to assume that oviduct fluids have a central role in the nourishment of the embryo and in gas exchange; however, the fluids also may contain substances that control or somehow enhance the development of the cleavage-stage embryo. The mammalian embryo can survive and progress in various artificial media. The embryo completes its passage through the oviduct at the early blastocyst stage and is propelled into the uterus. In the mouse, this occurs at day 3 of gestation, and in humans, at day 5 to 6. The blastocyst continues to develop in the uterus for another 24 to 48 hours, during which time it greatly expands its blastocoelic cavity until the inner cell mass is little more than a plaque of cells on the embryonic pole. Then, the embryo loses its zona pellucida. Despite attempts to isolate the factors involved in this process, little is known about the loss of the zona. In vitro, the zona can be removed by enzymatic digestion, mechanical disruption,16,17 or an acid milieu.46 In rabbits in vivo, the egg vestments are removed enzymatically at the implantation site and not while the blastocyst is free in the uterus.

X CHROMOSOME INACTIVATION In eutherian females (placental mammals, i.e., other than monotremes and marsupials), one of the two X chromosomes is inactivated early in embryonic development, thus providing a mechanism for genetic dosage compensation.47 In eutherians, this inactivation begins in the trophectoderm in the early blastocyst stage and is characterized by a preferential paternal X chromosome inactivation. This also is true of X chromosome inactivation that subsequently occurs in the primitive endoderm during the midblastocyst stage. During the late blastocyst stage, however, X chromosome inactivation occurs randomly in the inner cell mass, with no paternal or maternal preference, thus resulting in mosaic females composed of a mixture of cells that have either a maternally or paternally active X chromosome. In somatic cells, this inactivation becomes fixed such that all descendants from a particular cell maintain the same inactivated X chromosome. In the germline, however, this inactivation must be reversed at the time of meiosis so that each X chromosome has an equal chance of contributing to the gametes.48 In marsupials, this pattern of X chromosome inactivation is different in that it is always the paternal X chromosome that is inactivated. This may not necessarily be a functional difference, however, because the marsupial blastocyst has no inner cell mass. The coincidence of the timing of X chromosome inactivation and cell commitment to either trophectoderm or inner cell mass lineages in eutherians strongly suggests that these two processes are linked in some meaningful way. Perhaps the preferential X chromosome inactivation may be part of a system that is necessary to prevent rejection of the conceptus. Alternatively, it also has been suggested that preferential X chromosome inactivation may prevent the accumulation of genes necessary to the proper development of extraembryonic membranes on the paternal X chromosome. This would adversely affect the development of boys because they do not possess a paternally derived X chromosome.48,49 X chromosome inactivation was first proposed as a mechanism of gene dosage control by Lyon.50,51 The best evidence of it exists in women heterozygous at the glucose-6-phosphate dehydrogenase (G6PD) locus of the X chromosome. G6PD is a dimeric dimorphic enzyme; that is, there are two distinguishable allelic forms of the enzyme (isoenzymes) and the enzyme is composed of two subunits that must combine to form a holoenzyme. In heterozygous women, the two isoenzymes can combine to form heteropolymeric forms, which are distinguishable from the other two subunits. When X chromosome inactivation occurs in women heterozygous at the G6PD locus, two populations of cells are created: one with the paternal allele active and the other with the maternal allele active. The two isoenzymes can be distinguished by electrophoresis. No heteropolymeric form is present, however, indicating that the two alleles were not active simultaneously in the same cells at the time of sampling.52 Insights have been made into the multistep mechanism controlling X inactivation in mammals. First, the number of X chromosomes is counted by an as yet unknown process, tallying up the number of X inactivation centers (Xic). Second, a single X chromosome is chosen to remain active, and inactivation of any additional X chromosomes is initiated by expression of the Xist from the Xic. Finally, this inactivation spreads over the length of each inactive chromosome. Xist is only expressed on the inactive X in somatic cells of females, in male germ cells during spermatogenesis, and on the imprinted paternal X chromosome of the trophectoderm and primitive endoderm of the blastocyst. It does not encode a protein but instead remains as an RNA moiety that stays bound to the X chromosome undergoing inactivation. Furthermore, the expression of Xist coincides with the imprinted X inactivation that occurs in the trophectoderm and primitive endoderm of the blastocyst, which is then turned off before X inactivation in the embryonic lineage. DNA methylation of Xist correlates with its activity; it is unmethylated where it is expressed on an inactive X chromosome and methylated as an inactive allele on an active X chromosome. Xist can operate from multiple promoters, resulting in production of either stable or unstable RNA, suggesting one mechanism whereby Xist can be developmentally regulated. Stable Xist forms as a result of activation of one promoter on the imprinted paternal X chromosome of the trophectoderm. Alternatively, unstable Xist RNA results from activation of a different promoter when the imprint is erased before random X inactivation of the somatic cells.53,54 Questions that remain to be answered about X inactivation are how Xics are counted and how an inactivation signal can propagate throughout the entire length of the X chromosome and yet let certain genes escape this signal.

IMPLANTATION The embryo now undergoes implantation, which begins with attachment of the late blastocyst to the uterine tissue at a nidation site. The selection of this site is tightly regulated, because it usually occurs in a predictable manner, but little else is known. Implantation can be classified on the basis of the usual position of the site in the uterus and, hence, may be noninvasive and central, noninvasive and eccentric, or interstitial as in humans (Fig. 107-6). In humans, the embryo attaches to the lining of the uterine fundus, with the embryonic pole usually attaching to the antimesometrial lining. The endometrial cells of the uterus have microvilli on their luminal surfaces that begin to interdigitate with the microvilli of the trophectodermal cells. Pinocytosis (the cellular process of active engulfing of liquid) in the endometrial epithelial cells increases at this time and is thought to enhance or at least stabilize attachment, perhaps by removing uterine fluids from the attachment site. This pinocytosis is stimulated by progestins and inhibited by estrogens.21 Actual cell fusion between the embryonic trophectoderm and the uterine epithelium does not occur in most species. The presence of the blastocyst in the uterus undoubtedly provides some signal to the uterus and to the ovary to maintain the pregnancy.55 The blastocyst is

capable of producing human chorionic gonadotropin, which supports the corpus luteum, and the luteal phase of human conception cycles maintains higher progesterone levels from day 3 through day 8 than in nonconception cycles.56,57 and 58

FIGURE 107-6. Diagram of human implantation site. A, Trophoblast invasion of uterine epithelium at the time of attachment. B, Nidation site is completed with the embryo in its interstitial position. There is a single layer of abembryonic trophectodermal cells in contact with the uterine lumen. Primitive entoderm and primitive ectoderm are distinguishable. (From Tuchmann-Duplessis H, David G, Haegel P. Illustrated human embryology, vol I: embryogenesis. New York: Springer-Verlag, 1972.)

Implantation may be enhanced by proteases. These proteolytic enzymes are thought to have two functions: to cause the removal of the zona pellucida, which must precede the attachment of the embryo to the uterine lining, and to aid the embryo's invasion of the endometrial lining. The cells of the human trophoblast frankly invade uterine tissue as implantation proceeds. Early theories of the role of such enzymes suggested that they were necessary to digest maternal tissues; however, their actual role, if any, beyond the removal of the zona may be far more subtle. For example, such enzymes may act on the invasion process through limited proteolysis (e.g., blastolemmase) by beginning a cascade of activation of other enzymes.34 Implantation has at least three phases. The first is attachment, in which specific receptor sites may be responsible for binding of either the embryonic pole or the abembryonic pole, depending on the species, to the endometrial epithelium. The second phase is invasion. In humans, the trophectodermal cells invade through the basement membrane of the uterine epithelium to establish a nidation site in the stroma of the endometrium (see Fig. 107-6). The ability of the human embryo to invade tissue may explain the high frequency of ectopic pregnancies in women relative to other mammalian species. The third phase is the endometrial response to the implanted embryo. In a few eutherian species (including humans, other primates, and murine rodents), the uterine stromal cells undergo a specific reaction called decidualization. The name derives from the fact that these cells occasionally are shed at term. The stromal cells in the immediate area of the embryo become large, eosinophilic, and transcriptionally active. The cells of the decidual swelling may be important in the support of the pregnancy (e.g., by the production of luteotropin, which supports the corpus luteum); in the prevention of immune rejection of the implanted embryo; or in some other, unknown capacity.59 Among mammals, there is a great variation in the specific details of development at this stage. Although many attempts at generalizations across species have been made, by and large, they are either not helpful or are actually incorrect. For example, much of what is known concerning reproductive endocrinology is derived from experiments in mouse and rat. These animals, like other diapause mammals (i.e., those that can delay implantation while keeping the embryo alive), express an estrogen surge, which seems to be necessary for the progesteroneprimed uterus to accept the initiation of implantation. This is not true of humans. In the rabbit, the zona pellucida is removed at the site of implantation, and the blastocyst is invested with additional coverings that must be removed enzymatically. In the mouse, the blastocyst can exist free in the uterine cavity without its zona pellucida. Virtually nothing is known about the removal of the zona in humans. One important reason for these variations is that there are several successful solutions to problems of early development in viviparous animals, and many of the specific details of reproductive strategy do not allow for clear winners or clear losers. While the molecular control of implantation is not fully understood, many factors are necessary for proper implantation of the embryo. Most of the factors identified to date are produced or released by the uterus (many in response to estrogen or progesterone), but it is becoming evident that embryonic factors are important as well. COX-1, leukemia inhibitory factor (LIF), HB-EGF, and amphiregulin are expressed by the uterine epithelium at the time of implantation, while adhesion molecules such as lectins, carbohydrate moieties, and heparin sulfate proteoglycan interact between the surfaces of the blastocyst and endometrium. Interleukin (IL)-1a and IL-1b are released by the blastocyst and adhere to the endometrial epithelial b3-integrin subunit, while trophoblast giant cells produce proteinases such as gelatinases A and B and urokinase-type plasminogen activator (uPA) that mediate the invasion of the decidua.60,61,62,63 and 64 In diapause mammals such as the mouse or rat, there are uterine inhibitory factors that can prevent implantation. The blastocyststage embryo can overcome the inhibition by a process of activation, which occurs in response to the prenidation estrogen surge in these animals. This process does not occur in humans.

POSTIMPLANTATION DEVELOPMENT Development after implantation is rapid and complex. The embryo must establish both its placental compartment and its definitive fetal structures in a short time. The polar or embryonic trophectoderm (that overlying the inner cell mass) develops into an ectoplacental cone in the mouse, whereas in most primates, the trophoblast differentiates into syncytiotrophoblast and cytotrophoblast, the latter having a high mitotic rate. Rapid division produces a syncytial trophoblast surrounding the primate embryo, although the mural trophectoderm (that facing the uterine cavity) remains a single layer of cells. Lacunar spaces form within the syntrophoblast, which eventually becomes contiguous with the maternal capillary circulation, into which the chorionic villi will grow. The ectoplacental cone of the mouse and rat undergoes similar development, and the resulting placental structure is hemochorial, as in humans. The major placental classification among mammalian orders is derived from the number of tissue layers that separate the fetal and maternal circulations. There are six such potential barriers to exchange. Humans, like many other primates and murine species, have a hemochorial placenta in which three fetal tissues (endothelium, connective tissue, and chorionic epithelium) are bathed in maternal blood.65 As the process of placentation proceeds, definitive embryonic structures are developing. Immediately after implantation, a layer of cells appears at the blastocoele margin on the side of the inner cell mass. This layer is called the endoderm. The remaining cells of the inner cell mass are now called the epiblast or the primitive ectoderm. The endoderm proliferates rapidly and eventually surrounds the blastocoele. The epiblast cells (embryonic ectoderm) are now arranged in a columnar manner. Cells contiguous with the epiblast, called amnioblasts, appear; spaces between the amnioblasts develop (the proamnion) and eventually form the amnionic cavity. Although it is a matter of some debate, it seems possible that the amnioblast cells are the source of the amniotic fluid, which cushions and thereby protects the developing embryo. Apoptosis also plays a critical role in cavitation of the early embryo (not to be confused with cavitation of the blastocyst). A signal from the primitive endoderm acts over a short distance to induce apoptosis of the inner ectoderm cells, while survival of the outer ectodermal cells is mediated by interaction with the adjacent basement membrane that separates the ectoderm from the endoderm.66 By approximately 7 days in the mouse and 13 days in humans, all three germ layers are present. The primitive ectoderm (epiblast) gives rise to definitive ectoderm, definitive endoderm, and mesoderm, which appears between the primitive endoderm and the primitive ectoderm. The primitive endoderm gives rise to several extraembryonic tissues (see Fig. 107-3). The first indication of craniocaudal axis and bilateral symmetry in the embryo appears as a longitudinal depression in the columnar embryonic ectoderm. This depression is called the primitive streak and, in most primates, it seems to be the site of origin of mesodermal cells (see Fig. 107-4). Most of the available information concerning cell lineage in the early embryo is derived from experiments performed in the mouse, and it is not clear whether these principal features of the fates of early cells are applicable to human development. It may be some time before this can be determined because, at present, the only way this information can be obtained is by experimental manipulation and disruption of the embryo. A case in point is a series of experiments that defined the ultimate fates of areas of the egg cylinder–stage embryo of the mouse (day 7). This work required microsurgical removal of some structures from the embryo with development in culture, or transplantation of radioactively labeled structures to unlabeled embryos, after the development of the combined structure. Early postimplantation stages are responsible for establishing the structures that ultimately allow organogenesis to proceed.33,35 An understanding of the molecular biology of the control of differentiation of the definitive structures will have far-reaching implications for many gestational diseases and certainly for human cancer.67 Correct fetal development requires the coordinated expression of thousands of genes. The correct temporal and spatial expression of these genes could not occur without the intervention of some relatively small set of supervisor genes that can orchestrate the process. Such genes are being found based on information from diverse animal studies.68

SPONTANEOUS ABORTION

The most common manifestation of the failure of embryonic and fetal development is spontaneous abortion: the failure of conception to produce a live birth. Spontaneous abortion, then, is either the disruption of pregnancy once it can be recognized or the expulsion of a nonviable fetus. Precise clinical definitions are much more difficult. Most often, these definitions must invoke low birth weight, because below certain weights, the fetus is unlikely to survive. Other definitions include loss of pregnancy before 20 or 28 weeks of gestation. Accurate estimates of the incidence of spontaneous abortion, therefore, are difficult to obtain. The frequency of clinically evident spontaneous abortion is ~15% of pregnancies. Undoubtedly, the risk is much higher in women with a previous spontaneous abortion, with the risk as high as 46% after three consecutive abortions. However, if the abortus is karyotypically abnormal, the risk of consecutive abortion is substantially lower.69,70 The association of prior spontaneous abortion with subsequent poor pregnancy outcome has been well documented, even when all other risk factors have been controlled. The effects of specific risk factors seem to be much stronger than the history. Early pregnancy losses are occult. Early abortion has many causes and must not be considered a single disease entity. One of the principal observations in human embryos that fail to cleave normally is the presence of structural abnormalities of chromosomes. In a large series of fetal deaths, the karyotypes of the offspring were compared with the morphology of the conception products.71 More than half of the small or unformed fetuses had chromosomal abnormalities, whereas only 6% of fetuses of normal size with or without malformations had chromosomal aberrations (see Chap. 90). Intrauterine death may occur in association with chromosomal abnormalities that also can be seen in live births. These deaths result from the failure of embryonic development, not the gross anomalies frequently associated in live offspring with the deviant karyotype. There may exist a continuum of anomalies in the offspring into which spontaneous abortion fits, from failure of fetal development through to birth with malformations. Nonchromosomal causes of pregnancy loss include maternal metabolic disturbances such as endometrial growth factor disturbances or hyperglycemia.72,73 One potential source of disruption of pregnancy is exposure of the woman to toxic substances. Of particular concern are exposures in the early periods of pregnancy. Few data are available, however, in some measure because of the traditional view that preimplantation development is refractory to toxic insult. However, the general presumption that early-stage embryos are either killed or left unaffected to implant and develop normally is an oversimplification. For example, the blastocyst is sensitive to cyclophosphamide, heavy metals, and trypan blue. Such exposures decrease cell numbers in the early embryo and can lead to vascular anomalies in midgestation when exposure occurs at the blastocyst stage. Exposure to toxic substances can be environmental, such as in the workplace, or self-inflicted, such as maternal smoking.73a Maternal smoking is important because of the numerous persons involved, and has been implicated by association in a wide array of pregnancy complications involving both the mother and the offspring. These complications include low birth weight, spontaneous abortion, sudden infant death syndrome, placenta previa, excessive maternal bleeding, and perinatal mortality.74,75 and 76 Many laboratories have been investigating possible reasons for the adverse role of maternal smoking in pregnancy outcome (see Chap. 234), and several conclusions have emerged. First, the embryo can be affected directly by chemical exposure; there need not be any intervening maternal role. Nevertheless, injury to maternal pregnancy support systems, such as the corpus luteum, may occur, or maternal tissues may activate deleterious compounds in tobacco smoke. Second, the embryo is at risk for adverse effects much sooner than was previously suspected. The blastocyststage embryo is sensitive to compounds such as those in cigarette smoke with respect to implantation, decidual response, gross dysmorphogenesis, live birth rate, and perinatal mortality. These events can be manifested long after the exposure to the chemicals.77,78,79,80 and 81 Because mothers are unaware of early pregnancy, these data may require a reevaluation of the advice given to women who are considering pregnancy: It is becoming clear that one should not wait for evidence of the pregnancy to refine the potential environment of the developing embryo.82

PERSPECTIVES The study of the progression of embryonic tissue is the study of evolution, organization, differentiation, and molecular control. It has attracted the attention of endocrinologists, biologists, clinicians, and amateur naturalists for centuries. There can be no doubt that detailed investigation of the issues surrounding reproductive strategies of species both related and unrelated to humans will yield abundant insight that will help alleviate human ailments as diverse as birth defects and cancer. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

Fabro S. Reproductive toxicology: state of the art. Am J Ind Med 1983; 4:391. Roberts CJ, Lowe CR. Where have all the conceptions gone? Lancet 1975; 1:498. Hertig AT. The overall problem in man. In: Benirschke K, ed. Comparative aspects of reproductive failure. Berlin: Springer-Verlag, 1967:11. King CR, Pernoll ML, Prescott G. Reproductive wastage. Obstet Gynecol Annu 1982; 11:59. Newbold RR, Carter DB, Harris SE, et al. Molecular differentiation of the mouse genital tract: altered protein synthesis following prenatal exposure to diethylstilbestrol. Biol Reprod 1984; 30:459. Shettles LB. Ovum humanum. Munich: Urban und Schwarzenberg, 1960:79. Racowsky C, Satterlie RA. Metabolic, fluorescent dye and electrical coupling between hamster oocytes and cumulus cells during meiotic maturation in vivo and in vitro. Dev Biol 1985; 108:191. Jansen RP. Endocrine response in the fallopian tube. Endocr Rev 1984; 5:525. Chang MC. The meaning of sperm capacitation: a historical perspective. J Androl 1984; 5:45. O'Neill C, Quinn P. Inhibitory influence of uterine secretions on mouse blastocysts decreases at the time of blastocyst activation. J Reprod Fertil 1983; 68:269. Fraser LR, Umar G, Sayed S. Na+-requiring mechanisms modulate capacitation and acrosomal exocytosis in mouse spermatozoa. J Reprod Fertil 1993; 97:539. Fraser LR. Requirements for successful mammalian sperm capacitation and fertilization. Arch Pathol Lab Med 1992; 116:345. Spungin B, Levinshal T, Rubinstein S, Breitbart H. A cell free system reveals that capacitation is a prerequisite for membrane fusion during the acrosome reaction. FEBS Lett 1992; 311:155. Oura C, Toshimori K. Ultrastructural studies on the fertilization of mammalian gametes. Int Rev Cytol 1990; 122:105. Gilbert SF. Developmental biology, 3rd ed. Sunderland, MA: Sinauer Associates Inc, 1991:33. Bleil JD, Wassarman PM. Identification of a ZP3-binding protein on acrosome-intact mouse sperm by photoaffinity crosslinking. Proc Natl Acad Sci U S A 1990; 87:5563. Shur BD, Hall NG. A role for mouse sperm surface galactosyltransferase in sperm binding to the egg zona pellucida. J Cell Biol 1982; 95:574. Shur BD, Neely CA. Plasma membrane association purification and partial characterization of mouse sperm 1,4-galactosyltransferase. J Biol Chem 1988; 268:17706. Bedford JM. Significance of the need for sperm capacitation before fertilization in eutherian mammals. Biol Reprod 1983; 28:108. Hinrichsen-Kohane AC, Hinrichsen MJ, Schill WB. Molecular events leading to fertilization—a review. Andrologia 1984; 16:321. Hendrickx AG. Disorders of fertilization, transport, and implantation. Prog Clin Biol Res 1984; 160:211. Farooqui AA. Biochemistry of sperm capacitation. Int J Biochem 1983; 15:463.

22a. Espinosaal F, Lopez-Gonzaleza T, Munoz-Garaya C, et al. Dual regulation of the T-Type Ca (2+) current by serum albumin and beta-estradiol in mammalian spermatogenic cells. FEBS Lett 2000; 475:251. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.

Barton SC, Surani MAH, Norris ML. Role of paternal and maternal genomes in mouse development. Nature 1994; 311:374. Spindle A, Sturm KS, Flannery M, et al. Defective chorioallantoic fusion in mid-gestation lethality of parthenogenone-tetraploid chimeras. Dev Dyn 1996; 173:447. Sapienza C, Peterson AC, Rossant J, et al. Degree of methylation of trans-genes is dependent on gamete of origin. Nature 1987; 328:251. Pedersen RA, Sturm KS, Rappolee DA, Werb Z. Effects of imprinting on early development of mouse embryos. In: Bauister BD, ed. Preimplantation embryo development. New York: Springer-Verlag, 1993:212. Surani MA. Imprinting and the initiation of gene silencing in the germ line. Cell 1998; 93:309. Barlow DP. Gametic imprinting in mammals. Science 1995; 270:1610. Li E, Beard C, Jaenisch R. Role for DNA methylation in genomic imprinting. Nature 1993; 366:362. Junien C. Beckwith-Wiedemann syndrome, tumorigenesis and imprinting. Curr Opin Genet Dev 1992; 2:431. Okawa O, Eccles MR, Szeto J, et al. Relaxation of insulin-like growth factor II gene imprinting implicated in Wilms' tumour. Nature 1993; 362:749. Cassiday SB, Schwartz S. Prader-Willi and Angelman syndromes: disorders of genomic imprinting. Medicine 1998; 77:140. McLaren A. Early mammalian development. Prog Clin Biol Res 1985; 163A:29. Denker HW. Basic aspects of ovoimplantation. Obstet Gynecol Annu 1983; 12:15. Swartz WJ. Early mammalian embryonic development. Am J Ind Med 1983; 4:51. Telford NA, Watson AJ, Schultz GA. Transition from maternal to embryonic control in early mammalian development: a comparison of several species. Mol Reprod Dev 1990; 26:90. Watson AJ, Kidder GM, Schultz GA. How to make a blastocyst. Biochem Cell Biol 1992; 70:849. Latham KE, Garrels JI, Chang C, Solter D. Quantitative analysis of protein synthesis in mouse embryos I. Extensive reprogramming at the one- and two-cell stages. Development 1991; 112:921. Watson AJ. The cell biology of blastocyst development. Mol Reprod Dev 1992; 33:492. Becker DL, Leclerc-David C, Warner A. The relationship of gap junctions and compaction in the preimplantation mouse embryo. In: Stern C, Ingham P, eds. Gastrulation (Dev Suppl). Cambridge: The Company of Biologists Limited, 1992:113. Rossant J, Papaioannou VE. The biology of embryogenesis. In: Sherman MI, ed. Concepts in mammalian embryogenesis. Cambridge, MA: MIT Press, 1977:36. Sutherland AE, Calarco-Gillam PG. Analysis of compaction in the preimplantation mouse embryo. Dev Biol 1983; 100:328. Winkel GK, Ferguson JE, Takeichi M, Nuccitelli R. Activation of protein kinase C triggers premature compaction in the four-cell stage mouse embryo. Dev Biol 1990; 138:1. Ohsugi M, Ohsawa T, Yamamura H. Involvement of protein kinase C in nuclear migration during compaction and the mechanism of the migration: analyses in two-cell mouse embryos. Dev Biol 1993; 156:146. Bloom T, McConnell J. Changes in protein phosphorylation associated with compaction of the mouse preimplantation embryo. Mol Reprod Dev 1990; 26:199. Nicolson GL, Yanagimachi R, Yanagimachi H. Ultrastructural localization of lectin binding site on the zonae pellucidae and plasma membranes of mammalian eggs. J Cell Biol 1975; 66:263. Migeon BR. X-chromosome inactivation: molecular mechanisms and genetic consequences. Trends Genet 1994; 10:230. Gartler SM, Dyer KA, Goldman MA. Mammalian X-chromosome inactivation. Mol Genet Med 1992; 2:121. Solter D. Differential imprinting and expression of maternal and paternal genomes. Annu Rev Genet 1988; 88:127. Lyon MF. Mechanisms and evolutionary origins of variable X-chromosome activity in mammals. Proc R Soc Lond [Biol] 1974; 187:243. Chapman VM, Shows TB. Somatic cell genetic evidence for X-chromosome linkage of three enzymes in the mouse. Nature 1976; 259:665. Migeon BR. Glucose-6-phosphate dehydrogenase as a probe for the study of X-chromosome inactivation in human females. Curr Top Biol Med Res 1983; 9:189. Goto T, Monk M. Regulation of X-chromosome inactivation in development in mice and humans. Microbiol Mol Biol Rev 1998; 62:362. Johnston CM, Nesterova TB, Formstone EJ, et al. Developmentally regulated Xist promoter switch mediates initiation of X inactivation. Cell 1998; 94:809. Kennedy TG. Embryonic signals and the initiation of blastocyst implantation. Aust J Biol Sci 1983; 36:531.

56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73.

Casper RF, Wilson E, Collins JA, et al. Enhancement of human implantation by exogenous chorionic gonadotropin. (Letter). Lancet 1983; 2:1191. Kusuda M, Nakamura G, Matsukuma K, et al. Corpus luteum insufficiency as a cause of nidatory failure. Acta Obstet Gynecol Scand 1983; 62:199. Buster JE. Gestational changes in steroid hormone biosynthesis, secretion, metabolism, and action. Clin Perinatol 1983; 10:527. Bazer FW, Roberts RM. Biochemical aspects of conceptus-endometrial interactions. J Exp Zool 1983; 228:373. Rinkenberger JL, Cross JC, Werb Z. Molecular genetics of implantation in the mouse. Dev Genet 1997; 21:6. Smith SE, French MM, Julian J, et al. Expression of heparan sulfate proteoglycan (Perlecan) in the mouse blastocyst is regulated during normal and delayed implantation. Dev Biol 1997; 184:38. Cullinan EB, Abbondanzo SJ, Anderson PS, et al. Leukemia inhibitory factor (LIF) and LIF receptor expression in human endometrium suggests a potential autocrine/paracrine function in regulating embryo implantation. Proc Natl Acad Sci U S A 1996; 93:3115. Das SK, Wang X-N, Paria BC, et al. Heparin-binding EGF-like growth factor gene is induced in the mouse uterus temporally by the blastocyst solely at the site of its apposition: a possible ligand for interaction with blastocyst EGF-receptor in implantation. Development 1994; 120:1071. Schultz GA, Edwards DR. Biology and genetics of implantation. Dev Genet 1997; 21:1. Benirschke K. Placentation. J Exp Zool 1983; 228:385. Coucouvanis E, Martin GR. Signals for death and survival: a two-step mechanism for cavitation in the vertebrate embryo. Cell 1995; 83:279 Sanford JP, Chapman VM, Rossant J. DNA methylation in extraembryonic lineages of mammals. Trends Genet 1985; 1:89. Utset MF, Awqulewitsch A, Ruddle FH, McGinnis W. Region-specific expression of two mouse homeo box genes. Science 1987; 235:1379. Huisjes HJ. Spontaneous abortion. New York: Churchill Livingstone, 1984:205. Wilcox AJ. Surveillance of pregnancy loss in human populations. Am J Ind Med 1983; 4:285. Byrne J, Warburton D, Kline J, et al. Morphology of early fetal deaths and their chromosomal characteristics. Teratology 1985; 32:297. Freinkel N, Lewis NJ, Akazawa S, et al. The honeybee syndrome: implications of the teratogenicity of mannose in rat-embryo culture. N Engl J Med 1984; 310:223. Giudice LC. Growth factors and growth modulators in human uterine endometrium: their potential relevance to reproductive medicine. Fertil Steril 1994; 61:1.

73a. Zenzes MT. Smoking and reproduction: gene damage to human gametes and embryos. Hum Reprod Update 2000; 6:122. 74. Tovares R, Ramos P, Palminha J, et al. Transplacental exposure to genotoxins. Evaluation in hemoglobin of hydroxyethylvaline adduct levels in smoking and non-smoking mothers and their newborns. Carcinogenesis 1994; 15:1271. 75. Jacobson JL, Jacoboson SW, Sokol RJ, et al. Effects of alcohol use, smoking, and illicit drug use on fetal growth in black infants. J Pediatr 1994; 124:757. 76. Naeye RL. Common environmental influences on the fetus. Monogr Pathol 1981; 22:52. 77. Iannaccone PM, Tsao TY, Stols L. Effects on mouse blastocysts of in vitro exposure to methylnitrosourea and 3-methylcholanthrene. Cancer Res 1982; 42:864. 78. Iannaccone PM. Long-term effects of exposure to methylnitrosourea on blastocysts following transfer to surrogate female mice. Cancer Res 1984; 44:2785. 79. Iannaccone PM, Fahl WE, Stols L. Reproductive toxicity associated with endometrial cell mediated metabolism of benzo[a]pyrene: a combined in vitro, in vivo approach. Carcinogenesis 1984; 5:1437. 80. Bossert NL, Iannaccone PM. Midgestational abnormalities associated with in vitro preimplantation N-methyl-N-nitrosourea exposure with subsequent transfer to surrogate mothers. Proc Natl Acad Sci U S A 1985; 82:8757. 81. Dwivedi RS, Iannaccone PM. Effects of environmental chemicals on early development. In: Korach K, ed. Reproductive and developmental toxicology. New York: Marcel Dekker Inc, 1998:11. 82. Iannaccone PM, Bossert NL, Connelly CS. Disruption of embryonic and fetal development due to preimplantation chemical insults: a critical review. Am J Obstet Gynecol 1987; 157:476.

CHAPTER 108 THE MATERNAL-FETAL-PLACENTAL UNIT Principles and Practice of Endocrinology and Metabolism

CHAPTER 108 THE MATERNAL-FETAL-PLACENTAL UNIT BRUCE R. CARR Placental Compartment Progesterone Estrogen Human Chorionic Gonadotropin Human Placental Lactogen Human Growth Hormone Variant Other Placental Peptide Hormones Fetal Membranes and Decidua Fetal Compartment Hypothalamic-Pituitary Axis Growth Hormone, Prolactin, Vasopressin, and Oxytocin Thyroid Gland Gonads Adrenal Gland Parathyroid Gland and Calcium Homeostasis Pancreas Role of Hormones in Lung Maturation Maternal Compartment—Endocrine Alterations and Endocrine Diseases Associated with Pregnancy Hypothalamus and Pituitary Gland Thyroid Gland Adrenal Gland Parathyroid Gland Ovarian Androgen-Secreting Tumors Evaluation of the Fetal-Placental Unit by Endocrine Testing Early Pregnancy Late Pregnancy Chapter References

The hormonal changes and maternal adaptations of human pregnancy are among the most remarkable phenomena in nature. During pregnancy, the placenta, which is supplied with precursor hormones from the maternal-fetal unit, synthesizes large quantities of steroid hormones as well as various protein and peptide hormones and secretes these products into the fetal and maternal circulations. Near the end of pregnancy, a woman is exposed daily to ~100 mg estrogen, 250 mg progesterone, and large quantities of mineralocorticoids and glucocorticosteroids. The mother, and to a lesser extent the fetus, are also exposed to large quantities of human placental lactogen (hPL), human chorionic gonadotropin (hCG), prolactin, relaxin, and prostaglandins and to smaller amounts of proopiomelanocortin (POMC) derived peptides such as adrenocorticotropic hormone (ACTH) and endorphin, gonadotropin-releasing hormone (GnRH), thyroid-stimulating hormone (TSH), corticotropin-releasing hormone (CRH), somatostatin, and other hormones. Implantation, the maintenance of pregnancy, parturition, and finally lactation are dependent on a complex interaction of hormones in the maternal-fetal-placental unit. Moreover, there exists a complex regulation for the secretion of steroid hormones by means of protein and peptide hormones also produced within the placenta. In this chapter the discussion is focused on the hormones secreted by the placenta, the endocrinology of the fetus and the mother, the effect of various endocrine diseases on the maternal-fetal unit, and the use of endocrine tests to assess fetal well-being.

PLACENTAL COMPARTMENT In mammals, especially humans, the placenta has evolved into a complex structure that delivers nutrients to the fetus, produces numerous steroid and protein hormones, and removes metabolites from the fetus to the maternal compartment. The structure of the placenta is discussed in Chapter 111 PROGESTERONE The principal source of progesterone during pregnancy is the placenta, although the corpus luteum is the major source during the first 6 to 8 weeks of gestation,1 when progesterone is essential for the development of a secretory endometrium to receive and implant a blastocyst. Apparently, the developing trophoblast takes over as the principal source of progesterone by 8 weeks, since removal of the corpus luteum before this time, but not after, leads to abortion.2 After 8 weeks, the corpus luteum contributes only a fraction of the progesterone secreted. The placenta of a term pregnancy produces ~250 mg progesterone each day. Maternal progesterone plasma levels rise from 25 ng/mL during the late luteal phase to 40 ng/mL near the end of the first trimester to 150 ng/mL at term (Fig. 108-1A). Most progesterone (90%) secreted by the placenta enters the maternal compartment.

FIGURE 108-1. Range (mean ± 1 standard deviation) of progesterone (A), estradiol-17 (B), and estriol (C) in plasma of normal pregnant women as a function of a week of gestation. (Courtesy of Dr. C. Richard Parker, Jr.)

Although the placenta produces large amounts of progesterone, it normally has very limited capacity to synthesize precursor cholesterol from acetate. Radiolabeled acetate is only slowly incorporated into cholesterol in placental trophoblasts, and the activity of the rate-limiting enzyme of cholesterol biosynthesis—HMG-CoA reductase—in placental microsomes is low. Therefore, maternal cholesterol, in the form of low-density lipoprotein (LDL) cholesterol, is the principal substrate for the biosynthesis of progesterone.3,4 LDL cholesterol attaches to its receptor on the trophoblast and is taken up and degraded to free cholesterol, which then is converted to progesterone and secreted. These findings not only have provided new insights into the biochemical basis for placental progesterone formation, but they have also provided clues to other aspects of maternal-placental physiology. For example, the rate of progesterone secretion may depend on the number of LDL receptors on the trophoblast and may be independent of placental blood flow: (a) Cholesterol side-chain cleavage by the placental mitochondria is in a highly activated state, perhaps meaning that the placenta is under constant trophic stimulation and that hCG and GnRH produced by the placenta are the trophic substances; (b) cholesterol synthesis from acetate is limited, as discussed earlier; (c) the fetus does not contribute precursors for placental biosynthesis; and (d) the levels of maternal LDL are not rate-limiting for the placental products of cholesterol or progesterone.5 A functioning fetal circulation is unimportant for the regulation of progesterone levels in the maternal unit. In fact, fetal death, ligation of the umbilical cord, or anencephaly—which all are associated with a decrease in estrogen production—have no significant effect on progesterone levels in the maternal compartment.6,7 The physiologic role of the large quantity of progesterone includes binding to receptors in uterine smooth muscle, thereby inhibiting contractility and leading to myometrial quiescence. Progesterone also inhibits prostaglandin formation, which is critical in human parturition8 (see Chap. 109). Progesterone is essential for the

maintenance of pregnancy in all mammals, possibly because of its ability to inhibit the T-lymphocyte cell-mediated responses involved in graft rejection. The high local levels of progesterone can block cellular immune response to foreign antigens such as a fetus, creating immunologic privilege for the pregnant uterus.9 ESTROGEN During human pregnancy, the rate of estrogen production and the levels of estrogen in plasma increase markedly (Fig. 108-1B and C), and the levels of urinary estriol increase 1000-fold.10 In fact, it has been estimated that during a pregnancy, a woman produces more estrogen than a normal ovulatory woman could produce in 150 years!5 The corpus luteum of pregnancy is the principal source of estrogen during the first few weeks; subsequently, nearly all of the estrogen is formed by the trophoblast of the placenta. The mechanism by which estrogen is synthesized by the placenta is unique (Fig. 108-2). The placenta cannot convert progesterone to estrogens because of a deficiency of 17a-hydroxylase (CYP17). Thus, it must rely on androgens produced in the maternal and fetal adrenal glands. Estradiol-17b and estrone are synthesized by the placenta by conversion of dehydroepiandrosterone sulfate (DHEAS) that reaches it from both the maternal and the fetal blood. Near term, 40% of the estradiol-17b and estrone is formed from maternal DHEAS and 60% of the estradiol-17b and estrone arises from fetal DHEAS precursor.11 The placenta metabolizes DHEAS to estrogens through placental sulfatase, D4,5-isomerase and 3b-hydroxysteroid dehydrogenase, and aromatase enzyme complex. Estriol is synthesized by the placenta from 16a-hydroxydehydroepiandrosterone sulfate (16a-OHDS) formed in the fetal liver from circulating DHEAS. At least 90% of urinary estriol ultimately is derived from the fetal adrenal gland,11 which secretes steroid hormones at a high rate, sometimes up to 100 mg per day, mostly as DHEAS. The principal precursor for this DHEAS is LDL cholesterol circulating in fetal blood. A minor source is formation from pregnenolone secreted by the placenta. Only 20% of fetal cholesterol is derived from the maternal compartment, and because amniotic fluid cholesterol levels are negligible, the principal source of cholesterol appears to be the fetus itself. The fetal liver synthesizes cholesterol at a high rate and may supply sufficient cholesterol to the adrenals to maintain steroidogenesis.12

FIGURE 108-2. Sources of estrogen biosynthesis in the maternal-fetal-placental unit. (LDL, low-density lipoprotein; chol, cholesterol; OHDS, hydroxydehydroepiandrosterone sulfate; OHSDH, hydroxysteroid dehydrogenase; C2 pool, carbon-carbon unit; DHEAS, dehydroepiandrosterone sulfate; E1, estrone; E2, estradiol-17b7; E3, estriol.) (From Carr BR, Gant NE. The endocrinology of pregnancy-induced hypertension. Clin Perinatol 1983; 10:737.)

Estetrol is a unique estrogen, the 15a-hydroxy derivative of estriol, which is derived exclusively from fetal precursors and fetal metabolism. Although the measurement of estetrol in pregnant women had been proposed as an aid in monitoring a fetus at risk for intrauterine death, it is not superior to the measurement of urinary estriol.13 Several disorders lead to low urinary excretion of estriol by the mother. A particularly interesting one is placental sulfatase deficiency,14 also known as the steroid sulfatase deficiency syndrome, an X-linked metabolic disorder characterized during fetal life by decreased maternal estriol production secondary to this deficient enzymatic activity (see Fig. 108-2), which renders the placenta unable to cleave the sulfate moiety from DHEAS. Placental sulfatase deficiency is associated with prolonged gestation and difficulty in cervical dilatation at term, often necessitating cesarean section. Steroid sulfatase deficiency is thought to occur in 1 of every 2000 to 6000 neonates. The male offspring are, of course, sulfatase deficient, manifest clinically by ichthyosis during the first few months of life. The genetic locus for steroid sulfatase deficiency is on the distal short arm of the X chromosome.15 Most of the estrogen secreted by the placenta is destined for the maternal compartment, as is true for progesterone: 90% of the estradiol-17b and estriol enters the maternal compartment. Interestingly, estrone is the estrogen preferentially secreted into the fetal compartment.16 The physiologic role of the large quantity of estrogen produced by the placenta is not completely understood. It may regulate or fine-tune the events leading to parturition, because pregnancies are often prolonged when estrogen levels in maternal blood and urine are low, as in placental sulfatase deficiency or when the fetus is anencephalic. Estrogen stimulates phospholipid synthesis and turnover, increases incorporation of arachidonic acid into phospholipids, stimulates prostaglandin synthesis, and increases the number of lysosomes in the uterine endometrium.8 Estrogens increase uterine blood flow and may also play a role in fetal organ maturation and development.17 HUMAN CHORIONIC GONADOTROPIN The hCG secreted by the syncytiotrophoblast of the placenta is released into both the fetal and maternal circulation. This hormone is a glycoprotein with a molecular mass of ~38,000 daltons that consists of two noncovalently linked subunits: a and b.18 It has been used extensively as a pregnancy test and can be detected in the serum as early as 6 to 8 days after ovulation. Plasma levels rise rapidly during normal pregnancy, with a doubling in concentration every 2 to 3 days,19 reaching a peak between 60 and 90 days of gestation (Fig. 108-3). Thereafter, the maternal concentration declines and plateaus from ~120 days until delivery.20 The levels of hCG are higher in multiple pregnancies, in pregnancies associated with Rh isoimmunization, and in diabetic women. Levels also are higher in pregnancies associated with hydatidiform moles or in women with choriocarcinoma (see Chap. 112).

FIGURE 108-3. Mean concentration of chorionic gonadotropin (hCG) and placental lactogen (hPL) in sera of women throughout normal pregnancy. (From Pritchard JA, MacDonald PC, Gant NF. Williams obstetrics, 17th ed. Norwalk, CT: Appleton-Century-Crofts, 1985:121.)

There is some evidence that the rate of secretion of hCG is regulated by a paracrine mechanism involving the release of GnRH by the cytotrophoblast.21 Fetal concentrations of hCG reach a peak at 11 to 14 weeks' gestation, thereafter falling progressively until delivery. The most accepted theory regarding the role of hCG in pregnancy is the maintenance of the early corpus luteum to ensure continued progesterone and, possibly, relaxin secretion by the ovary until this function is taken over by the growing trophoblast. Likewise, some investigators have demonstrated that hCG promotes steroidogenesis (progesterone) by the trophoblast.21 Others have suggested a role for hCG in promoting early growth and androgen secretion by the developing fetal zone of the human adrenal gland.22 It is more likely that a primary role for hCG is to regulate the development as well as the secretion of testosterone by the fetal testes. Male sexual differentiation occurs at an early but critical time when hCG is present in fetal serum. At this time, fetal hCG levels are higher—before the vascularization of the fetal pituitary, when fetal plasma luteinizing hormone (LH) levels are low.23 Another role may be to create immunologic privilege to the developing

trophoblast.24 Finally, the excess thyrotropic activity during the clinical development of hyperthyroidism observed in some women with neoplastic trophoblastic disease is secondary to excessive hCG secretion. hCG and TSH have similar structures, and purified hCG inhibits binding to thyroid membranes and stimulates adenylate cyclase in thyroid tissues25 (see Chap. 15 and Chap. 112). HUMAN PLACENTAL LACTOGEN Placental lactogen is a single-chain polypeptide of 191 aminoacid residues with a molecular mass of ~22,000 daltons.26 The hormone has both lactogenic and growth hormone (GH)–like activity and is also referred to as chorionic growth hormone or chorionic somatomammotropin. However, hPL exhibits principally lactogenic activity, having only 3% or less of the growth-stimulating activity of human GH. The amino-acid sequences of hPL and GH are similar,27 and their genes are close together on chromosome 17: It has been proposed that the two hormones evolved from a similar ancestral polypeptide (see Chap. 12 and Chap. 13). The nucleotide sequence for hPL has been reported, and the gene has been cloned.28 hPL is secreted by the syncytiotrophoblast and can be detected in serum by radioimmunoassay as early as the third week after ovulation.26 The serum level of the hormone continues to rise with advancing gestational age and appears to plateau at term (see Fig. 108-3), the concentration closely following and being correlated with increasing placental weight.29 The serum half-life of hPL is short. For example, although the serum level of hPL before delivery is the highest of all the protein hormones secreted by the placenta, hPL cannot be detected after the first postpartum day. The time sequence and peak of hPL secretion are significantly different from those of hCG (see Fig. 108-3), which suggests a different regulation for each hormone. This is interesting, because both are secreted by the syncytiotrophoblast rather than by the cytotrophoblast. Moreover, hPL secretion is not limited to the trophoblast, since immunoreactive hPL has been detected in patients with various malignant tumors including lymphomas, hepatomas, and bronchogenic carcinomas. Interestingly, hPL appears to be secreted primarily into the maternal circulation; only low levels are found in cord blood of neonates. Thus, most of the physiologic roles proposed for hPL have centered on its sites of action in maternal tissues. It has been suggested that hPL has a significant effect on maternal glucose, thereby providing adequate and continued nourishment for the developing fetus.27 It has been proposed that hPL exerts metabolic effects in pregnancy similar to those of GH, including stimulation of lipolysis, thus increasing the circulating free fatty acids available for maternal and fetal nutrition; inhibition of glucose uptake in the mother, yielding increased maternal insulin levels; development of maternal insulin resistance; and inhibition of gluconeogenesis, which favors transportation of glucose and protein to the fetus.30 A few cases of deficient hPL in maternal serum have been described in otherwise normal pregnancies, however, raising issue with this proposed role of hPL.31 HUMAN GROWTH HORMONE VARIANT A “true” placental GH has been shown to be produced by the syncytiotrophoblast of the placenta and secreted in parallel with hPL.32,33,34 and 35 This GH variant is now recognized to be the product of the hGH-V gene34 and differs from the major 22-kDa GH in 13 amino-acid residues.36 A glycosylated variant of this GH form has also been described in an in vitro system,37 but it is not known if this form circulates. Because concentrations of the placental GH variant in maternal plasma correlate with plasma levels of insulin-like growth factor-I (IGF-I), it has been suggested that placental GH is involved in the control of serum IGF-I levels in normal pregnant women.38 OTHER PLACENTAL PEPTIDE HORMONES In addition to hCG and hPL, several other placental hormones that are similar or closely related to hypothalamic, pituitary, or other hormones in their biologic and immunologic activity have been described (e.g., POMC, human chorionic follicle-stimulating hormone [FSH], human chorionic gonadotropin–releasing hormone [hCGnRH], human chorionic thyrotropin [hCT]–releasing hormone, human chorionic corticotropin–releasing hormone, relaxin, somatostatin, gastrin, and vasoactive intestinal peptide). Information regarding these hormones is limited.21 The regulation of their secretion is poorly understood, although it appears that classic negative feedback inhibition does not exist. Furthermore, their function and significance are speculative. Most of these hormones do not cross the placenta and are believed to enter primarily the maternal compartment. HUMAN CHORIONIC PROOPIOMELANOCORTIN PEPTIDES Considerable evidence exists for a chorionic corticotropin or ACTH produced and secreted by placental tissue. Along with ACTH, other products that are processed from a similar 31-kDa POMC peptide are found in placental tissue, including b-endorphin, b-lipotropin, and a-melanocyte–stimulating hormone.21,39,40 HUMAN CHORIONIC THYROTROPIN A substance with TSH-like activity has been identified presumptively in placental tissue. However, the structure of this “hCT” is not identical to that of human pituitary TSH,41 and its physiologic role is unclear. The increased thyroid activity observed in some women with gestational trophoblastic disease is believed to be secondary to the action of excessive hCG secretion and not to hCT. HUMAN CHORIONIC GONADOTROPIN–RELEASING HORMONE AND OTHER HORMONES A substance with bioimmunoreactivity similar to that of hypothalamic GnRH has been localized to and shown to be synthesized by the cytotrophoblast layer of the placenta. It has been proposed that hCG secretion by the syncytiotrophoblast is regulated in part by hCGnRH.21 Similarly, substances similar to thyrotropin-releasing hormone (TRH), somatostatin, and CRH are also synthesized by the trophoblast. CRH mRNA has been localized in the placenta, principally in the cytotrophoblast.42 CRH levels increase in maternal plasma and amniotic fluid throughout pregnancy, but the role for this increase is unclear.43 The FSH-suppressing hormone follistatin has been found in human placenta. Inhibin and activin are secreted by the placenta, and maternal levels increase near term.44 (See Chap. 112 for a discussion of these and other placental hormones.)

FETAL MEMBRANES AND DECIDUA Fetal membranes consisting of amnion and chorion were originally thought to be inactive endocrinologically. The amnion is a thin structure (0.02–0.5 mm) and contains no blood vessels or nerves. However, the fetal membranes play important roles during pregnancy in the transport and metabolism of hormones and in the events that lead eventually to parturition.45 Thus, although fetal membranes apparently do not synthesize hormones de novo, they have extensive enzymatic capabilities for regulating steroid hormone metabolism. Some of these enzymes are 5a-reductase, 3b-hydroxysteroid dehydrogenase, D4,5-isomerase, 20a-hydroxysteroid oxidoreductase, 17b-hydroxysteroid dehydrogenase, aromatase, and sulfatase. Also, fetal membranes contain large quantities of arachidonic acid, the obligate precursor of prostaglandins. Furthermore, they contain phospholipase A2 and other enzymes that stimulate the release of arachidonic acid from glycerophospholipids in the amnion or chorion.8 The decidua is a complex structure of specialized endometrial stromal cells that proliferate in response to progesterone secreted during the luteal phase of the menstrual cycle and later in response to hormones secreted by the developing trophoblast. Evidence suggests that the decidua is also a rich source of enzymatic activity and secretion of hormones. The decidua may be important in fetal homeostasis and in the maintenance of pregnancy, since the decidua appears to communicate directly with the fetus via transport through the fetal membranes and into the amniotic fluid as well as directly into the myometrium by simple diffusion. The hormones and enzymatic activities localized to the decidua include prolactin, relaxin, prostaglandins, and 1a-hydroxylase. The concentration of prolactin in amniotic fluid is extremely high compared with that in fetal or maternal plasma; it arises from the decidua.46 Prolactin is secreted by decidual cells in culture but not by trophoblast or placental membranes. The prolactin secreted by the decidua is immunologically, structurally, and biologically similar to that from pituitary sources.47 However, the regulation of decidual prolactin formation and secretion is more complex. Bromocriptine treatment of pregnant women reduces maternal and fetal plasma levels but not amniotic fluid levels of prolactin. Prolactin secretion by decidual cells or tissues is not affected by treatment with dopamine, dopaminergic agonists, or TRH. The function of decidual prolactin remains speculative. Because most of the prolactin synthesized and secreted by the decidua reaches amniotic fluid, a regulatory role in amniotic fluid osmolality and homeostasis has been proposed.48 Relaxin is a peptide consisting of two chains (A and B) of 22 and 31 amino acids covalently linked.49 Relaxin is secreted by the corpus luteum, decidua, and basal plate and septa of the placenta.50 The greatest source appears to be the corpus luteum of pregnancy, and it is thought to be regulated by hCG. That the decidua and placenta can synthesize relaxin is intriguing because of the proximity of the pregnant uterus. This is relevant because relaxin is believed to play a role along with progesterone in reducing uterine activity as well as in the softening of pelvic tissues and cervix before parturition (see Chap. 94).50

FETAL COMPARTMENT The understanding and elucidation of the human fetal endocrine system have required the development of assays for minute quantities of hormone. The regulation of the fetal endocrine system, like that of the placenta, is not completely independent, since synthesis relies to some extent on precursor hormones secreted directly by the placenta or obtained from the maternal unit. As the fetus develops, its endocrine system becomes more independent in preparation for extrauterine existence. HYPOTHALAMIC-PITUITARY AXIS The fetal hypothalamus begins differentiation from the forebrain during the first few weeks of fetal life, and by 12 weeks, hypothalamic development is well advanced. Most of the hypothalamic-releasing hormones, including GnRH, TRH, dopamine, norepinephrine, and somatostatin, and their respective hypothalamic nuclei have been identified as early as 6 to 8 weeks of fetal life.23 The neurohypophysis is detected first at 5 weeks, and by 14 weeks, the supraoptic and paraventricular nuclei are fully developed.23,51 Rathke pouch appears in the human fetus at 4 weeks. The premature anterior pituitary cells that develop from the cells lining Rathke pouch can secrete GH, prolactin, FSH, LH, and ACTH in vitro as early as 7 weeks of fetal life.52 Evidence suggests that the intermediate lobe of the pituitary is a significant source of POMC hormones.53,54 The hypothalamic-pituitary portal system is the functional link between the hypothalamus and the anterior pituitary. Vascularization of the anterior pituitary begins by 13 weeks of fetal life, but a functioning intact portal system is absent until 18 to 20 weeks.55 However, there is indirect evidence that hypothalamic secretion of releasing hormones influences anterior pituitary function before this time by simple diffusion, given their proximity in early fetal development. There is also evidence that fetal adrenal feedback is operative at the hypothalamic-pituitary axis as early as week 14 of fetal life. Elevated levels of fetal androgens are detected at this time in amniotic fluid in fetuses affected with congenital adrenal hyperplasia secondary to 21-hydroxylase deficiency.56 GROWTH HORMONE, PROLACTIN, VASOPRESSIN, AND OXYTOCIN GH is detected in the fetal pituitary as early as 12 weeks' gestation, and fetal pituitary GH concentrations increase until 25 to 30 weeks' gestation, thereafter remaining constant until term. Fetal plasma GH levels peak at 20 weeks and then fall rapidly until birth57 (Fig. 108-4). However, fetal plasma concentrations of GH always exceed maternal concentrations, which are suppressed, possibly by the high circulating levels of hPL.

FIGURE 108-4. Ontogeny of pituitary hormones in human fetal serum. (Prl, prolactin; TSH, thyroid-stimulating hormone; ACTH, adrenocorticotropic hormone; GH, growth hormone; LH, luteinizing hormone; FSH, follicle-stimulating hormone.) (From Parker CR Jr. The endocrinology of pregnancy. In: Carr BR, Blackwell RE, eds. Textbook of reproductive medicine. Norwalk, CT: Appleton & Lange, 1993:17.)

The regulation of GH release in the fetus appears to be more complex than in the adult. To explain the high levels of plasma GH at midgestation and a fall thereafter, unrestrained release of growth hormone–releasing hormone (GHRH) leading to excessive release of GH at midgestation has been postulated.23 Thereafter, as the hypothalamus matures, somatostatin may increase and GHRH levels decline, reducing GH release. The role of GH in the fetus is also unclear. There is considerable evidence that GH is not essential to somatic growth in primates.58 For example, in neonates with pituitary agenesis, congenital hypothalamic hypopituitarism, or familial GH deficiency, birth size and length are usually normal. However, the somatomedins, in particular IGF-I and IGF-II, increase in fetal plasma, and IGF-I and IGF-II levels correlate better than do GH levels with fetal growth.59 Although GH is an important trophic hormone for somatomedin production in the fetus, somatomedin regulation may be independent of GH, depending instead on other factors. Prolactin, hPL, and insulin also stimulate somatomedin production.23 Prolactin is present in lactotropes by 19 weeks of life. The prolactin content of the fetal pituitary increases throughout gestation,60 whereas plasma levels increase slowly until 30 weeks' gestation, when the levels rise sharply until term and remain elevated until the third month of postnatal life (see Fig. 108-4). In humans, TRH and dopamine as well as estrogens appear to affect fetal prolactin secretion. Bromocriptine both lowers maternal prolactin levels and crosses the placenta to inhibit fetal prolactin release and lower prolactin levels in fetal blood.61 It has been suggested that prolactin in the fetus influences adrenal growth, lung maturation, and amniotic fluid volume. Arginine vasopressin (AVP) and oxytocin are found in hypothalamic nuclei and in the neurohypophysis during early fetal development.52 However, there are relatively few studies on the regulation and secretion of these hormones. The levels of AVP are high in fetal plasma and cord blood at delivery. The principal stimulus to AVP release appears to be fetal hypoxia, although acidosis, hypercarbia, and hypotension also play a role.62 The elevated AVP levels in fetal blood may lead to increased blood pressure, vasoconstriction, and the passage of meconium by the fetus. Oxytocin levels in the fetus are not affected by hypoxia but appear to increase during labor and delivery (see Chap. 109). THYROID GLAND The placenta apparently is relatively impermeable to TSH and thyroid hormone, so that the fetal hypothalamic–pituitary– thyroid axis develops and functions independently of the maternal system. Although thyroxine (T4) may cross the placenta to a slight degree, human thyroid-stimulating immunoglobulins (TSI) as well as iodine and antithyroid drugs given to women with hyperthyroidism pass through the placenta and may affect fetal thyroid function.63 TRH is detectable in hypothalamic nuclei, and TSH is found in pituitary tissues by 10 to 12 weeks of fetal life.64 High concentrations of TRH are detected in fetal blood, and the source is thought to be the fetal pancreas. However, this source of TRH appears to have little effect on the pituitary release of TSH, and the function of pancreatic TRH is unknown.65 The fetal thyroid has developed sufficiently by the end of the first trimester that it can concentrate iodine and synthesize iodothyronines. The levels of TSH and thyroid hormone are relatively low in fetal blood until midgestation. At 24 to 28 weeks' gestation, serum TSH concentrations rise abruptly to a peak but decrease slightly thereafter until delivery.66 In response to the surge of TSH, T 4 levels rise progressively after midgestation until term (Fig. 108-5). During this time, both thyroid responsiveness to TRH and pituitary TRH content increase. At birth, there is an abrupt release of TSH, T4, and triiodothyronine (T3), and the levels of these hormones fall during the first few weeks after birth. 66 The relative hyperthyroid state of the fetus is believed to be necessary to prepare it for the thermoregulatory adjustments of extrauterine life. The abrupt changes of TSH and T4 that occur at birth are believed to be stimulated by the cooling associated with delivery.67 Finally, 3,3',5'-triiodo-L thyronine (reverse T3) levels are high during early fetal life, begin to fall at midgestation, and continue to fall after birth (see Chap. 47). The difference between the formation of T3 and reverse T3 is thought to be related to maturation of peripheral iodothyronine metabolism (see Chap. 30).

FIGURE 108-5. Maturation of serum thyroid-stimulating hormone (TSH), thyroxine (T4), and triiodothyronine (T3) during the last half of gestation and early neonatal life. The increase in thyrotropin-releasing hormone (TRH) effect or content is also illustrated. (From Fisher DA. Maternal-fetal neurohypophyseal system. Clin Perinatol 1983; 10:615.)

GONADS Bioactive and immunoreactive GnRH has been detected in the fetal hypothalamus by 9 to 12 weeks of life. The amount increases with fetal age, with the maximum noted between 22 and 25 weeks in females and between 34 to 38 weeks in males.23 The dominant gonadotropin fraction in the fetal pituitary is the subunit. However, the fetal pituitary in vitro is capable of secreting intact LH by 5 to 7 weeks.53 The pituitary content of LH increases from 10 weeks to 24 weeks and then falls slowly near term. The content of LH is higher in females than in males, a difference thought to be secondary to a greater negative feedback in response to higher concentrations of plasma testosterone in fetal male plasma.68 The FSH content of the fetal pituitary increases until midgestation, then falls until term. The FSH content is higher in female than in male fetuses because of greater negative feedback in the latter. The plasma concentration of FSH rises slowly to a peak near week 25 and then falls to low levels by term (see Fig. 108-4). The FSH levels parallel the pituitary content of FSH with respect to sexual dimorphism, being higher in females than in males. The pattern of LH levels in fetal plasma parallels that of FSH. The fall in gonadotropin pituitary content and plasma concentration after midgestation is thought to be attributable to the maturation of the hypothalamus.23 The hypothalamus also becomes more sensitive to sex steroids circulating in fetal blood that originate in the placenta. The differentiation of the bipotential fetal gonad into a testis or an ovary is discussed in Chapter 94. In the male, testosterone secretion begins soon after differentiation of the gonad into a testis and the formation of Leydig cells at 7 weeks of fetal life. Maximal levels are observed at ~15 weeks and decrease thereafter.69 This early secretion of testosterone is important in regulating sexual differentiation. It is believed that hCG is the primary stimulus to the early development and growth of Leydig cells and the subsequent peak of testosterone. The pattern of hCG, the concentration of testicular hCG receptors, and the pattern of plasma testosterone are related closely.70 Thus, it appears that sexual differentiation of the male does not rely solely on fetal pituitary gonadotropins. However, fetal LH and FSH are still required for complete differentiation of the fetal ovary and testis. For example, anencephalic fetuses with low levels of circulating LH and FSH have appropriate secretion of testosterone at 15 to 20 weeks secondary to adequate levels of hCG, but they have a decreased number of Leydig cells, exhibit hypoplastic external genitalia, and often have undescended testes.71 Likewise, male fetuses with congenital hypopituitarism often have an associated micropenis. These observations suggest that beginning about midgestation fetal pituitary gonadotropins affect testosterone secretion from the testes. The regulation of testosterone secretion also appears to depend on fetal cholesterol, principally LDL cholesterol, to maintain maximal rates of testosterone secretion. Fetal testicular LDL receptors and rates of de novo synthesis of cholesterol also parallel the secretion of hCG and testosterone.72 The fetal ovary is involved primarily with the formation of follicles and germ cells. Although follicular development appears to be relatively independent of gonadotropins, the anencephalic female fetus has small ovaries and a decreased number of ovarian follicles. However, the fetal ovaries do not contain hCG receptors, at least by 20 weeks of gestation. The ovaries appear to be relatively inactive with respect to steroidogenesis during fetal life, but they can aromatize androgens to estrogens in vitro as early as 8 weeks of life.73 ADRENAL GLAND The human fetal adrenal glands secrete large quantities of steroid hormones (up to 200 mg per day near term).74 This rate of steroidogenesis may be five times that observed in the adrenal glands of adults at rest. The principal steroids are C-19 steroids (mainly DHEAS), which serve as substrate for estrogen biosynthesis in the placenta. It was recognized early that the human fetal adrenal gland contains a unique fetal zone that accounts for the rapid growth of the gland and that this zone disappears during the first few weeks after birth.12 The fetal zone differs histologically and biochemically from the neocortex (also known as the definitive or adult zone). The uniqueness of a transient fetal zone has been reported in certain higher primates and some other species, but only humans possess the extremely large fetal zone that involutes after birth. The cells of the adrenal cortex arise from coelomic epithelium. Those cells comprising the fetal zone can be identified in the 8- to 10-mm embryo and before the appearance of the cells of the neocortex (14-mm embryo).75 Growth is most rapid during the last 6 weeks of fetal life. By 28 weeks' gestation, the adrenal gland may be as large as the fetal kidney and may be equal to the size of the adult adrenal by term (Fig. 108-6). The fetal zone accounts for the largest percentage of growth; after birth, the gland shrinks secondary to involution and necrosis of fetal zone cells.

FIGURE 108-6. Size of adrenal gland and its component parts during fetal life, infancy, and childhood. (From Carr BR, Simpson ER. Lipoprotein utilization and cholesterol synthesis by the human fetal adrenal gland. Endocr Rev 1981; 2:306.)

Histologically, the central portion of the adrenal contains the fetal zone cells, which are eosinophilic cells with palestaining nuclei that at term make up 80% to 85% of the volume of the gland. The neocortex is the outer rim of cells containing a small quantity of cytoplasm and dark-staining nuclei. The neocortex is thought to originate the zona glomerulosa, zona fasciculata, and zona reticularis after birth.12 In vitro studies utilizing fetal adrenal tissues or cells, in vivo perfusion studies of previable fetuses, and cord blood measurements of steroid hormones demonstrate that the fetal zone can secrete the full complement of steroid hormones secreted by the adult adrenal cortex. However, the fetal zone has a reduced capacity to secrete C-21 steroids because of the low activity of 3b-hydroxysteroid dehydrogenase and D4,5-isomerase complex,76 probably secondary to estrogen or other factors produced by the placenta that may inhibit enzyme action. Thus, the principal steroids secreted by the fetal zone cells are D5-sulfoconjugates, namely, DHEAS and pregnenolone sulfate.77,78 In contrast, the principal secretory product of the neocortex is cortisol. Electron microscopic investigations suggest fetal zone activity as early as the seventh week and indicate that it is the most steroidogenic zone throughout gestation. The neocortex cells exhibit little steroidogenic activity until the third trimester.79

During gestation, DHEAS levels in fetal plasma rise, peaking between 34 and 40 weeks.80 This pattern coincides with the marked increase in fetal adrenal growth. After birth, DHEAS levels decline, paralleling the regression of the fetal zone. Cortisol plasma levels also increase during fetal life, but there is little evidence after 25 weeks' gestation of a sharp rise like that of DHEAS. Moreover, a significant portion of the circulating cortisol in fetal plasma arises from placental transfer from the maternal compartment.81 REGULATION OF FETAL ADRENAL GROWTH ACTH stimulates steroidogenesis in vitro,12 and there is clinical evidence that ACTH is the principal trophic hormone of the fetal adrenal gland in vivo. For example, in anencephalic fetuses, the plasma levels of ACTH are very low, and the fetal zone is markedly atrophic. Maternal glucocorticosteroid therapy suppresses fetal adrenal steroidogenesis by suppressing fetal ACTH secretion.82 Further evidence that ACTH regulates steroidogenesis early in fetal life is provided by the observation of elevated levels of 17a-hydroxyprogesterone in the amniotic fluid of fetuses with congenital adrenal hyperplasia secondary to the absence of 21-hydroxylase. Despite these observations, other ACTH-related peptides (e.g., fetal pituitary or placental POMC derivatives) have been proposed as trophic hormones for the fetal zone, but the evidence is weak.12 Other hormones or growth factors, including prolactin, hCG, GH, hPL, and epidermal and fibroblast growth factor, have no consistent significant effect on steroidogenesis or adenylate cyclase activity in cultures of fetal zone organs or monolayer cells or of membrane preparations. However, a role of these or other hormones in promoting growth of adrenal cells is possible.83,84 After birth, the adrenal gland shrinks by more than 50% secondary to regression of fetal zone cells. This suggests that a trophic substance other than ACTH is withdrawn from the maternal or placental compartment or that the secretion rates of some other trophic hormone are altered to initiate regression of this fetal zone. FETAL ADRENAL STEROIDOGENESIS The availability of precursor substrates may assist ACTH in regulating the rate of steroid hormone production by the fetal zone. Circulating pregnenolone and progesterone have long been suggested as the principal precursors of fetal adrenal steroidogenesis, but a number of factors make this unlikely. For example, in view of the fetal adrenal blood flow and the levels of unconjugated pregnenolone in fetal plasma, 70 different cancer cell lines have been shown to contain them.25,75,76,77 and 78,78a The presence of hCG and/or one of its subunits in cancer cells is probably due to synthesis rather than sequestration. The regulatory mechanisms involved in the expression of hCG-subunit genes in cancer cells is not known. Ectopic production of hCG is considered a recapitulation of the embryonic state, as is cancer. The expression of hCG and/or one of its subunits increases in advanced cancers, suggesting that they might be involved in the progression of the disease. In fact, contraceptive hCG vaccine is now being tested, especially against cancers of the colon and pancreas.79,80 and 81 Like other hormones, hCG acts via binding to its receptors. A demonstration of hCG/LH receptors in cancers of nongonadal tissues has reinforced a belief that hCG may indeed play a role. In fact, studies suggest that hCG may have dual roles in cancers. It promotes some cancers (endometrial cancer,82,83,84,85 and 86 choriocarcinomas4,5 [see Chap. 111], and lung cancer87,88), whereas it inhibits others (prostate cancer89,90,91 and 92 and breast cancer93,94,95 and 96). Some controversies on whether hCG prevents or promotes cancers could be due to whether they produce intact hCG or just its b subunit, which may have a stimulatory effect, probably due to the formation of homodimers.97,98 and 99 When intact hCG or LH promotes cancer, its presence in cancer tissues can be expected to be associated with a poor prognosis. When these hormones protect against cancer, their presence indicates a good prognosis. In the latter case, injection of the hormone into the lesion may slow the cancer progression. Gestational trophoblastic neoplasms (GTNs) contain high hCG/LH receptor levels, which further increase in more malignant phenotypes such as choriocarcinomas.100 These high receptor levels are due to a loss of self-regulation of hCG biosynthesis. This may explain how GTNs can produce much higher levels of hCG than do normal trophoblasts.101 The high hCG levels produced by choriocarcinomas may promote their growth, development, and metastasis in the host body.5 hCG-producing tumors (see Chap. 120 and Chap. 219) in young boys can cause precocious puberty by virtue of constant stimulation of the testis to produce testosterone102 (see Chap. 92). Such tumors in young girls usually do not have obvious adverse effects unless they are associated with the ovaries. The reason for this gender difference is that both LH and FSH are required for ovarian estradiol synthesis, whereas LH alone is capable of stimulating testicular synthesis of testosterone. This gender difference is also seen when there is an activating LH receptor mutation.103 hCG-producing tumors in some men can cause gynecomastia, possibly due to direct actions of hCG on the breast. hCG-producing tumors in women cause disruption of the menstrual cycle and dysfunctional uterine bleeding. POTENTIAL THERAPEUTIC USES OF HUMAN CHORIONIC GONADOTROPIN There are several potential therapeutic uses of hCG. Since hCG has pervasive actions during pregnancy, some unexplained pregnancy losses could be due to aberrant or inadequate actions of hCG; this may be corrected by the administration of hCG. hCG levels progressively decrease during threatened abortion, but whether this is a cause or consequence is not known. Administration of hCG also may help in some of these cases. hCG treatment may work by increasing the placental endocrine activity, by preventing immunologic mechanisms that promote fetal rejection, by increasing uterine blood flow, by decreasing uterine activity, and so forth.49 This treatment may not work if infection, anatomic defects, fetal anomalies, and so forth, are responsible for these conditions. The ability of hCG to maintain myometrial quiescence suggests it may be used in the treatment of preterm labor and delivery, unless it is caused by infection, premature rupture of membranes, and so forth. In fact, administration of hCG has a tocolytic effect in a mouse preterm-labor model. 104 If it is proved that hCG works in women, it would be the most natural means of preventing preterm labor. The rationale for giving hCG when women already have it in their circulation is that perhaps their levels are not adequate, and increasing levels by giving exogenous hormone might delay events that lead to preterm labor and delivery. Epidemiologic data, the rat breast cancer model studies, and the anticancer effects of hCG in human breast cancer cells suggest that the decreased incidence of

breast cancer in women who complete a full-term pregnancy before 20 years of age could be due to hCG.105,106,107 and 108 This hormone may act on breasts to promote nonreversible differentiation of proliferation-competent epithelial cells into secretory cells in terminal end buds. Coincidentally, this differentiation, which is a physiologic phenomenon to prepare the breast for lactation, also makes the cells less susceptible to carcinogenic transformation.109 Additional mechanisms, such as inhibition of cell growth and invasion,94,95,110 increase of apoptosis,111,112 and the cell's ability to repair DNA damage, also may play roles in the protective actions of hCG in the breast.113,114 Increased inhibin and insulin-like growth factor (IGF)–binding proteins and/or decreased IGF-I and its receptors may mediate hCG actions in breast cancer cells.109,115,116 These findings do not necessarily mean that every woman who completes a full-term pregnancy at a young age will never get breast cancer. Several other factors, such as family history, radiation, environment, and so forth, contribute to the development of this disease; therefore, pregnancy may not be able to overcome some or all of these factors. Potential uses of hCG in the treatment of HIV infections and Kaposi sarcomas are controversial.117,118 Nonetheless, these treatment strategies may have some merit because of the reported antiviral properties of hCG as well as its ability to act on cells of the immune system and numerous other target tissues throughout the body.46,49,67,119,120 and 121 SUMMARY AND PERSPECTIVES Trophoblastic tissue is a transient and unique endocrine organ that is capable of producing a vast array of bioactive substances. Among them, hCG is the best known and perhaps most important. hCG is not just a gonadal-regulating hormone as once was believed. It is a pluripotent regulatory molecule with the actions of a classic hormone, a growth factor, and a cytokine. It plays a pivotal role in the regulation of the functions of the fetoplacental unit and of a number of other tissues during pregnancy. Although the evolutionary significance of the broad spectrum of hCG actions is unknown, it could have evolved to orchestrate numerous functions during pregnancy in women. LH may fulfill some of the roles of hCG in other species. These farranging actions are not unique to hCG, as prolactin is another example of a hormone having multiple targets in the body. hCG research has helped to explain many unknown, and to rationalize several known, effects of hCG. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.

Bernischke K, Kaufmann P. Pathology of the human placenta, 3rd ed. New York: Springer-Verlag. 1995. Rodway MR, Rao ChV. A novel perspective on the role of human chorionic gonadotropin during pregnancy and in gestational trophoblastic disease. Early Pregnancy Biol Med 1995; 1(3):176. El-Hendy KA, Subramanian MG, Diamond MP, Yelian FD. hCG modulates MMP-9 activity in first trimester trophoblast cells. J Soc Gynecol Invest 1998; 5:118A. Zygmunt M, Hahn D, Munstedt K, et al. Invasion of cytotrophoblastic JEG-3 cells is stimulated by hCG in vitro. Placenta 1998; 19(8):587. Lei ZM, Taylor DD, Gercel-Taylor C, Rao ChV. Human chorionic gonadotropin promotes tumorigenesis of choriocarcinoma JAR cells. Trophoblast Res 1999; 13:147. Lei ZM, Rao ChV, Kornyei JL, et al. Novel expression of human chorionic gonadotropin/luteinizing hormone receptor gene in brain. Endocrinology 1993; 132(5):2262. Kliman HJ. Placental hormones. Endocrinol Pregnancy 1994; 5(4):591. Petraglia F, Florio P, Nappi C, Genazzani AR. Peptide signaling in human placenta and membranes: autocrine, paracrine, and endocrine mechanisms. Endocr Rev 1996; 17(2):156. Rao ChV. Potential novel roles of luteinizing hormone and human chorionic gonadotropin during early pregnancy in women. Early Pregnancy Biol Med 1997; 3(1):1. Walsh SW. Prostaglandins in pregnancy. In: Sciarra JJ, ed. Gynecology and obstetrics. Philadelphia: JB Lippincott Co, 1992. Reyes FI. Protein hormones of the placenta. In: Sciarra JJ, ed. Gynecology and obstetrics. Philadelphia: JB Lippincott Co, 1992. Guilbert L, Robertson SA, Wegmann TG. The trophoblast as an integral component of a macrophage-cytokine network. Immunol Cell Biol 1993; 71(Pt 1):49. Senaris R, Garcia-Caballero T, Casabiell X, et al. Synthesis of leptin in human placenta. Endocrinology 1997; 138(10):4501. Conrad KP, Benyo DF. Placental cytokines and the pathogenesis of preeclampsia. Am J Reprod Immunol 1997; 37(3):240. Petraglia F, Santuz M, Florio P, et al. Paracrine regulation of human placenta: control of hormonogenesis. J Reprod Immunol 1998; 39(1– 2):221. He Y, Smith SK, Day KA, et al. Alternative splicing of vascular endothelial growth factor (VEGF)-R1 (FLT-1) pre-mRNA is important for the regulation of VEGF activity. Mol Endocrinol 1999; 13(4):537. Shi QJ, Lei ZM, Rao ChV, Lin J. Novel role of human chorionic gonadotropin in differentiation of human cytotrophoblasts. Endocrinology 1993; 132(3):1387. Sawai K, Azuma C, Koyama M, et al. Leukemia inhibitory factor (LIF) enhances trophoblast differentiation mediated by human chorionic gonadotropin (hCG). Biochem Biophys Res Commun 1995; 211(1):137. Sawai K, Matsuzaki N, Kameda T, et al. Leukemia inhibitory factor produced at the fetomaternal interface stimulates chorionic gonadotropin production: its possible implication during pregnancy, including implantation period. J Clin Endocrinol Metab 1995; 80(4):1449. Mochizuki M, Maruo T, Matsuo H, et al. Biology of human trophoblast. Int J Gynaecol Obstet 1998; 60(Suppl 1):S21. Morrish DW, Dakour J, Li H. Functional regulation of human trophoblast differentiation. J Reprod Immunol 1998; 39(1–2):179. Yang M, Lei ZM, Rao ChV. Mechanism of leukemia inhibitory factor induced differentiation of human cytotrophoblasts into syncytiotropho-blasts. In: The Endocrine Society Annual Meeting, San Diego, CA; 1999. Abstract P1–6. Yang M, Lei ZM, Rao ChV. How does epidermal growth factor promote the differentiation of human cytotrophoblasts into syncytiotrophoblasts. J Soc Gynecol Invest 1999; 6(Suppl 1):Abstract 31. Lei ZM, Toth P, Rao ChV, Pridham D. Novel coexpression of human chorionic gonadotropin (hCG)/human luteinizing hormone receptors and their ligand hCG in human fallopian tubes. J Clin Endocrinol Metab 1993; 77:863. Acevedo HF, Hartsock RJ, Maroon JC. Detection of membrane-associated human chorionic gonadotropin and its subunits on human cultured cancer cells of the nervous system. Cancer Detect Prev 1997; 21(4):295. Pierce JG, Parsons TF. Glycoprotein hormones: structure and function. Annu Rev Biochem 1981: 50:465. Iles RK, Chard T. Molecular insights into the structure and function of human chorionic gonadotrophin. J Mol Endocrinol 1993; 10(3):217. Lapthorn AJ, Harris DC, Littlejohn A, et al. Crystal structure of human chorionic gonadotropin. Nature 1994; 369(6480):455. Fiddes JC, Goodman HM. The gene encoding the common alpha subunit of the four human glycoprotein hormones. J Mol Appl Genet 1981; 1(1):3. Boorstein WR, Vamvakopoulos NC, Fiddes JC. Human chorionic gonadotropin beta-subunit is encoded by at least eight genes arranged in tandem and inverted pairs. Nature 1982; 300(5891):419. Policastro P, Ovitt CE, Hoshina M, et al. The beta subunit of human chorionic gonadotropin is encoded by multiple genes. J Biol Chem 1983; 258(19):11492. Bo M, Boime I. Identification of the transcriptionally active genes of the chorionic gonadotropin beta gene cluster in vivo. J Biol Chem 1992; 267(5):3179. Talmadge K, Vamvakopoulos NC, Fiddes JC. Evolution of the genes for the beta subunits of human chorionic gonadotropin and luteinizing hormone. Nature 1984; 307(5946):37. Fiddes JC, Goodman HM. The cDNA for the beta-subunit of human chorionic gonadotropin suggests evolution of a gene by readthrough into the 3'-untranslated region. Nature 1980; 286(5774):684. Albanese C, Colin IM, Crowley WF, et al. The gonadotropin genes: evolution of distinct mechanisms for hormonal control. Recent Prog Horm Res 1996; 51:23. Jameson JL, Hollenberg AN. Regulation of chorionic gonadotropin gene expression. Endocr Rev 1993; 14(2):203. Liu L, Roberts RM. Silencing of the gene for the beta subunit of human chorionic gonadotropin by the embryonic transcription factor Oct-3/4. J Biol Chem 1996; 271(28):16683. Liu L, Leaman D, Villalta M, Roberts RM. Silencing of the gene for the alpha-subunit of human chorionic gonadotropin by the embryonic transcription factor Oct-3/4. Mol Endocrinol 1997; 11(11):1651. Fuh VL, Burrin JM, Jameson JL. Cyclic AMP (cAMP) effects on chorionic gonadotropin gene transcription and mRNA stability: labile proteins mediate basal expression whereas stable proteins mediate cAMP stimulation. Mol Endocrinol 1989; 3(7):1148. Licht P, Cao H, Lei ZM, et al. Novel self-regulation of human chorionic gonadotropin biosynthesis in term pregnancy human placenta. Endocrinology 1993; 133(6):3014. Merz WE. Biosynthesis of human chorionic gonadotropin: a review. Eur J Endocrinol 1996; 135(3):269. Blithe DL, Richards RG, Skarulis MC. Free alpha molecules from pregnancy stimulate secretion of prolactin from human decidual cells: a novel function for free alpha in pregnancy. Endocrinology 1991; 129(4):2257. Moy E, Kimzey LM, Nelson LM, Blithe DL. Glycoprotein hormone alpha-subunit functions synergistically with progesterone to stimulate differentiation of cultured human endometrial stromal cells to decidualized cells: a novel role for free alpha-subunit in reproduction. Endocrinology 1996; 137(4):1332. Wolkersdorfer GW, Bornstein SR, Hilbers U, et al. The presence of chorionic gonadotrophin beta subunit in normal cyclic human endometrium. Mol Hum Reprod 1998; 4(2):179. Ringler GE, Strauss JF. In vitro systems for the study of human placental endocrine function. Endocr Rev 1990; 11(1):105. Rao ChV. The beginning of a new era in reproductive biology and medicine: expression of low levels of functional luteinizing hormone/human chorionic gonadotropin receptors in nongonadal tissues. J Physiol Pharmacol 1996; 47(Suppl 2):41. Rao ChV. Novel concepts in neuroendocrine regulation of reproductive tract functions. In: Bazer FW, ed. The endocrinology of pregnancy. Totowa, NJ: Humana Press, 1998:125. Ziecik AJ, Derecka-Reszka K, Rzucidlo SJ. Extragonadal gonadotropin receptors, their distribution and function. J Physiol Pharmacol 1992; 43(4 Suppl 1):33. Rao ChV. A paradigm shift on the targets of luteinizing hormone/human chorionic gonadotropin actions in the body. J Bellevue Obstet Gynecol Soc 1999; 15:26. Lei ZM, Rao ChV. Signaling and transacting factors in the transcriptional inhibition of gonadotropin releasing hormone gene by human chorionic gonadotropin in immortalized hypothalamic GT1-7 neurons. Mol Cell Endocrinol 1995; 109(2):151. Lei ZM, Rao ChV. Cis-acting elements and trans-acting proteins in the transcriptional inhibition of gonadotropin-releasing hormone gene by human chorionic gonadotropin in immortalized hypothalamic GT1-7 neurons. J Biol Chem 1997; 272(22):14365. Toth P, Lukacs H, Hiatt ES, et al. Administration of human chorionic gonadotropin affects sleepwake phases and other associated behaviors in cycling female rats. Brain Res 1994; 654(2):181. Lukacs H, Hiatt ES, Lei ZM, Rao ChV. Peripheral and intracerebroventricular administration of human chorionic gonadotropin alters several hippocampus- associated behaviors in cycling female rats. Horm Behav 1995; 29(1):42. Han SW, Lei ZM, Rao ChV. Up-regulation of cyclooxygenase-2 gene expression by chorionic gonadotropin during the differentiation of human endometrial stromal cells into decidua. Endocrinology 1996; 137(5):1791. Han SW, Lei ZM, Rao ChV. Treatment of human endometrial stromal cells with chorionic gonadotropin promotes their morphological and functional differentiation into decidua [In Process Citation]. Mol Cell Endocrinol 1999; 147(1–2):7. Ambrus G, Rao ChV. Novel regulation of pregnant human myometrial smooth muscle cell gap junctions by human chorionic gonadotropin. Endocrinology 1994; 135(6):2772. Eta E, Ambrus G, Rao ChV. Direct regulation of human myometrial contractions by human chorionic gonadotropin. J Clin Endocrinol Metab 1994; 79(6):1582. Toth P, Li X, Rao ChV, et al. Expression of functional human chorionic gonadotropin/human luteinizing hormone receptor gene in human uterine arteries. J Clin Endocrinol Metab 1994; 79(1):307. Toth P, Gimes G, Rao ChV. hCG treatment in early gestation: its impact on uterine blood flow and pregnancy outcome. In: 16th World Congress on Fertility and Sterility and 54th Annual Meeting of the American Society for Reproductive Medicine; 1998, San Francisco, CA; Abstract S46. Rao ChV, Li X, Toth P, et al. Novel expression of functional human chorionic gonadotropin/luteinizing hormone receptor gene in human umbilical cords. J Clin Endocrinol Metab 1993; 77(6):1706. Toth P, Li X, Lei ZM, Rao ChV. Expression of human chorionic gonadotropin (hCG)/luteinizing hormone receptors and regulation of the cyclooxygenase- 1 gene by exogenous hCG in human fetal membranes. J Clin Endocrinol Metab 1996; 81(3):1283. McGregor WG, Kuhn RW, Jaffe RB. Biologically active chorionic gonadotropin: synthesis by the human fetus. Science 1983; 220(4594):306.

63. Dobozy O, Brindak O, Csaba G. Influence of pituitary hormones (hCG, TSH, Pr, GH) on testosterone level and on the functional activity of the Leydig cell in rat fetuses. Acta Physiol Hung 1988; 72(2):159. 64. Al-Hader AA, Lei ZM, Rao ChV. Novel expression of functional luteinizing hormone/chorionic gonadotropin receptors in cultured glial cells from neonatal rat brains. Biol Reprod 1997; 56(2):501. 65. Al-Hader AA, Lei ZM, Rao ChV. Neurons from fetal rat brains contain functional luteinizing hormone/chorionic gonadotropin receptors. Biol Reprod 1997; 56(5):1071. 66. Al-Hader AA, Tao YX, Lei ZM, Rao ChV. Fetal rat brains contain luteinizing hormone/human chorionic gonadotropin receptors. Early Pregnancy Biol Med 1997; 3(4):323. 67. Harris PJ. Human chorionic gonadotropin hormone is antiviral. Med Hypotheses 1996; 47(2):71. 68. De SK, Wohlenberg CR, Marinos NJ, et al. Human chorionic gonadotropin hormone prevents wasting syndrome and death in HIV-1 transgenic mice. J Clin Invest 1997; 99(7):1484. 69. Shapira A, Bao S, Lei ZM, et al. Treatment of homozygous HIV-1 transgenic mouse pups with human chorionic gonadotropin (hCG) upregulates the skin hCG/luteinizing hormone receptor levels. In the program of The Endocrine Society Annual Meeting, 1998, Abstract P2–149. 70. Zuo J, Lei ZM, Rao ChV. Human myometrial chorionic gonadotropin/luteinizing hormone receptors in preterm and term deliveries. J Clin Endocrinol Metab 1994; 79(3):907. 71. Bird J, Li X, Lei ZM, et al. Luteinizing hormone and human chorionic gonadotropin decrease type 2 5 alpha-reductase and androgen receptor protein levels in women's skin. J Clin Endocrinol Metab 1998; 83(5):1776. 72. Tao YX, Heit M, Lei ZM, Rao ChV. The urinary bladder of a woman is a novel site of luteinizing hormone–human chorionic gonadotropin receptor gene expression. Am J Obstet Gynecol 1998; 179(4):1026. 73. Thompson DA, Othman MI, Lei ZM, et al. Localization of receptors for luteinizing hormone/chorionic gonadotropin in neural retina. Life Sci 1998; 63(12):1057. 74. Jacobs EL, Haskell CM. Clinical use of tumor markers in oncology. Curr Probl Cancer 1991; 15(6):299. 75. Acevedo HF, Krichevsky A, Campbell-Acevedo EA, et al. Expression of membrane-associated human chorionic gonadotropin, its subunits, and fragments by cultured human cancer cells. Cancer 1992; 69 (7):1829. 76. Acevedo HF, Krichevsky A, Campbell-Acevedo EA, et al. Flow cytometry method for the analysis of membrane-associated human chorionic gonadotropin, its subunits, and fragments on human cancer cells. Cancer 1992; 69(7):1818. 77. Acevedo HF, Tong JY, Hartsock RJ. Human chorionic gonadotropin-beta subunit gene expression in cultured human fetal and cancer cells of different types and origins [see comments]. Cancer 1995; 76(8):1467. 78. Lazar V, Diez SG, Laurent A, et al. Expression of human chorionic gonadotropin beta subunit genes in superficial and invasive bladder carcinomas. Cancer Res 1995; 55(17):3735. 78a. Fujikawa K, Matsui Y, Oka H, et al. Prognosis of primary testicular seminoma: a report of 57 new cases. Cancer Res 2000; 60:2152. 79. Triozzi PL, Martin EW, Gochnour D, Aldrich W. Phase Ib trial of a synthetic beta human chorionic gonadotropin vaccine in patients with metastatic cancer. Ann N Y Acad Sci 1993; 690:358. 80. Triozzi PL, Stevens VC, Aldrich W, et al. Effects of a beta-human chorionic gonadotropin subunit immunogen administered in aqueous solution with a novel nonionic block copolymer adjuvant in patients with advanced cancer. Clin Cancer Res 1997; 3(12 Pt 1):2355. 81. Triozzi PL, Stevens VC. Human chorionic gonadotropin as a target for cancer vaccines. Oncol Rep 1999; 6(1):7. 82. Lin J, Lei ZM, Lojun S, et al. Increased expression of luteinizing hormone/human chorionic gonadotropin receptor gene in human endometrial carcinomas. J Clin Endocrinol Metab 1994; 79(5):1483. 83. Bax CR, Chatzaki E, Davies S, Gallagher CJ. Elucidating the role of gonadotropins in endometrial cancer cell growth. Biochem Soc Trans 1996; 24:443S. 84. Konishi I, Koshiyama M, Mandai M, et al. Increased expression of LH/hCG receptors in endometrial hyperplasia and carcinoma in anovulatory women. Gynecol Oncol 1997; 65(2):273. 85. Nagamani M, Cao HA. Specific binding and proliferative effects of luteinizing hormone in human endometrial cancer cell lines. J Soc Gynecol Invest 1997; 4:132A. 86. Han SW, Zhou XL, Lei ZM, Rao ChV. Role of luteinizing hormone in human endometrial carcinoma. In: The Endocrine Society Annual Meeting; 1999, Abstract P1–593. 87. Rivera RT, Pasion SG, Wong DT, et al. Loss of tumorigenic potential by human lung tumor cells in the presence of antisense RNA specific to the ectopically synthesized alpha subunit of human chorionic gonadotropin. J Cell Biol 1989; 108(6):2423. 88. Kumar S, Talwar GP, Biswas DK. Necrosis and inhibition of growth of human lung tumor by anti-alpha human chorionic gonadotropin antibody. J Natl Cancer Inst 1992; 84(1):42. 89. Tao YX, Bao S, Ackermann DM, et al. Expression of luteinizing hormone/human chorionic gonadotropin receptor gene in benign prostatic hyperplasia and in prostate carcinoma in humans. Biol Reprod 1997; 56(1):67. 90. Bao S, Lei ZM, Rao ChV. The presence of functional luteinizing hormone/chorionic gonadotropin receptors in human prostate cell lines. In: The Endocrine Society Annual Meeting; 1997, Abstract P3–403. 91. Dirnhofer S, Berger C, Hermann M, et al. Coexpression of gonadotropic hormones and their corresponding FSH- and LH/CG-receptors in the human prostate. Prostate 1998; 35(3):212. 92. Lei ZM, Rao ChV. Direct luteinizing hormone regulation of male reproductive tract In: Coutinho EMSP, ed. Reproductive medicine: a millennium review. The proceedings of the 10th World Congress on Human Reproduction; 2000. London: Parthenon Publishing. 2000, In press. 93. Russo IH, Russo J. Chorionic gonadotropin: a tumoristatic and preventive agent in breast cancer. In: Teicher BA, ed. Drug resistance in oncology. New York: Marcel Dekker Inc, 1993:537. 94. Lojun S, Bao S, Lei ZM, Rao ChV. Presence of functional luteinizing hormone/chorionic gonadotropin (hCG) receptors in human breast cell lines: implications supporting the premise that hCG protects women against breast cancer. Biol Reprod 1997; 57(5):1202. 95. Li X, Lei ZM, Rao ChV. The actions of human chorionic gonadotropin in MCF-7 cells support the premise that it may protect woman against breast cancer. In: The Endocrine Society Annual Meeting, 1999. Abstract P3. 96. Lei ZM, Rao ChV. Protective role of human chorionic gonadotropin and luteinizing hormone against breast cancer. In: Barnea ER, Jaunaiux JE, Schwartz PE, Schofield PN, eds. Cancer and pregnancy. London: Springer-Verlag, 2000. 97. Gillott DJ, Iles RK, Chard T. The effects of beta-human chorionic gonadotrophin on the in vitro growth of bladder cancer cell lines. Br J Cancer 1996; 73(3):323. 98. Bieche I, Lazar V, Nogues C, et al. Prognostic value of chorionic gonadotropin beta gene transcripts in human breast carcinoma. Clin Cancer Res 1998; 4(3):671. 99. Butler SA, Laidler P, Porter JR, et al. The beta-subunit of human chorionic gonadotrophin exists as a homodimer [In Process Citation]. J Mol Endocrinol 1999; 22(2):185. 100. Lei ZM, Rao ChV, Ackerman DM, Day TG. The expression of human chorionic gonadotropin/human luteinizing hormone receptors in human gestational trophoblastic neoplasms. J Clin Endocrinol Metab 1992; 74(6):1236. 101. Licht P, Cao H, Zuo J, et al. Lack of self-regulation of human chorionic gonadotropin biosynthesis in human choriocarcinoma cells. J Clin Endocrinol Metab 1994; 78(5):1188. 102. Perilongo G, Rigon F, Murgia A. Oncologic causes of precocious puberty. Pediatr Hematol Oncol 1989; 6(4):331. 103. Themmen AP, Martens JW, Brunner HG. Activating and inactivating mutations in LH receptors. Mol Cell Endocrinol 1998; 145(1–2):137. 104. Kurtzman JT, Spinnato JA, Zimmerman MJ, et al. Human chorionic gonadotropin exhibits potent inhibition of preterm delivery in a small animal model. Am J Obstet Gynecol 1999; 181:853. 105. Hildreth NG, Shore RE, Dvoretsky PM. The risk of breast cancer after irradiation of the thymus in infancy. N Engl J Med 1989; 321(19):1281. 106. Tokunaga M, Land CE, Tokuoka S. Followup studies of breast cancer incidence among atomic bomb survivors. J Radiat Res (Tokyo) 1991; 32 (Suppl):201. 107. MacMahon B, Cole P, Lin TM, et al. Age at first birth and breast cancer risk. Bull WHO 1970; 43(2):209. 108. Trapido EJ. Age at first birth, parity, and breast cancer risk. Cancer 1983; 51(5):946. 109. Russo J, Russo IH. Hormonally induced differentiation: a novel approach to breast cancer prevention. J Cell Biochem Suppl 1995;22:58. 110. Alvarado MV, Alvarado NE, Russo J, Russo IH. Human chorionic gonadotropin inhibits proliferation and induces expression of inhibin in human breast epithelial cells in vitro. In Vitro Cell Dev Biol Anim 1994; 30A(1):4. 111. Srivastava P, Russo J, Russo IH. Chorionic gonadotropin inhibits rat mammary carcinogenesis through activation of programmed cell death. Carcinogenesis 1997; 18(9):1799. 112. Srivastava P, Russo J, Mgbonyebi OP, Russo IH. Growth inhibition and activation of apoptotic gene expression by human chorionic gonadotropin in human breast epithelial cells. Anticancer Res 1998; 18(6A):4003. 113. Huang Y, Bove B, Wu Y, et al. Microsatellite instability during the immortalization and transformation of human breast epithelial cells in vitro. Mol Carcinog 1999; 24(2):118. 114. Russo J, Yang X, Hu YF, et al. Biological and molecular basis of human breast cancer. Front Biosci 1998; 3:D944. 115. Alvarado MV, Russo J, Russo IH. Immunolocalization of inhibin in the mammary gland of rats treated with hCG. J Histochem Cytochem 1993; 41(1):29. 116. Huynh H. In vivo regulation of the insulin-like growth factor system of mitogens by human chorionic gonadotropin. Int J Oncol 1998; 13(3):571. 117. Bourinbaiar AS, Nagorny R. Inhibitory effect of human chorionic gonadotropin (hCG) on HIV-1 transmission from lymphocytes to trophoblasts. FEBS Lett 1992; 309(1):82. 118. Darzynkiewicz Z. The butler did it: search for killer(s) of Kaposi's sarcoma cells in preparations of human chorionic gonadotropin. J Natl Cancer Inst 1999; 91(2):104. 119. Lin J, Lojun S, Lei ZM, et al. Lymphocytes from pregnant women express human chorionic gonadotropin/luteinizing hormone receptor gene. Mol Cell Endocrinol 1995; 111:R13. 120. Zhang YM, Lei ZM, Rao ChV. Human macrophages contain luteinizing hormone and chorionic gonadotropin receptors. J Soc Gynecol Invest 1998; 5(Suppl 1):Abstract 212. 121. Zhang YM, Lei ZM, Rao ChV. Functional importance of human monocyte luteinizing hormone and chorionic gonadotropin receptors. J Soc Gynecol Invest 1999;6(Suppl 1):Abstract 46.

CHAPTER 113 MORPHOLOGY AND PHYSIOLOGY OF THE TESTIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 113 MORPHOLOGY AND PHYSIOLOGY OF THE TESTIS DAVID M. DE KRETSER Structural Organization of the Testis Spermatogenesis Replication of Stem Cells Meiosis Spermiogenesis Nuclear Changes Acrosome Formation Flagellar Development Redistribution of Cytoplasm Spermiation Spermatogenic Cycle Sertoli Cells Control of the Seminiferous Tubule Intertubular Tissue Leydig Cells Intercompartmental Modulation in the Testis Immunologic Control in the Testis Chapter References

STRUCTURAL ORGANIZATION OF THE TESTIS In most mammalian species, the testis is located within the scrotum, having descended from an intraabdominal position during fetal development1 (see Chap. 93). The intrascrotal posi-tion allows the testis to function at a lower temperature than is found within the abdomen. This is a requirement for normal spermatogenesis in many mammals including humans, although in some, such as the elephant, the testes do not descend and spermatogenesis is unaffected by the higher temperature within the abdomen. Because of testicular descent, the vascular supply originates relatively proximally, from the aorta near the origin of the renal arteries. The venous drainage, commencing as an anastomotic plexus of veins (the pampiniform plexus that surrounds the testicular artery), terminates in the renal vein on the left and in the inferior vena cava on the right. This arrangement of vessels acts as a countercurrent mechanism to maintain lower testicular temperature; the cooler venous blood surrounds the testicular artery, decreasing its temperature as it approaches the testis. The remnant of the peritoneal sac, the processus vaginalis, surrounds the testis on its anterior and lateral sides as the tunica vaginalis. The outer dense connective tissue covering, the tunica albuginea, sends septa, which run posteriorly toward the mediastinum of the testis and divide the testis into a series of lobules. Within these lobules lie the seminiferous tubules, which are coiled, extending as loops from the region of the mediastinum of the testis, from which they drain by straight tubules into an anastomotic network of ducts, the rete testis2 (Fig. 113-1). The products of the tubules drain from the rete testis by a series of 15 to 20 ducts, the ductuli efferentes, which in humans constitute part of the head of the epididymis. In turn, the ductuli drain into the duct of the epididymis, whose coils form the remainder of the head, body, and tail of the epididymis.

FIGURE 113-1. A, The arrangement of the seminiferous tubules, rete testis, efferent ducts, and epididymis is illustrated. B, The structure of the rete testis is detailed. (From de Kretser DM, Temple-Smith PD, Kerr JB, et al. Anatomical and functional aspects of the male reproductive organs. In: Bandhauer K, Frick J, eds. Handbuch der Urologie, vol XVI. Berlin: Springer-Verlag, 1982:1.)

The connective tissue surrounding the seminiferous tubules contains vascular and lymphatic vessels and the Leydig cells, which are responsible for the androgenic output of the testis.

SPERMATOGENESIS The sequence of cytologic changes that produce spermatozoa from spermatogonia is called spermatogenesis. It can be subdivided into three major phases: the replication of stem cells, meiosis, and spermiogenesis. REPLICATION OF STEM CELLS The replication of stem cells commences in fetal life with the migration of the primordial germ cells into the mesenchyme of the gonadal ridge. There is evidence that this migration is controlled by stem cell factor and its receptor c-kit.3 After a period of prenatal mitotic division, the gonocytes, which populate the seminiferous cords at birth, remain quiescent until immediately before puberty, when they divide by mitosis to form the spermatogonial population.4 The types of spermatogonia (Fig. 113-2) that can be characterized cytologically vary with each species. However, a feature common to all is the division of the stem cell population to provide two pools of cells, one that moves through the subsequent steps of spermatogenesis and the other that retains its stem cell function.5,6 In the human male, three types of spermatogonia are recognized classically: the A dark, thought to represent the most primitive; the A pale; and the B spermatogonia. They lie adjacent to the basement membrane of the tubule, interspersed with the basal aspects of the Sertoli cells. During cell division, the spermatogonia do not complete cytokinesis and remain linked by intercellular bridges.7Populations of spermatogonia that are injected into tubules devoid of germ cells will partially restore spermatogenesis.8 These studies will provide the experimental basis for further evaluation of the nature and control of the stem cells in the testis.

FIGURE 113-2. Light micrograph of a section from a normal human testis illustrates spermatogonia (SG), primary spermatocytes (SC), spermatids (SD), Sertoli cell nuclei (arrows), and Leydig cells (L). Stages of the human seminiferous cycle are denoted where identifiable (STG). ×950

MEIOSIS Responding to unknown signals, groups of type B spermatogonia begin meiosis, which involves two cell divisions. The spermatogonia lose their contact with the basement membrane of the tubule and are then called primary spermatocytes.9 During the prophase of the first division (see Chap. 90), the primary spermatocytes, which have already replicated their DNA and contain twice the diploid content, undergo a series of characteristic nuclear changes consistent with the appearance and pairing of homologous chromosomes (Fig. 113-3). During the leptotene stage, the chromosomes appear as single, randomly coiled threads, which thicken and commence pairing during zygotene. During pachytene, the chromosomes appear as condensed, closely paired structures, which begin to repel each other during diplotene. Diakinesis is associated with further repulsion of the pairs of chromosomes, which each consist of a pair of daughter chromatids. During the metaphase of this first division, each member of the pair of homologous chromosomes moves to each daughter cell, reducing the number of chromosomes to the diploid number; however, because each chromosome is composed of two chromatids, the DNA content of a diploid cell is retained. Incomplete cytokinesis causes the formation of a pair of joined secondary spermatocytes from each primary spermatocyte.

FIGURE 113-3. This diagram illustrates the process by which homologous chromosomes pair in the first meiotic division involving the primary spermatocytes. In the leptotene phase, the chromosomes are represented by unpaired fine “threads.” These pair during zygotene and thicken during pachytene, eventually repelling each other during diplotene. The pairing process involves the formation of a tripartite structure, synaptinemal complex, which can be identified under the electron microscope. During the pairing, exchange of genetic material occurs between the maternal and paternal chromosomes in a process called crossing over. (From de Kretser DM, Kerr JB. The cytology of the testes. In: Knobil E, Neill JD, eds. The physiology of reproduction, 2nd ed. New York: Raven Press, 1994.)

During the second division, the 23 chromosomes, each comprising a pair of chromatids, attach to the spindle, and the chromatids separate. This yields a cell, a spermatid, containing the haploid DNA and chromosomal complement. SPERMIOGENESIS Cell division stops after the formation of the spermatids. However, a dramatic metamorphosis, spermiogenesis, transforms a conventional cell into a highly specialized cell with the capability of flagellar-derived motility (Fig. 113-4 and Fig. 113-5). Little is known of the mechanisms by which these dramatic cytologic changes are controlled. The developmental phases are termed the Sa1, Sb1, Sb2, Sc2, Sd1, and Sd2 stages according to Clermont, but many of the details can be determined only by electron microscopy.9,10 and 11 With the increasing use of spermatids extracted directly from the testis by biopsy, the specific features that characterize each of the above stages become crucial in identifying the nature of the cell type injected into the cytoplasm of the oocyte to achieve fertilization. The changes can be subdivided into nuclear, acrosome formation, flagellar development, redistribution of cytoplasm, and spermiation.

FIGURE 113-4. Initial cytologic changes during spermiogenesis are shown. See Figure 113-7 for explanation of Sa, Sb1, and Sb2. (From de Kretser DM. The light and electron microscope anatomy of the normal human testis. In: Santen RJ, Swerdloff RS, eds. Male sexual dysfunction: diagnosis and management of hypogonadism, infertility and impotence. New York: Marcel Dekker Inc, 1986:3.)

FIGURE 113-5. Subsequent cytologic changes during spermiogenesis. (See also Fig. 113-7.) (From de Kretser DM. The light and electron microscope anatomy of the normal human testis. In: Santen RJ, Swerdloff RS, eds. Male sexual dysfunction: diagnosis and management of hypogonadism, infertility and impotence. New York: Marcel Dekker Inc, 1986:3.)

FIGURE 113-7. The light microscopic features of the stages of the human seminiferous cycle are illustrated. (L, leptotene primary spermatocyte; M, spermatocytes in meiosis; P, pachytene primary spermatocyte; Z, zygotene primary spermatocyte; Sa1, Sb1, Sb2, Sc, Sd1, spermatids at different steps of spermiogenesis; Ser, Sertoli cell nucleus.) (From de Kretser DM, Temple-Smith PD, Kerr JB, et al. Anatomical and functional aspects of the male reproductive organs. In: Bandhauer K, Frick J, eds. Handbuch der Urologie, vol XVI. Berlin: Springer-Verlag, 1982:1.)

NUCLEAR CHANGES During the Sa1 and Sb1 stages, the nucleus, which ultimately forms the head of the sperm, remains centrally placed, but it is subsequently displaced peripherally, coming into apposition with the cell membrane, separated only by the acrosomal cap. There is also a progressive decrease in nuclear volume associated with chromatin condensation, causing the development of resistance by the DNA to degradation by the enzyme DNAase. ACROSOME FORMATION During the Sa1 and Sb1 stages, the Golgi complex of the spermatid produces several large vacuoles, which are applied to one pole of the nucleus. These vacuoles form a cap-like structure, the acrosome, which contains substances for penetration of the ovum during fertilization.11a Sperm that do not contain an acrosome cannot fertilize the ovum; because of the globular shape of the sperm head, the condition is known as globozoospermia. The Golgi complex subsequently migrates to the abacrosomal pole of the spermatid and is lost in the residual cytoplasm. FLAGELLAR DEVELOPMENT The initial development of the tail occurs from the pair of centrioles located near the Golgi complex of the Sa1 spermatid. The microtubular structure, comprising nine peripheral doublets surrounding a pair of single microtubules, grows out from the distal centriole and forms the axoneme or core of the tail. The developing axoneme lodges in a facet at the abacrosomal pole of the nucleus by a complex articulation called the neck of the spermatid. Initially, the axoneme distal to the neck is surrounded by the cell membrane, but several specializations develop. Immediately adjacent to the outer nine doublets, a set of nine electron-dense fibers develop, which are connected cranially to the neck but distally taper to disappear eventually. A pair of these dense fibers, which lie diametrically opposite to each other, persist distally and become surrounded by a collection of microtubules. These microtubules form the precursors for solid electron-dense fibers, or ribs, which run transversely around the axoneme, joining the pair of longitudinal, dense fibers. The region of the tail surrounded by these ribs is the principal piece. The region between the principal piece and the neck is the last segment to become organized. Some of the genes, which encode the proteins comprising the outer dense fibers and the fibrous sheath, have been identified; these will provide the basis for future studies on the formation and function of the components.12,13,14 and 15 Relatively late in spermiogenesis, between the Sd1 and Sd2 stages, the mitochondria that lie in the peripheral regions of the cell coalesce to form several helical arrays around the axoneme, forming the midpiece of the sperm. The axoneme is identical to the core structure of cilia throughout the plant and animal kingdoms. The two centrally placed, single microtubules appear to be connected to the nine peripheral doublets by a series of radial spokes (Fig. 113-6). One of each pair of doublets, termed subfiber A, is smaller and more electron dense; it sends a pair of hook-like extensions, or dynein arms, toward the adjacent doublet. They represent condensations of the protein, dynein, which has adenosine triphosphatase activity and is vital to the generation of flagellar motility; the absence of dynein arms is associated with immobility of sperm and cilia.16,17

FIGURE 113-6. The structure of a typical axoneme of a sperm tail.

REDISTRIBUTION OF CYTOPLASM Associated with the eccentric placement of the nucleus is a significant repositioning of the cytoplasm and organelles. This probably is caused by a palisade of microtubules, the manchette, which extends as a cylindrical collection from the head, in the region where the acrosome ends, to the caudal pole of the spermatid. Ultimately, most of this cytoplasm is shed by the spermatid and effectively appears to be pulled off by processes of Sertoli cell cytoplasm, which invaginate into pouches developing in the spermatid cytoplasm.10 This cytoplasmic remnant, the residual body, is phagocytosed and degraded by the adjacent Sertoli cells. SPERMIATION The release of spermatids at the Sd2 stage occurs in association with the loss of the residual cytoplasm. SPERMATOGENIC CYCLE Careful and extensive cytologic studies have demonstrated a characteristic sequence of cell associations through which the seminiferous epithelium passes and that constitute the cycle of the seminiferous epithelium.6 In the rat, this consists of 14 cell associations, each of which may extend over several millimeters of tubule. However, in the human, six cell associations or stages have been identified, and each may occupy only 25% to 33% of a cross section of a seminiferous tubule9 (Fig. 113-7). Studies in which dividing cells are labeled by tritiated thymidine have shown that the time taken for the daughter cells of spermatogonial divisions to mature to Sd2 spermatids released from the seminiferous tubule is 64 ± 6 days or 4.5 cycles of the human seminiferous cycle.18 The time for spermatogenesis is unique for each species and represents a biologic constant. The germ cells either pass through these stages at a specified speed or degenerate. It is uncertain whether these time constraints are an innate function of each cell or whether they are imposed by other controlling factors. Coordination of the spermatogenic process may be achieved by the unusual organization of the seminiferous epithelium, wherein the germ cells remain joined by intercellular bridges.7 Additionally, coordination may be achieved by the Sertoli cells, whose arborizing branches maintain contact with many germ cell stages around the radial axis of the

seminiferous tubules. SERTOLI CELLS Sertoli Cell Cytology. The general features of the Sertoli cell are similar in most mammalian species.19,20 The cells extend from the basement membrane of the tubule to the luminal surface of the epithelium, and, although the base of the cell is clearly identifiable, the central portions cannot be distinguished by light microscopy. This results from the formation of many thin cytoplasmic prolongations that extend in an arborizing network around the germ cells undergoing spermatogenesis. Electron microscopy has demonstrated that these processes are always surrounded by the cell membrane of the Sertoli cell. In the Sertoli cell, the cytoplasmic organelles exhibit some degree of polarity. The basal aspects, which abut the basement membrane of the tubule, interspersed between spermatogonia, often show collections of mitochondria. The nucleus is pleomorphic, is aligned perpendicular to the basal lamina, and contains a prominent nucleolus. The nucleolus is a feature of the postpubertal Sertoli cell, probably related to the follicle-stimulating hormone (FSH)–induced maturation of the protein synthetic capacity of the mature Sertoli cell.21,22 In the perinuclear region, small collections of rough endoplasmic cisternae and Golgi membranes can be seen. Significant numbers of lipid inclusions, often surrounded by smooth endoplasmic reticulum, are found adjacent to the nucleus; in some species, these inclusions exhibit significant changes in number with the stages of the seminiferous cycle. Lysosomes, lipofuscin pigment, and residual bodies, which are the phagocytosed cytoplasmic remnants of the spermatids, may be seen deep within the Sertoli cell cytoplasm. The shape of the Sertoli cell is probably maintained by the cytoskeleton, consisting of a perinuclear array of fine filaments and microtubules that often extend into the smaller processes of cytoplasm surrounding the germ cells. Inter–Sertoli Cell Junctions. One of the characteristic features of the Sertoli cell is the specialized inter–Sertoli cell junction, which occurs where adjacent Sertoli cells abut. These usually commence at a level in the epithelium just luminal to the basal row of the spermatogonia and extend centrally. This unique cell junction consists of small occluding junctions, representing points of fusion of the cell membranes that obliterate the intercellular space. Adjacent to these points of occlusion, smooth-membraned cisternae run parallel to the cell membrane and demarcate a narrow band of cytoplasm containing bundles of fine filaments.23 The effect of these complexes is to prevent transport by way of the intercellular space to the centrally placed germ cells.23 The inter–Sertoli cell junctions divide the seminiferous epithelium into two compartments: a basal one containing the spermatogonia and preleptotene primary spermatocytes and an adluminal one containing the subsequent stages of germ cell maturation. The cell junctions represent the site of the blood–testis barrier; by preventing intercellular transport, they create a highly selective permeability barrier based on transport systems within the Sertoli cell.24 These cell junctions and the blood–testis barrier are absent in the immature testis but develop during pubertal maturation.21,25 The resultant development of the cell junctions and barrier coincides with the onset of meiosis and seminiferous fluid secretion.22,26 These cell junctions are not permanent structures since they must disassemble and reform basally as preleptotene spermatocytes lose their attachment to the basement membrane and move centrally. Sertoli Cell Function. There has been a great expansion in the knowledge of Sertoli cell function.27 Because of the intimate relationship of the Sertoli cells to germ cells, the function of these cells is probably crucial for the successful completion of spermatogenesis. The concept that the number of Sertoli cells determines the spermatogenic output of the testis has gained considerable support, and clearly, the Sertoli cell is critically important for the transport of substances into the seminiferous tubule. Numerous examples indicate that the Sertoli cells are essential for the metabolic activities of the germ cells: For example, the germ cells, which are unable to metabolize glucose, are provided with lactate by the Sertoli cells.28 Key roles of the Sertoli cells are discussed later. Sertoli Cell Replication. Studies in the rat demonstrate that the Sertoli cell population divides by mitosis in fetal and post-natal life until day 15 but remains stable thereafter.29 Neonatal hypothyroidism in the rat can extend the period of Sertoli cell replication, and in adult life the increased numbers of Sertoli cells lead to a marked increase in sperm output.30 These data strongly suggest that the number of Sertoli cells in the testis is a major factor in controlling the spermatogenic potential of the testis. Use of this model has also shown that the functional maturation of the Sertoli cells is delayed during hypothyroidism31and that there is a marked delay in the process of spermatogenesis.32 The control of Sertoli cell proliferation involves FSH, thyroid hormones, and growth factors such as activin A.33 Seminiferous Tubule Fluid Production. The Sertoli cells secrete a fluid, with characteristics distinct from plasma, into the lumen of the seminiferous tubule.34 This secretion commences during sexual maturation, after the formation of the blood–testis barrier; it is dependent on FSH for its production.35,36 In animals, the production of seminiferous tubule fluid can be measured by unilateral efferent duct ligation; the increasing difference in weight, with time, between the ligated and unligated sides provides this index.35 Under such conditions, fluid production is maintained for 24 hours, after which a pressure atrophy of the seminiferous epithelium occurs and fluid production decreases, eventually ceasing altogether.35,36 Androgen-Binding Protein. The discovery that the Sertoli cell produces an androgen-binding protein (ABP) that is capable of binding dihydrotestosterone and testosterone with high affinity has provided a biochemical marker of Sertoli cell function.27,37 Purification and characterization have revealed that, in the rat, ABP is a protein with a molecular mass of 85,000 daltons, for which a radioimmunoassay has been developed.38 The amount of ABP secreted by the Sertoli cells varies among species; it is absent in some, and in others, such as humans, it is uncertain whether it is produced in the testis, because there is contamination of testis tissue with blood, which contains an ABP formed in the liver. ABP and the sex steroid–binding globulin (SSBG) found in some species are very similar, and represent the same protein produced in the liver and testis under different regulatory mechanisms.39,40 ABP/SSBG acts as a hormone or growth factor, as evidenced by several studies that have identified high-affinity binding sites.41,42 Furthermore, there are multiple alternate RNA transcripts of ABP in different tissues, and one of these encodes a nuclear targeting signal.43,44 The secretion of ABP occurs principally into the lumen of the seminiferous tubules (Fig. 113-8), but it also passes across the basal aspect of the Sertoli cells into interstitial fluid and blood.38 ABP may provide a mechanism for ensuring a large store of testosterone within the tubule, stabilizing fluctuations in testosterone secretion by the Leydig cells. It may also provide a mechanism for ensuring a high concentration of testosterone for the caput epididymis, because ABP and its bound testosterone are transported into the epididymis, where they are absorbed by the principal cells.

FIGURE 113-8. A diagram of the testis shows the relationship of Sertoli cells to germ cells. Factors controlling the testicular compartments are outlined. Evidence for the secretion of estradiol (E2) and gonadotropin-releasing hormone (GnRH) by the adult Sertoli cell is tentative, but the hormones have been suggested as local modulators of Leydig cell function. (T, testosterone; ABP, androgen binding protein; FSH, follicle-stimulating hormone; LH, luteinizing hormone.)

Secretion of Other Plasma Proteins. The Sertoli cell may produce a number of other proteins found in plasma, perhaps circumventing the presence of the blood–testis barrier and ensuring adequate local concentrations. Albumin, transferrin, and plasminogen activator are such proteins secreted by the Sertoli cell.45,46 and 47 Their specific functions are unclear. Plasminogen activator has been proposed as a mechanism for enabling the discrete disruption of the inter–Sertoli cell junctions to allow germ cells to pass through from the basal to the adluminal compartments. Transferrin production by the Sertoli cells facilitates the transport of iron across the blood–testis barrier and enables the provision of adequate amounts of iron for the metabolism of germ cells.47 Inhibins. The Sertoli cell secretes inhibin (see Fig. 113-8), which increases in the testis after efferent duct ligation.48,49 and 50 Inhibin, originally isolated from follicular fluid, is a disulfide-linked dimer of two dissimilar subunits termed a and b.51 Two forms exist that share a common a subunit but distinct b subunits: These forms are inhibin A (abA) and inhibin B (abB).51 Both forms suppress FSH, whereas dimers of the b subunits—termed activins (activin A [bA bA], activin AB [b A b B], activin B [bB bB])—all stimulate FSH. The genes coding the subunits produce larger precursor molecules that dimerize and are proteolytically cleaved to form the mature substances.52,53 The subunits of inhibin show significant homology to transforming growth factor-b, antimüllerian hormone (AMH), and numerousother proteins, such as

the protein coded for by the decapentaplegic gene complex of Drosophila.52,53 and 54 The Sertoli cells secrete inhibin B as the principal protein feedback on FSH secretion, and there is a good correlation between sperm output and inhibin B levels in the circulation.55,56 Inhibin B levels are inversely correlated to FSH concentrations; therefore, there have been suggestions that inhibin B is a useful marker for Sertoli cell function.57 Experiments using recombinant inhibin A indicate that, although both testosterone and inhibin modulate FSH, inhibin can normalize FSH levels in castrated rams in the absence of testosterone.58,59 Antimüllerian Hormone. Antimüllerian hormone is produced by the Sertoli cell during fetal and early postnatal life.60,60a It has been purified and shown to be a glycoprotein. In fetal life it directs the regression of the müllerian ducts during male sexual differentiation (see Chap. 90). Steroidogenic Function. The presence of smooth endoplasmic reticulum, lipid droplets, and mitochondria with tubular cristae in the cytoplasm of the Sertoli cells led to the proposal that the Sertoli cells could be the site of steroid biosynthesis.61 By separating seminiferous tubules from interstitial tissue, it was demonstrated that the tubules do not have the capacity to metabolize cholesterol to products more distant in the steroidogenic pathway. However, if provided with substrates such as progesterone, the seminiferous tubules have the capacity to metabolize these materials to androgens.62 Unique metabolites of C-19 steroids have been identified in Sertoli cells, and the pattern of metabolism can be altered by FSH treatment. Despite the uncertainty about the steroidogenic capacity of the seminiferous tubules and Sertoli cells, cultures of immature rat Sertoli cells have the ability to metabolize androgens, with the enzyme aromatase, to estradiol.63 However, in the rat, this activity decreases rapidly with age and is not detectable 20 days after birth. Evidence suggests that estradiol production by the testis after this time occurs within the Leydig cells.64 In the immature rat, the activity of aromatase is FSH inducible, and this reaction has been used as a basis for an in vitro bioassay for FSH.65 Age-Dependent Changes in Sertoli Cell Function. Before 20 days, the Sertoli cells can respond dramatically to FSH by an increase in cyclic adenosine monophosphate (AMP) and in protein synthesis, together with the stimulation of the enzyme aromatase. Additionally, the immature Sertoli cells produce AMH in fetal life and for a short period after birth. These functions disappear ~20 days after birth in the rat and are replaced by increases in ABP secretion, the onset of fluid production, and an increase in inhibin secretion.66 The mechanisms causing these changes in Sertoli cell function are unknown, but some may be related to the onset of pubertal secretion of FSH by the pituitary. Stage-Dependent Changes in Sertoli Cell Function. The seminiferous tubule undergoes a sequence of cytologic changes, causing the formation of specific cell associations, which have been identified as the seminiferous cycle. In association with these changes in the germ cell complement of the seminiferous tubule, there are distinct cytologic changes that have been identified within the Sertoli cell, particularly in the rat, in which there is a very well-defined seminiferous cycle. In this species, a number of biochemical parameters vary according to the stage of the seminiferous cycle.67 Thus, FSH-receptor levels, ABP production, and cyclic AMP production change according to the stage of the seminiferous cycle in response to FSH stimulation.9 The nature of the products from germ cells that modulate Sertoli cell function is still unknown. The concept that these may arise from late spermatids, possibly through the phagocytosis of residual bodies, has been reviewed.68 CONTROL OF THE SEMINIFEROUS TUBULE Hormonal Control. The function of the testis depends on the secretion of the gonadotropic hormones, FSH and luteinizing hormone (LH), by a functional hypothalamic-pituitary unit (see Chap. 16). Moreover, the action of LH on spermatogenesis is mediated through the secretion of testosterone by the Leydig cells. However, there is considerable controversy concerning the relative roles of FSH and testosterone in the control of seminiferous tubule function. Initiation of Spermatogenesis. The role of testosterone in this process is not questioned, and current data indicate that constitutive-activating mutations in the LH receptor cause precocious puberty, while inactivating mutations result in familial testosterone resistance and male pseudohermaphroditism.69,70 The earlier view that, in humans and other mammalian species, both FSH and LH are required for the initiation of spermatogenesis during pubertal maturation71,72 and 73 has been challenged by three studies in mice. First, testosterone alone has been shown to induce spermatogenesis in the hpg mouse, which lacks the capacity to secrete gonadotropin-releasing hormone and, thus, cannot produce both FSH and LH.74 Secondly, inactivation of the gene encoding the b subunit of FSH did not prevent the onset of full spermatogenesis during pubertal maturation in these mice.75 The third study of several males with inactivating mutations of the FSH receptor found that the males were able to complete spermatogenesis, but in most, the testicular volumes and sperm counts were very significantly impaired.76 In all these studies it was noted that the testes were smaller and that the sperm output was lower than normal. The possibility that this results from an impairment of Sertoli cell multiplication (normally stimulated by FSH) is being explored. The view that both FSH and LH were required originated from studies of patients with hypogonadotropic hypogonadism during the induction of spermatogenic activity using FSH and LH. Some patients respond to LH or human chorionic gonadotropin (hCG) alone, but most require the action of both gonadotropic hormones.77 There are conflicting data for the rat, suggesting that LH, through the action of testosterone, may play a more dominant role during pubertal maturation; however, the rat is a very poor model for studies of sexual maturation because it does not have a prepubertal period. The spermatogenic process in the rat occupies 48 to 50 days, and mature spermatozoa can be seen in the rat 45 to 50 days after birth. Consequently, changes closely related to the time of birth in the rat may be involved in the initiation of spermatogenesis. Maintenance of Spermatogenesis. Given the results of the studies in the hpg mouse and those in genetically modified mice in which the b subunit of FSH is not produced, the controversy regarding the role of FSH in the maintenance of spermatogenesis after its establishment at puberty is still unresolved. In the rat, the observations that testosterone alone, given immediately after hypophysectomy, could maintain spermatogenesis without FSH have received some support from experiments in humans using an alternative design. In these studies, the suppression of FSH and LH secretion by the administration of contraceptive doses of testosterone caused azoospermia or severe oligospermia. These changes could be reversed by the administration of hCG or LH, which presumably act by stimulating Leydig cells to increase the intratesticular levels of testosterone.77,78 However, when researchers used the same design to suppress spermatogenesis, highly purified FSH was also able to restore the sperm count, presumably without altering intratesticular levels of testosterone.79 Several studies have shown that the levels of testosterone normally found within the testis are not required to maintain spermatogenesis, which can proceed successfully at concentrations ~10% of normal.80,81 Nonetheless, these levels still represent twice the normal circulating concentrations, in turn raising questions as to why the testicular androgen receptor requires significantly greater stimulation than other androgen-dependent tissues, for example, prostate. Both in rats and in humans, if regression of the spermatogenic epithelium has occurred after hypophysectomy, testosterone alone is insufficient to restore spermatogenesis.80,81 and 82 However, this view has been challenged by studies in stalk-sectioned monkeys in whom testosterone treatment alone reversed the testicular regression that had occurred, although testicular volumes returned to only 60% of presurgical levels.83 The claim that no biochemical action of FSH was found in the adult rat testis84 has now been shown to be erroneous since the sensitivity to FSH stimulation varies with the stage of the seminiferous cycle within the seminiferous tubule.67 These studies demonstrate that an FSH-induced effect can be obtained in the adult testis provided an FSH-sensitive phase of the seminiferous cycle is selected. Positive evidence for the role of FSH in the spermatogenic process includes the fact that receptors for FSH have been identified on the Sertoli cell and on spermatogonia.85 The action of FSH is mediated by the cyclic AMP–protein kinase system and stimulates protein synthesis by the Sertoli cell. A number of proteins, such as ABP, aromatase, plasminogen activator, RNA polymerase, inhibin, and proteoglycan, are responsive to FSH or cyclic AMP.49,66,84 Studies in primates and in normal men have shown that FSH is important in maintaining the transition of type A to type B spermatogonia.86,87 Testosterone is clearly important in maintaining the seminiferous epithelium. This action of testosterone is mediated through androgen receptors found within the Sertoli cell and on peritubular and Leydig cells.88,89 and 90 Further evidence for the role of androgens in the stimulation of Sertoli cell function was obtained from Sertoli cells in culture, in which RNA polymerase and ABP production could be stimulated independently of any action of FSH.91,92 After hypophysectomy, testosterone alone can maintain fluid production and the secretion of inhibin by the rat testis.36,93 Unfortunately, the rat proves to be a difficult model in which to explore the effect of high doses of testosterone, since these stimulate FSH secretion.94 Morphometric studies have shown that spermiogenesis is exquisitely sensitive to testosterone,81,94 and further data suggest that withdrawal of testosterone disrupts the conversion of step 7 to 8 because of premature sloughing of round spermatids into the epididymis.95,96 It is likely that this cell loss is due to disruption of the ectoplasmic specializations between spermatids and Sertoli cells (Fig. 113-9). Further, the studies in the hpg model indicate that testosterone is important in facilitating the survival of primary spermatocytes.74

FIGURE 113-9. The relationship between germ cells and Sertoli cells in the rat is shown. In the presence of low intratesticular testosterone levels, spermatids at stages VII to VIII of the cycle are shed from the epithelium. (From McLachlan RI, Wreford NG, Robertson DM, et al. Hormonal control of spermatogenesis. Trends Endocrinol Metab 1995; 6:95.)

The relative role of testosterone and dihydrotestosterone in spermatogenesis has been perplexing, given that the concentrations of intratesticular testosterone are significantly greater. However, studies with 5a-reductase inhibitors have shown that dihydrotestosterone is a significant stimulator of spermatogenesis when intratesticular testosterone concentrations decline.96 It is recognized that the testis secretes estradiol, arising from the conversion of androgens by the enzyme aromatase. Evidence for an effect of estradiol on spermatogenesis has emerged from gene-targeted disruption of the P450 aromatase gene, which demonstrated that the initially fertile mice progressively became infertile as a result of decreases in spermatid numbers, increased apoptosis, and abnormal acrosome development.97,98 and 99 More direct evidence of the actions of estradiol was obtained from gene knock-out of the estradiol a receptor: These mice were infertile due to the actions of estradiol on fluid reabsorption in the epithelium of the efferent ductules.100 The resultant back pressure resulted in loss of germ cells from the seminiferous epithelium. However, estradiol can still act on the seminiferous epithelium in these mice, since a functional estradiol b receptor remains in the testis.101 Initial reports of the knock-out of this gene indicate that the mice are fertile at 6 weeks of age.102 In summary, while recent studies shed some doubt as to the critical importance of FSH in enabling spermatogenesis to proceed to completion, others have defined specific points at which FSH appears to be very important in maintaining a normal throughput of germ cells. These points appear to be in its action on Sertoli cell mitosis and in facilitating the formation and survival of type B spermatogonia. Nonhormonal Control. There are numerous steps that must be successfully completed before the testis can successfully produce a normal sperm output. These involve molecular mechanisms that require key regulators that are not hormones. As these molecular controllers are identified, usually through experiments that involve the exploration of the function of a protein through gene-targeted disruption, additional regulators of spermatogenesis emerge. It is not possible to consider these proteins exhaustively in this chapter, but a few examples are given that illustrate these developments. For successful sperm production, the development and differentiation of the testis must proceed normally. Any mutation or rearrangement in genes, which is crucial for normal testis development, will impair sperm output or testicular development. Mutations in key domains of the androgen receptor can disrupt hormone binding and result in testicular feminization, but data indicate that there are mutations that occur in other regions of the gene encoding the receptor that do not interfere with sexual differentiation but can impair spermatogenesis. These include expansions of the CAG repeat sequence in the amino-terminal region of the protein.103 The deletion of genes on the long arm of the Y chromosome has demonstrated the presence of testis-specific genes that are essential to enable normal sperm production. One of these, DAZ (deleted in azoospermia), encodes an apparent RNA-binding protein and exhibits homology to the boule gene in Drosophila, mutations that cause sterility in flies. Mutations or deletions of the DAZ genes, for which there are multiple copies in humans, result in severe disruption of spermatogenesis, without any effect on sexual differentiation.104,105 Numerous studies in mice have shown that mutations in the gene encoding stem cell factor or its receptor c-kit result in testes devoid of germ cells because of disruption of the migration of the primordial germ cells and their transformation into spermatogonia.3,106 Later stages of spermatogenesis can show disruption by interference in molecular mechanisms. For instance, targeted disruption of the gene encoding heat shock protein 70-2, a molecular chaperone, results in the arrest of spermatogenesis at the primary spermatocyte stage since this protein appears to be crucial to enabling the completion of meiosis.107 There are numerous studies to indicate that apoptotic mechanisms are important regulatory pathways in the testis, especially following hormonal modulation.108 Evidence supporting this view has emerged from studies of targeted disruption of the gene encoding bcl-w, a cell survival molecule; the first wave of spermatogenesis in mice almost progressed to completion but collapsed, resulting in infertility and ultimately in a Sertoli cell–only phenotype.109 Whether there are hormonal mechanisms that regulate this and other proteins, which control cell survival, remains to be established. INTERTUBULAR TISSUE The seminiferous tubules are supported by a loose connective tissue, which is supplied by a rich vascular network. It is bounded by the basement membrane of the seminiferous tubule, which is surrounded by a varying number of layers of contractile myoid cells interspersed with collagen fibers and a basement membrane–type material that is applied to some of the layers of the myoid cells. The myoid cells are modified smooth muscle cells that cause the contraction of the seminiferous tubules. The general organization of the intertubular tissue varies among species, based on the number of Leydig cells, which are responsible for androgen secretion, the arrangement of the lymphatics, and the extent of the connective tissue.110 The intertubular tissue contains varying numbers of Leydig cells, fibroblasts, macrophages, mast cells, and small unmyelinated nerve fibers. LEYDIG CELLS The Leydig cells are derived from the mesenchyme of the gonadal ridge; two generations of Leydig cells are developed. In fetal life, the differentiation of mesenchyme into Leydig cells induces the secretion of androgens that generate the sexual differentiation of the external genitalia. These fetal Leydig cells degenerate shortly after birth, and the prepubertal period is characterized by the absence of Leydig cells from the intertubular tissue. Associated with the pubertal secretion of gonadotropins, adult Leydig cells redifferentiate from connective tissue precursors 111 within the intertubular tissue. In other species, such as the rat, in which the interval from birth to sexual maturation is short, some overlap occurs between the fetal and adult generations. The testosterone is secreted into the intertubular tissue, where it is absorbed by blood vessels, lymphatics, and the seminiferous tubules. Leydig Cell Cytology. Leydig cells form small collections around blood vessels and have a variable appearance by light microscopy, usually attributed to the lipid inclusions, which cause a variable vacuolation (see Fig. 113-2). The Leydig cells are characterized by an ovoid nucleus exhibiting a conspicuous nucleolus. The cytoplasm contains a large amount of smooth endoplasmic reticulum in the form of interconnected tubules. Mitochondria contain both lamellar and tubular cristae, the latter being typical of steroid-secreting cells. The content of lipid, lysosomes, and lipofuscin pigment is variable. The human Leydig cells are characterized by the presence of crystalloid inclusions, the crystals of Reinke, the functional significance of which is unknown. The amount of smooth endoplasmic reticulum and mitochondria declines after hypophysectomy and increases after LH or hCG stimulation.112 Leydig Cell Function. The Leydig cells secrete testosterone and are able to synthesize cholesterol from acetate, with the cholesterol acting as a substrate for steroidogenesis. The amount of cholesterol obtained by lipoprotein uptake relative to synthesis from acetate varies among species. The enzymatic steps involved in the synthesis of testosterone from cholesterol are outlined in Figure 113-10. Besides the secretion of testosterone, the Leydig cells secrete estradiol, contributing 20% to 30% of the total circulating estradiol; the remainder is derived from the peripheral aromatization of androgenic substrates. There is a subcellular localization of the enzymes involved in steroido-genesis, with the conversion of cholesterol to pregnenolone localized in mitochondria; the remaining steps of steroid bio-synthesis in the testis depend on enzymes located within the smooth endoplasmic reticulum of the Leydig cell.

FIGURE 113-10. These are the biochemical path-ways in the synthesis of testosterone.

Several studies have expanded the understanding of the mechanisms underlying cholesterol transport into the mito-chondria, a step that is critical in enabling cleavage of the side chain of pregnenolone. The isolation and characterization of steroidogenic acute regulatory protein (StAR)113 and the cloning of the gene encoding this protein demonstrated that it had a crucial role in the transport of cholesterol into mitochondria. Mutations in this gene or its targeted disruption profoundly interfered with steroid hormone biosynthesis in all steroid-secreting endocrine tissues; in humans, mutations were shown to be responsible for the condition of lipoid congenital adrenal hyperplasia.114,115 Influence of Luteinizing Hormone. The Leydig cell contains receptors for LH on its cell membrane; this hormone controls testosterone secretion by means of cyclic AMP.116 The principal enzyme controlled by LH is the side-chain cleavage enzyme involved in the conversion of cholesterol to pregnenolone. Besides the immediate events initiated through the phosphorylation of proteins that induce testosterone secretion, LH is trophic to the Leydig cell and stimulates the incorporation of labeled amino acids into specific proteins. This trophic activity causes a hypertrophy of the Leydig cells and probably increases the number of Leydig cells. The stimulation of Leydig cells with large doses of LH or hCG rapidly reduces the number of their receptors, a phenomenon termed down-regulation.116,117 Although these changes decrease testosterone secretion 24 to 48 hours after an injection, repeated stimulation does not yield the same results. Additionally, a single injection of hCG is followed by a prolonged steroidogenic response characterized by two phases of testosterone secretion, one initially occurring over the first 6 to 18 hours and the second occurring 48 to 72 hours later. 117 The results indicate that hCG can be administered at 6- to 7-day intervals due to the prolonged steroidogenic response. The nadir between the two phases of testosterone secretion is due to the block in testosterone production induced by the large injection of hCG, and recovery from the inhibition allows restimulation of testosterone secretion by the existing plasma hCG, because of the long half-life of hCG. INFLUENCE OF OTHER FACTORS. There is increasing evidence, principally from rat studies, that other factors may be involved in the stimulation of testosterone secretion by Leydig cells. This concept originated from observations that damage to the seminiferous tubules by a number of experimental agents caused changes within the Leydig cells,118 which included Leydig cell hypertrophy, partial loss of LH receptors, and a hyper-responsiveness of the Leydig cells to hCG stimulation in vitro. If unilateral damage to the testis was induced, such as after uni-lateral cryptorchidism, the Leydig cell changes were present exclusively in the damaged testis, thereby ruling out circulating humoral factors, such as LH, participating in these changes. The resultant hypothesis suggested that the seminiferous tubules somehow modulate the Leydig cells (see Fig. 113-8). Reasoning that any factor passing from the seminiferous tubules to the Leydig cells must pass across the lymphatic sinusoidal system in the rat, investigators have demonstrated a proteinaceous factor that was not LH and that stimulated steroidogenesis.119 This substance stimulates testosterone more than the maximum levels generated by LH or hCG. Also, a number of studies have suggested that the macrophages present in both the testis and ovary may secrete substances capable of stimulating steroidogenesis, thereby increasing the potential for local control of Leydig cells.118 This view has been substantiated by the impaired development and function of Leydig cells in mice wherein the gene encoding CSF-1 has been disrupted.120 The relative importance of these local factors and of LH in the physiology of testosterone secretion requires clarification. The possibility that these changes may be explained by modulation of StAR or the peripheral benzodiazepine receptor121 requires further work.

INTERCOMPARTMENTAL MODULATION IN THE TESTIS Throughout this chapter, the seminiferous tubules and intertubular tissue have been considered as independent entities. However, there is increasing evidence that this approach is unwarranted. It is well documented that testosterone is required for the process of spermatogenesis, indicating that the Leydig cells at a local level are able to influence the seminiferous tubules.122 Additionally, the seminiferous tubules are involved in the modulation of Leydig cell function. There is also evidence that the Leydig cells exhibit changes in size according to the stage of the seminiferous cycle in the tubules immediately adjacent to them.123 Furthermore, the Sertoli cells show marked changes in function in association with spermatogenic damage and the stage of the seminiferous cycle.124 All of the agents used to induce experimental damage to spermatogenesis decrease the parameters of Sertoli cell function, such as seminiferous tubule fluid production, ABP production, and inhibin production. These changes occur despite a relatively well-maintained morphology of the Sertoli cell; they indicate the importance of obtaining sensitive biochemical indices of Sertoli cell function. The testis should be considered as a functional unit and not as individual compartments with few functional interrelationships. Consideration of these factors may soon provide explanations about why spermatogenesis in certain infertile men does not proceed to completion despite a well-maintained stem cell population within the seminiferous epithelium.

IMMUNOLOGIC CONTROL IN THE TESTIS There is increasing interest in the concept that the testis is an immunologically privileged site; this is based on observations that grafts in the testis survive for prolonged periods.125,126 The maintenance of this environment may be crucial in preventing the formation of autoantibodies to sperm components. Attention has been drawn to the large population of macrophages in the intertubular tissue of the testis that have an impaired capacity to respond to an inflammatory stimulus by the secretion of proinflammatory cytokines. The reason for this observation is still unclear but may involve the production of unidentified substances by the Leydig and Sertoli cells. These cells have the capacity to produce cytokines such as interleukin (IL)-1 and IL-6 in response to inflammatory stimuli.127 CHAPTER REFERENCES 1. Wartenberg H. Differentiation and development of the testes. In: Burger HG, de Kretser DM, eds. The testis. New York: Raven Press, 1981:39. 2. de Kretser DM, Temple-Smith PD, Kerr JB. Anatomical and functional aspects of the male reproductive organs. In: Bandhauer K, Frick J, eds. Handbuch der Urologie, vol XVI. Berlin: Springer-Verlag, 1982:1. 3. Marziali G, Lazzaro D, Sorrentino V. Binding of germ cells to mutant SI d Sertoli cells is defective and is rescued by expression of the transmembrane form of the c-kit ligand. Dev Biol 1993; 157:182. 4. Muller J, Skakkebaek NE. Quantification of germ cells and seminiferous tubules by stereological examination of testicles from 50 boys who suffered from sudden death. Int J Androl 1983; 6:143. 5. Huckins C. The spermatogonial stem cell population in adult rats: I. Their morphology, proliferation and maturation. Anat Rec 1971; 169:533. 6. Clermont Y. Kinetics of spermatogenesis in mammals: seminiferous epithelium cycle and spermatogonial renewal. Physiol Rev 1972; 52:198. 7. Dym M, Fawcett DW. Further observations on the numbers of spermatogonia, spermatocytes and spermatids connected by bridges in the mammalian testis. Biol Reprod 1971; 4:195. 8. Brinster RL, Zimmerman JW. Spermatogenesis following male germ-cell transplantation. Proc Natl Acad Sci U S A 1994; 11298. 9. Clermont Y. The cycle of the seminiferous epithelium in man. Am J Anat 1963; 112:35. 10. de Kretser DM. Ultrastructural features of human spermiogenesis. Z Zell-forsch 1969; 98:477. 11. Holstein AF, Roosen-Runge EC. Atlas of human spermatogenesis. Berlin: Grosse Verlag, 1981. 11a. Abou-Haila A, Tulsiani DR. Mammalian sperm acrosome: formation, contents, and function. Arch Biochem Biophys 2000; 379:173. 12. Fulcher KD, Mori C, Welch JE, et al. Characterization of FSC-1 complementary deoxyribonucleic acid for a mouse sperm fibrous sheath component. Biol Reprod 1995; 52: 41. 13. Schalles U, Shao X, van der Hoorn FA, Oko R. Developmental expression of the 84 kDa ODF sperm protein: localization to both the cortex and medulla of outer dense fibers and to the connecting piece. Develop Biol 1998; 199:250. 14. Burfeind P, Hoyer-Fender S. Sequence and developmental expression of a mRNA encoding a putative protein of rat sperm outer dense fibers. Develop Biol 1991; 148:195. 15. O'Bryan MK, Loveland KL, Herzfield D, et al. Identification of a rat testis specific gene encoding a potential rat outer dense fibre protein. Mol Reprod Develop 1998; 50:313. 16. Gibbons IRL. Mechanisms of flagellar motility. In: Afzelius BA, ed. The functional anatomy of the spermatozoon. Oxford: Pergamon Press, 1975:127. 17. Afzelius BA, Eliasson R, Johnsen O, Lindholmer C. Lack of dynein arms in immotile human spermatozoa. J Cell Biol 1975; 66:225.

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Heller CG, Clermont Y. Kinetics of the germinal epithelium in man. Recent Prog Horm Res 1964; 20:545. Fawcett DW. Ultrastructure and function of the Sertoli cell. In: Hamilton DW, Greep RO, eds. Handbook of physiology, section 7, vol 5. Baltimore: Williams & Wilkins, 1975:21. de Kretser DM, Kerr JB. The cytology of the testis. In: Knobil E, Neill JD, eds. Physiology of reproduction. New York: Raven Press, 1994:1177. de Kretser DM, Burger HG. Ultrastructural studies of the human Sertoli cell in normal men and males with hypogonadotropic hypogonadism before and after gonadotropic treatment. In: Saxena BB, Beling CG, Gandy HM, eds. Gonadotropins. New York: Wiley Interscience, 1972:640. Means AR, Fakunding JL, Huckins C, et al. Follicle stimulating hormone, the Sertoli cell and spermatogenesis. Recent Prog Horm Res 1976; 32:477. Dym M, Fawcett DW. The blood–testis barrier in the rat and the physiological compartmentation of the seminiferous epithelium. Biol Reprod 1970; 3:308. Setchell BP, Waites GMH. The blood–testis barrier. In: Hamilton DW, Greep RO, eds. Handbook of physiology, section 7, vol 5. Baltimore: Williams & Wilkins, 1975:143. Flickinger CJ. The postnatal development of the Sertoli cells of the mouse. Z Zellforsch 1967; 78:92. Gilula NB, Fawcett DW, Aoki A. The Sertoli cell occluding junctions and gap junctions in mature and developing mammalian testis. Dev Biol 1976; 50:142. Russell LD. Form, dimensions and cytology of mammalian Sertoli cell. In: Russell LD, Griswold MD, eds. The Sertoli cell. Vienna, IL: Cache River Press, 1993:1. Jutte NHPM, Jansen R, Grootegoed AJ, et al. Regulation of survival of rat pachytene spermatocytes by lactate supply from Sertoli cells. J Reprod Fertil 1982; 65:431. Orth JM, Gunsalus GL, Lamperti AA. Evidence from Sertoli cell–depleted rats indicates that spermatid number in adults depends on numbers of Sertoli cells. Endocrinology 1988; 122:787. Cooke PS, Hess RA, Porcelli J, Meisami E. Increased sperm production in adult rats after transient neonatal hypothyroidism. Endocrinology 1991; 129:244. Bunick KD, Kirby J, Hess RA, Cooke PS. Developmental expression of testis messenger ribonucleic acids in the rat following propylthiouracil– induced neonatal hypothyroydism. Biol Reprod 1994; 51:706. Simorangkir DR, Wreford NG, de Kretser DM. Impaired germ cell development in the testis of immature rats with neonatal hypothyroidism. J Androl 1997; 18:186. Boitani C, Stefanini M, Fragale A, Morena AR. Activin stimulates Sertoli cell proliferation in a defined period of rat testis development. Endocrinology 1995; 136:5438. Setchell BP. Do Sertoli cells secrete fluid into the seminiferous tubules? J Reprod Fertil 1969; 19:391. Jégou B, Le Gac F, de Kretser DM. Seminiferous tubule fluid and interstitial fluid production: I. Effects of age and hormonal regulators in immature rats. Biol Reprod 1982; 27:590. Jégou B, Le Gac F, Irby D, de Kretser DM. Studies on seminiferous tubule fluid production in the adult rat: effect of hypophysectomy and treatment with FSH, LH, and testosterone. Int J Androl 1983; 6:249. French FS, Ritzen EM. A high-affinity androgen binding protein (ABP) in rat testis: evidence for secretion into efferent duct fluid and absorption by epididymis. Endocrinology 1973; 93:88. Gunsalus GL, Musto NA, Bardin CW. Immunoassay of androgen binding protein in blood: a new approach for study of the seminiferous tubule. Science 1978; 200:65. Hansson V, Ritzen EM, French FS, et al. Testicular androgen-binding protein (ABP): comparison of ABP in rabbit testis and epididymis with a similar androgen-binding protein (TeBG) in rabbit serum. Mol Cell Endocrinol 1975; 3:1. Joseph DR, Hall SH, French FS. Rat androgen binding protein: evidence for identical subunits and amino acid sequence homology with human sex hormone binding globulin. Proc Natl Acad Sci U S A 1987; 84:337. Hryb DJ, Khan MS, Romus NA, Rosner W. The solubilization and partial characterization of the sex hormone–binding globulin receptor from human prostate. J Biol Chem 1989; 264:5378. Porto CS, Abreu LC, Gunsalus GL, Bardin CW. Binding of sex hormone– binding globulin (SHBG) to testicular membranes and solubilized receptors. Mol Cell Endocrinol 1992; 89:33. Joseph BR, Becchis M, Fenstermacher DA, Petrusz P. The alternate N-terminal sequence of rat androgen binding protein/sex hormone binding globulin contains a nuclear targetting signal. Endocrinology 1996; 137:1138. Joseph DR, Wang YM, Sullivan PS. Characterization and sex hormone regulation of multiple alternate androgen–binding protein/sex hormone-binding globulin RNA transcript in rat brain. Endocr J 1994; 2:749. Wright WW, Musto NA, Mather JP, Bardin CW. Sertoli cells secrete both testis-specific and serum proteins. Proc Natl Acad Sci U S A 1981; 78:7565. Lacroix M, Smith FE, Fritz IB. Secretion of plasminogen activator by Sertoli cell enriched cultures. Mol Cell Endocrinol 1977; 9:227. Huggenvik J, Sylvester SR, Griswold MD. Control of transferrin in RNA synthesis in Sertoli cells. Ann NY Acad Sci 1984; 438:1. Steinberger A, Steinberger E. Secretion of an FSH-inhibiting factor by cultured Sertoli cells. Endocrinology 1976; 99:918. Le Gac F, de Kretser DM. Inhibin production by Sertoli cells. Mol Cell Endocrinol 1982; 28:487. Au CL, Robertson DM, de Kretser DM. An in vivo method for estimating inhibin production by adult rat testes. J Reprod Fertil 1985; 71:259. Robertson DM, Foulds LM, Leversha L, et al. Isolation of inhibin from bovine follicular fluid. Biochem Biophys Res Commun 1985; 126:220. Mason AJ, Hayflick JS, Ling N, et al. Complementary DNA sequences of ovarian follicular fluid inhibin show precursor structure and homology with transforming growth factor. Nature 1985; 318:659. Forage RG, Ring JM, Brown RW, et al. Cloning and sequence analysis of cDNA species coding for the two subunits of inhibin from bovine follicular fluid. Proc Natl Acad Sci U S A 1986; 83:3091. Massague J. The TGF-b family of growth and differentiation factors. Cell 1987; 49:437. Jensen TK, Andersson AM, Hjollund NHI, et al. Inhibin B as a serum marker of spermatogenesis: correlation to differences in sperm concentration and follicle stimulating hormone levels. A study of 349 Danish men. J Clin Endocrinol Metab 1997; 82:4059. Anderson RA, Wallace EM, Groome NP, et al. Physiological relationships between inhibin B, follicle stimulating hormone secretion and spermatogenesis in normal men and response to gonadotrophin suppression by exogenous testosterone. Hum Reprod 1997; 12:746. Anawalt BD, Bebb RA, Matsumoto AM, et al. Serum inhibin B levels reflect Sertoli cell function in normal men and men with testicular dysfunction. J Clin Endocrinol Metab 1996; 81:3341. Tilbrook AJ, de Kretser DM, Clarke IJ. Human recombinant inhibin A suppresses plasma follicle stimulating hormone to intact levels but has no effect on luteinizing hormone in castrated rams. Biol Reprod 1993; 49:779. Tilbrook AJ, de Kretser DM, Clarke IJ. Human recombinant A and testosterone act directly at the pituitary to suppress plasma concentrations of FSH in castrated rams. J Endocr 1993; 138:181. Josso N, Picard J, Tran D. The antimullerian hormone. Recent Prog Horm Res 1977; 33:117.

60a. de Santa Barbara P, Moniot B, Poulat K, Berta P. Expression and subcellular localization of SF-1, Sox 9, WT1, and AMH proteins during early human testicular development. Dev Dyn 2000; 217:293. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104.

Brokelmann J. Fine structure of germ cells and Sertoli cells during the cycle of the seminiferous epithelium in the rat. Z Zellforsch 1963; 59:820. Christensen AK, Mason NR. Comparative ability of seminiferous tubules and interstitial tissue of rat testes to synthesize androgen from progesterone-4- 15 C in vitro. Endocrinology 1965; 76:646. Dorrington JH, Armstrong DT. Follicle stimulating hormone stimulates estradiol-17b synthesis in cultured Sertoli cells. Proc Natl Acad Sci U S A 1975; 72:2677. Tsai-Morris CH, Aquilano DR, Dufau ML. Cellular localization of rat testicular aromatase activity during development. Endocrinology 1985; 116:38. Van Damme MP, Robertson DM, Marana R, et al. A sensitive and specific in vitro bioassay method for the measurement of follicle stimulating hormone activity. Acta Endocrinol (Copenh) 1979; 91:224. Hodgson Y, Robertson DM, de Kretser DM. The regulation of testicular function. In: Greep RO, ed. International review of physiology, vol 27, reproductive physiology IV. Baltimore: University Park Press, 1983:275. Parvinen M. Regulation of the seminiferous epithelium. Endocr Rev 1982; 3:404. Jégou B, Syed V, Sourdaine P, et al. The dialogue between late spermatids and Sertoli cells in vertebrates: a century of research. In: Nieschlag E, Habernicht UF, eds. Spermatogenesis, fertilization, contraception. Berlin: Springer-Verlag, 1992:57. Toledo SP. Leydig cell hypoplasia leading to two different phenotypes: male pseudohermaphroditism and primary hypogonadism not associated with this. Clin Endocrinol 1992; 36:521. Latronico AC, Anasti J, Arnhold IJ, et al. Brief report: testicular and ovarian resistance to luteinizing hormone caused by inactivating mutations of the luteinizing hormone–receptor gene. N Engl J Med 1996; 334:507. Steinberger E, Root A, Ficher M, Smith KD. The role of androgens in the initiation of spermatogenesis in man. J Clin Endocrinol Metab 1973; 37:746. Burger HG, Baker HWG. Therapeutic considerations and results of gonadotropin treatment in male hypogonadotropic hypogonadism. Ann N Y Acad Sci 1984; 438:447. Paulsen CA. The effect of human menopausal gonadotropin on spermato-genesis in hypogonadotropic hypogonadism. In: Gual C, ed. Proceedings of the sixth Pan-American Congress in Endocrinology: International Congress Series 112. Amsterdam: Excerpta Medica, 1966:398. Singh J, O'Neill C, Handlesman DJ. Induction of spermatogenesis by androgens in gonadotropin deficient (hpg) mice. Endocrinology 1995; 136:5311. Kumar TR, Wang Y, Lu N, Matzuk M. Follicle stimulating hormone is required for ovarian follicle maturation but not male fertility. Nature Genet 1997; 15:201. Tapanainen JS, Aittomaki K, Min J, et al. Men homozygous for an inactivating mutation of the follicle stimulating hormone (FSH) receptor gene present variable suppression of spermatogenesis and fertility. Nature Genet 1997; 15:205. Bremner WJ, Matsumoto AM, Paulsen CA. Gonadotropin control of spermatogenesis in man: studies of gonadotropin administration in spontaneous and experimentally induced hypogonadotropic states. Ann N Y Acad Sci 1984; 438:465. Matsumoto AM, Paulsen CA, Bremner WJ. Stimulation of sperm production by human luteinizing hormone in gonadotrophin-suppressed normal men. J Clin Endocrinol Metab 1984; 55:882. Matsumoto AM, Karpas AE, Bremner WJ. Chronic human chorionic gonadotropin administration in normal men: evidence that follicle stimulating hormone is necessary for the maintenance of quantitatively normal spermatogenesis in man. J Clin Endocrinol Metab 1986; 62:1184. Cunningham GR, Huckins C. Persistence of complete spermatogenesis in the presence of low intratesticular concentration of testosterone. Endocrinology 1979; 105:177. Sun YT, Wreford NG, Robertson DM, de Kretser DM. Quantitative cytological studies of spermatogenesis in intact and hypophysectomised rats: identi-fication of androgen-dependent stages. Endocrinology 1990; 127:1215. Macleod J, Pazianos A, Ray B. The restoration of human spermatogenesis and of the reproductive tract with urinary gonadotropins following hypo-physectomy. Fertil Steril 1966; 17:7. Marshall GR, Wickings EJ, Lüdecke DK, Nieschlag E. Stimulation of spermatogenesis in stalk-sectioned rhesus monkeys by testosterone alone. J Clin Endocrinol Metab 1983; 57:152. Means AR, Dedman JR, Tash JS, et al. Regulation of the testis Sertoli cell by follicle stimulating hormone. Annu Rev Physiol 1980; 42:59. Orth J, Christensen AK. Autoradiographic localization of specifically bound 125 I-labelled follicle stimulating hormone on spermatogonia of the rat testis. Endocrinology 1978; 103:1944. Zhengwei Y, Wreford NG, Royce P, et al. Stereological evaluation of human spermatogenesis following suppression by testosterone treatment: hetero-geneous pattern of spermatogenic impairment. J Clin Endocrinol Metab 1998; 83:1284. Zhengwei Y, Wreford NG, Schlatt S, et al. GnRH antagonist-induced gonadotropin withdrawal acutely and specifically impairs spermatogonial development in the adult macaque (Macaca fascicularis). J Reprod Fertil 1998; 112:139. Grootegoed JA, Peters MJ, Mulder E, et al. Absence of nuclear androgen receptor in isolated germinal cells of rat testes. Mol Cell Endocrinol 1977; 9:159. Sanborn BM, Steinberger A, Tcholakian RK, Steinberger E. Direct measurement of androgen receptors in cultured Sertoli cells. Steroids 1977; 29:493. Bremner WJ, Millar MR, Sharpe RM. Immunohistochemical localization of androgen receptors in the rat testis: evidence of a stage-dependent expression and regulation by androgens. Endocrinology 1994; 135:1227. Lamb DJ, Tsai YH, Steinberger A, Sanborn BM. Sertoli cell nuclear transcriptional activity: stimulation by follicle stimulating hormone and testosterone in vivo. Endocrinology 1981; 108:1020. Louis BG, Fritz IB. FSH and testosterone independently increase the production of ABP by Sertoli cells in culture. Endocrinology 1979; 104:454. Au CL, Irby DC, Robertson DM, de Kretser DM. Effects of testosterone on testicular inhibin and fluid production in intact and hypophysectomized adult rats. J Reprod Fertil 1986; 76:257. Sun YT, Irby DC, Robertson DM, de Kretser DM. The effects of exogenously administered testosterone on spermatogenesis in intact and hypophysectomised rats. Endocrinology 1989; 125:1000. O'Donnell L, McLachlan RI, Wreford NG, et al. Testosterone withdrawal promotes stage-specific detachment of round spermatids from the rat seminiferous epithelium. Biol Reprod 1996; 55:895. O'Donnell L, Stanton PG, Wreford NG, et al. Inhibition of 5a reductase activity impairs T-dependent restoration of spermiogenesis in rats. Endocrinology 1996; 137:2703. Janulis L, Bahr JM, Hess RA, et al. Rat testicular germ cells and epididymal sperm contain active P450 aromatase. J Androl 1998; 19:65. Carreau S, Bilinska B, Levallet J. Male germ cells: a new source of estrogens in the mammalian testis. Ann Endocrinol (Paris) 1998; 59:79. Robertson DM, O'Donnell L, Jones ME, et al. Impairment of spermatogenesis in mice lacking a functional aromatase ( cyp 19) gene. Proc Natl Acad Sci U S A 1999; 96:7986. Hess RA, Bunick D, Lee KH, et al. A role for oestrogens in the male reproductive tract. Nature 1997; 390:509. Saunders PTK, Maguire SM, Gaughan J, Millar MR. Expression of oestrogen receptor beta (ER beta) in multiple cell types including some germ cells, in the rat testis. J Endocrinol 1997; 156:R13. Krege JG, Hodgin JB, Couse JF, et al. Generation and reproductive phenotypes of mice lacking estrogen receptor beta. Proc Natl Acad Sci U S A 1998; 95:15677. Yong EL, Wang Q, Tut TG, et al. Male infertility and the androgen receptor: molecular, clinical and therapeutic aspects. Reprod Med Rev 1997; 6:113. Reijo R, Lee TY, Salo P, et al. Diverse spermatogenic defects in humans caused by Y chromosome deletions encompassing a novel RNA-binding protein gene. Nature Genet 1995; 10:383.

105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127.

Eberhart CG, Maines JZ, Wasserman SA. Meiotic cell cycle requirement for a fly homologue of human deleted in azoospermia. Nature 1996; 381:783. Loveland KL, Schlatt S. Stem cell factor and c-kit in the mammalian testis: lessons originating from mother nature's gene knockouts. J Endocr 1997; 153:337. Dix DJ, Allen JW, Collins BW, et al. Targetted disruption of Hsp 70-2 results in failed meiosis, germ cell apoptosis and male infertility. Proc Natl Acad Sci U S A 1996; 93:3264. Sinha Hikim AP, Swerdloff RS. Hormonal and genetic control of germ cell apoptosis in the testis. Rev Reprod 1999; 4:38. Print CG, Loveland K, Gibson L, et al. Apoptosis regulator Bcl-w is essential for spermatogenesis but is otherwise dispensable. Proc Natl Acad Sci U S A 1998; 95:12424. Fawcett DW, Neaves WB, Flores MN. Comparative observations on intertubular tissue of the mammalian testis. Biol Reprod 1973; 9:500. Lording DW, de Kretser DM. Comparative ultrastructural and histochemical studies of the interstitial cells of the rat testis during fetal and postnatal development. J Reprod Fertil 1972; 29:261. Ewing LL, Zirkin B. Leydig cell structure and steroidogenic activity. Recent Prog Horm Res 1983; 39:599. Clark BJ, Wells J, King SR, Stocco DM. The purification, cloning, and expression of a novel LH-induced mitochondrial protein in MA-10 mouse Leydig tumor cells: characterization of the steroidogenic acute regulatory protein (StAR). J Biol Chem 1994; 269:28314. Bose HS, Pescovitz OH, Miller WL. Spontaneous feminization in a 46XX female patient with congenital lipoid adrenal hyperplasia due to a homozygous frameshift mutation in the acute steroid regulatory protein. J Clin Endocrinol Metab 1997; 82:1511. Fujieda K, Tajima T, Nakae J, et al. Spontaneous puberty in 46XX subjects with congenital lipoid adrenal hyperplasia. J Clin Invest 1997; 99:1265. Catt KJ, Harwood JP, Aguilera G, Dufau ML. Hormonal regulation of peptide receptors and target cell responses. Nature 1979; 280:109. Padron RS, Wischusen J, Hudson B, et al. Prolonged biphasic response of plasma testosterone to single intramuscular injections of human chorionic gonadotropin. J Clin Endocrinol Metab 1980; 50:1100. de Kretser DM. Sertoli cell-Leydig cell interaction in the regulation of testicular function. Int J Androl 1982; 5(Suppl):11. Sharpe RM, Cooper I. Intratesticular secretion of a factor(s) with major stimulating effects on Leydig cell testosterone secretion in vitro. Mol Cell Endocrinol 1984; 37:159. Cohen PE, Hardy MP, Pollard JW. Colony-stimulating factor-1 plays a major role in the development of reproductive function in male mice. Mol Endocrinol 1997; 11:1636. Papadopoulos V. Peripheral-type benzodiazepine/diazepam binding inhibitor receptor: biological role in steroidogenic cell function. Endocrinol Rev 1993;14:222. McLachlan RI, Wreford NG, de Kretser DM, Robertson DM. The effects of testosterone on spermatogenic cell populations in the adult rat. Biol Reprod 1994;51:945. Bergh A. Local differences in Leydig cell morphology in the adult rat testis: evidence for a local control of Leydig cells by adjacent seminiferous tubules. Int J Androl 1982; 5:325. Rich KA, de Kretser DM. Spermatogenesis and the Sertoli cell. In: de Kretser DM, Burger HG, Hudson B, eds. The pituitary and testis: clinical and experimental studies. Berlin: Springer-Verlag, 1983:85. Head JR, Billingham RE. Immune privilege in the testis. II: evaluation of potential local factors. Transplantation 1985; 40:269. Hedger MP. Testicular leucocytes: what are they doing? Rev Reprod 1997; 2:38. Cucicini C, Kercret H, Touzlain AM, et al. Vectorial production of interleukin 1 and interleukin 6 by rat Sertoli cells cultured in a dual culture compartment system. Endocrinology 1997; 138:2863.

CHAPTER 114 EVALUATION OF TESTICULAR FUNCTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 114 EVALUATION OF TESTICULAR FUNCTION STEPHEN J. WINTERS Clinical Evaluation Laboratory Evaluation Testosterone Sex Hormone–Binding Globulin Free or Bioavailable Testosterone Methodology Luteinizing Hormone and Follicle-Stimulating Hormone Dihydrotestosterone Prolactin Estrogens Gonadotropin Subunits Inhibin Müllerian Inhibitory Hormone Gonadotropin Radioreceptor Assays and in Vitro Bioassays Androgen Receptors Functional Tests Semen Analysis Sperm Antibodies Genital Tract Infection Biochemical Analysis of Semen Study of Testicular Tissue Genetic Studies Chapter References

Male hypogonadism signifies impaired production of testosterone by Leydig cells, or deficient spermatogenesis, but in most clinical disorders both compartments of the testis are abnormal. This is not surprising because extensive biochemical communication occurs between the Leydig cells and the seminiferous tubules (see Chap. 113). Hypogonadism due to pathology intrinsic to the testis (primary testicular failure) or to deficient gonadotropin drive to the testis (hypogonadotropic hypogonadism) may produce similar clinical features, and the tests used to diagnose most forms of hypogonadism overlap. Therefore, a general clinical and laboratory approach to evaluating testicular function will be presented.

CLINICAL EVALUATION Patients with impaired testicular function present variably, depending on the age at onset of their disease. Hypogonadism in the fetus results in genital ambiguity (see Chap. 90). When the disturbance begins in childhood, puberty is delayed or does not occur (see Chap. 92). Adult men with hypogonadism often present with a decrease in libido and energy, or with infertility, among other symptoms (Table 114-1), and although these symptoms are sensitive indicators of androgen deficiency, they are nonspecific. For example, a reduced libido in men is also a characteristic of depression as well as of performance anxiety.

TABLE 114-1. Symptoms and Signs of Hypogonadism in Adult Men

A complete physical examination is needed to evaluate hypogonadal males. Prepubertal Leydig cell insufficiency causes eunuchoidism, including a juvenile voice; childlike facies; scant facial, pubic, and body hair; smooth skin; poorly developed skeletal musculature; fat accumulation in the hips, buttocks, and lower abdomen; a small prostate; and failure of epiphyseal closure of the long bones, resulting in an arm span that exceeds the height by >6 cm and disproportionately long legs. The testis size is readily measured by approximating its length and width with a ruler or with a commercially available series of ovoids (see Chap. 93). The testes generally reach adult size by 18 years of age. The median testis length among normal men is 5 cm, equivalent to 25 mL in volume.1 The left testis is often slightly smaller than the right. Although testes that are 4 cm long may be normal, many men with hypospermatogenesis will have testes of this size. There are genetic differences in testis size, with smaller testes found among Asian men. Testis size declines with aging.2 Varicocele (see Chap. 118), a distention of the pampiniform plexus within the scrotum due to dysfunctional valves within the spermatic vein, is a common finding in infertile men, and is associated with a reduction in testis size beginning in adolescence, particularly on the left.3,4 Usually, the venous distention is visible or palpable in the upright posture, increases if the patient performs a Valsalva maneuver, and disappears as soon as the patient is recumbent. Color Doppler ultrasonography can be used to confirm the presence of a varicocele.5 The skin of hypogonadal men may be soft and smooth. Muscle mass may decline, and fat mass may increase. Gynecomastia is a frequent finding in hypogonadal teenagers and in adults. Examination of the male breast can be difficult, however, because the distinction between fat and breast tissue is often inexact. Galactorrhea, on the other hand, is a rare finding in males (see Chap. 13). Physical changes regress slowly in sexually mature men who acquire Leydig cell dysfunction. If some of the physical changes of eunuchoidism in previously normal men are present, androgen deficiency is both severe and long standing. Therefore, testosterone deficiency is often present in men with limited physical findings. Reduced visual acuity or a visual field disturbance suggests a mass lesion in the hypothalamus-pituitary (see Chap. 19). The digital rectal examination is useful in assessing sphincter tone in men with erectile dysfunction, and for estimating prostate size.

LABORATORY EVALUATION TESTOSTERONE Physiologic Aspects of Testosterone Testing. Testosterone is both a paracrine regulator of spermatogenesis (see Chap. 113) and a hemocrine hormone. The testicular content of testosterone is ~50 µg (1 µg/g testis), whereas the blood production rate is ~5000 µg per day, indicating that only a small portion of the testosterone produced each day is stored in the testes. The testosterone precursor steroids, including pregnenolone, 17-OH pregnenolone, dehydroepiandrosterone (DHEA), progesterone, 17a-hydroxyprogesterone, and androstenedione, are also secreted by the testis, and the relative concentrations of these steroids in spermatic vein blood are proportional to their testicular concentrations.6 The release of precursor steroids into the circulation may indicate that they are unnecessary by-products in the orderly biotransformation of pregnenolone to testosterone, because none is known to have a physiologic action in the male. Androstenedione is of special interest because it is used as a performance-enhancing drug. Androstenedione is bioconverted to testosterone and to estrone, but there is little published information on the endocrine profile following androstenedione administration. A single sample of blood, generally drawn in the morning, can be used to measure testosterone. The usual normal range in morning samples is 3 to 10 ng/mL (10–40 nmol/L). There is a diurnal variation in testosterone in adult men, with highest levels in the early morning, followed by a progressive fall throughout the day, reaching the lowest levels in the evening and during the first few hours of sleep. Peak and nadir values differ by ~15%, although more pronounced differences are sometimes

observed.7 The diurnal rhythm is blunted with aging8 and in men with testicular failure.9 The metabolic clearance of testosterone is not thought to vary throughout the day, so that the diurnal testosterone rhythm is presumed to result from a day-night difference in testosterone production. Because there are no parallel changes in serum LH levels, however, the origin as well as physiologic significance of the diurnal testosterone rhythm remains uncertain. It is important to measure testosterone in the morning, because reference ranges are based on morning values. Frequent sampling of spermatic venous blood reveals testosterone secretory pulses at a frequency of ~1 pulse per hour,6 but because of this rapid frequency and relatively low pulse amplitude only small fluctuations occur in peripheral plasma. Testosterone secretory bursts are more readily defined in the peripheral blood when pulse frequency is low.7 There is a good correlation between the plasma testosterone level at first sampling and the mean of multiple samples taken over 1 year, so that one morning sample is reasonably representative,10 although abnormal and borderline values should be confirmed. There is a prominent sleep-related increase in serum testosterone in pubertal boys, with abrupt rises from female to adult male levels.11 This difference can be used clinically to evaluate boys with delayed puberty (see Chap. 92), because the rise to a higher testosterone value in the morning may precede pubertal testis growth and indicates that puberty has begun.12 Men with hypogonadism and hyperprolactinemia have an exaggerated diurnal testosterone rhythm, leading to very low levels in the afternoon and evening, which can explain clinical hypogonadism despite a normal morning total testosterone level,9 as sometimes occurs in men with a prolactinoma. The testosterone level tends to rise during intense exercise because of hemoconcentration13 and to decline 12 to 24 hours later because gonadotropin-releasing hormone (GnRH) secretion is reduced.14 Testosterone levels are also reduced in acute and chronic illness15 and with fasting for at least 48 hours. These factors can confound an evaluation of testicular function. Testosterone Metabolism. The metabolism of testosterone is shown in Figure 114-1. Testosterone is metabolized into two biologically important products, dihydrotestosterone (DHT) and estradiol, by the enzymes 5a-reductase and aromatase, respectively. Most of the metabolism of testosterone occurs in the liver, however, via 3a- and 3b-hydroxysteroid dehydrogenase, 5a- and 5b-steroid reductase, and oxidation of the D-ring to the 17-keto-steroids androsterone (3a-hydroxy-5a-androstane-17-one) and etiocholanolone (3a-hydroxy-5b-androstane-17-one), which are then excreted in the urine. Most of the 10 to 25 mg per day of keto-steroids in the urine of men is of adrenal origin, however, so that the urinary 17-ketosteroid excretion is not a test of testicular function. Testosterone metabolites are also conjugated to sulfuric and glucuronic acids at the 3- or 17-position and excreted in the urine and bile. A small fraction (2%) of the circulating testosterone is excreted unchanged in the urine.

FIGURE 114-1. Metabolic pathways for testosterone. Enzymes are (A) aromatase, (B) 5a-reductase, (C) 17b-hydroxysteroid dehydrogenase, (D) various hydroxylases and transferases, and (E) 3a-hydroxysteroid dehydrogenase. The percentages represent the average percent bioconversion in normal men to these active and inactive metabolities.

SEX HORMONE–BINDING GLOBULIN Of the circulating testosterone in normal men, 6 WBCs per high-power microscopic field in expressed prostatic secretions suggests infection. Seminal plasma pH may be elevated. Cytokines produced by WBCs in response to infection could damage sperm cell membrane integrity and impair fertilization. Infection with Chlamydia trachomatis is now the most common sexually transmitted disease, and is known to cause symptomatic pelvic inflammatory disease in women. Chlamydia urethritis in men has been proposed to produce chronic prostatitis and seminal vesicle infection. Detection of chlamydia in a first-void urine sample or in semen can be accomplished by specific PCR or ligase chain reaction assays. C. trachomatis seems to be rare in the semen and urine of asymptomatic infertile men, however.106 BIOCHEMICAL ANALYSIS OF SEMEN There is a tremendous array of seminal plasma constituents, each of which presumably plays a role in maintaining the proper milieu for fertilization.107 After ejaculation, human semen coagulates because of the formation of a dense fibrous network. Proteolytic enzymes of prostatic origin lyse the fibers in 10 to 30 minutes. Sperm can then be separated from seminal plasma by gentle centrifugation. Because of cell breakage, however, intracellular constituents invariably are present in seminal plasma, and specific seminal plasma constituents may be reduced or absent because they bind to the sperm surface and are removed. The protein content of seminal plasma ranges from 3.5 to 5.0 g/dL. Some of these proteins are identical to that of blood plasma, including transferrin, insulin-like growth factor (IGF)-I and -II, and inhibin; others are specific for semen such as sperm adherins, which are glycoproteins that are thought to play a role in sperm binding to the zona pellucida. The prostate contributes protective redox enzymes such as superoxide dismutase as well as a number of peptidases such as prostate-specific antigen, which, although a marker for prostate cancer, plays a physiologic role in semen liquefaction. Carbohydrates are present in semen, both free and associated with proteins. Fructose is the principal sugar of seminal plasma. Bilateral agenesis or complete obstruction of the seminal vesicles results in ejaculates that are nearly free of fructose. Androgen deficiency impairs the function of the accessory organs, with a reduction in the concentrations of many substances normally present in semen. Among the steroid hormones, testosterone, several of its precursor steroids, DHT, and 17b-estradiol are present in semen.108 Seminal plasma contains many trace elements, with calcium and zinc being the most abundant.109 Carnitine, acetyl-carnitine, glycerylphosphorylcholine, and citric acid are among the products of the human prostate. Carnitine is also present in sperm, and treatment with oral carnitine has been proposed to improve sperm motility. High levels of the cytokine interleukin-6 (IL-6) in semen have been proposed as a marker for infection of the male accessory glands, although no relationship has been shown between seminal plasma IL-6 levels and sperm parameters.110 Prosta-glandins of both the E and F series are produced by the testis and throughout the excurrent duct system, and presumably regulate ejaculatory function. There is substantial concern that environmental pollutants damage the male reproductive system. An estimation of internal dosing can be made by measuring chemicals in serum or semen. Blood lead levels have been a better indicator of seminiferous tubule dysfunction than is the level of lead in semen,111 whereas the concentration of aluminum in spermatozoa may be a more reliable biomarker of aluminum toxicity.112 STUDY OF TESTICULAR TISSUE Over the past 30 years, physical examination of the testes and measurement of plasma testosterone, LH, and FSH levels have replaced testicular biopsy for distinguishing gonadotropin deficiency from primary testicular failure. The use of routine testicular biopsy in men with unexplained infertility has also declined because the finding of damaged seminiferous tubular epithelium with incomplete spermatogenesis has provided little insight into the pathogenesis of male infertility, and the ultimate therapeutic impact of the biopsy results has been limited. However, testicular biopsy is performed in azoospermic men with normal plasma FSH levels in an effort to identify genital tract obstruction, which can be successfully treated by microsurgery. In many centers, fine-needle aspiration biopsy of the testis has replaced open biopsy.113 In addition, transrectal ultrasonography and transurethral vasography can be used to localize the site of an obstruction.114 In men with marked hypospermatogenesis, testicular sperm aspiration (TESA) and epididymal sperm aspiration (PESA) are used to obtain sperm for intracytoplasmic sperm injection (ICSI) when few or no sperm are present in the ejaculate.115 GENETIC STUDIES Cytogenetic studies are helpful in clarifying the cause of primary testicular failure, and for genetic counseling in men planning ICSI because mutations can be passed on to the progeny. Standard chromosomal analyses were abnormal in 13.7% of azoospermic men and 4.6% of oligospermic men.116 Klinefelter syndrome (47,XXY) and its variants (e.g., 46,XY/47,XXY) are the most common cause of azoospermia and are detected by peripheral blood leukocyte karyotyping with banding procedures, and no longer by the examination of buccal mucosal cells for condensed chromatin (Barr body). Genes on the long arm of the Y chromosome are required for spermatogenesis. This region of the Y is known as the AZF region because it contains genes related to azoospermia. Using the PCR to analyze DNA from peripheral blood leukocytes, small deletions of AZF genes, which escape detection under the microscope, can be identified in 15% to 30% of men with azoospermia.117 The absence of these deletions in the fathers of infertile men indicates that they represent de novo mutations, and provides good evidence that they relate to male infertility. Because, with ICSI, Y microdeletions will be passed on to sons, screening tests for Y microdeletions have been recommended. Congenital bilateral absence of the vas deferens accounts for 3.5% to 8.0% of cases of azoospermia, and represents a mild form of cystic fibrosis in ~70% of cases.118 This autosomal recessive disorder results from mutations involving the cystic fibrosis transmembrane conductase regulator gene, which codes for a membrane protein that functions as an ion channel, and appears to play a role in the development of the epididymis, seminal vesicles, and vas deferens. Testing for this mutation should be performed in men with obstructive azoospermia. Computed tomographic (CT) scans of the paranasal sinuses, chest radiography, and pulmonary function tests should also be obtained, and affected men should be cautioned not to smoke cigarettes. CHAPTER REFERENCES 1. Takihara H, Sakatoku J, Fujii M, et al. Significance of testicular size measurement. In: Andrology. I. A new orchiometer and its clinical application. Fertil Steril 1983; 39:836. 2. Stearns EL, MacDonnell JA, Kaufman BJ, et al. Decline of testicular function with age: hormonal and clinical correlates. Am J Med 1974; 57:761.

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24.

Sawczun IS, Hensle TW, Burbie KA, Nagler HM. Varicoceles: effect on testicular volume in prepubertal and pubertal males. Urology 1993; 41:466. Haans LCF, Laven JSE, Mali WPThM, et al. Testis volumes, semen quality, and hormonal patterns in adolescents with and without a varicocele. Fertil Steril 1991; 56:731. Chiou RK, Anderson JC, Wobig RK, et al. Color Doppler ultrasound criteria to diagnose varicoceles: correlation of a new scoring system with physical examination. Urology 1997; 50:953. Winters SJ, Takahashi J, Troen P. Secretion of testosterone and its delta-4 precursor steroids into spermatic vein blood in men with varicocele-associated infertility. J Clin Endocrinol Metab 1999; 84:997. Spratt DI, O'Dea L St L, Schoenfeld D, et al. Neuroendocrine-gonadal axis in men: frequent sampling of LH, FSH and testosterone. Am J Physiol 1988; 254:E658. Tenover JS, Matsumoto AM, Clifton DK, Bremner WJ. Age-related alterations in the circadian rhythms of pulsatile luteinizing hormone and testosterone secretion in healthy men. J Gerontol 1988; 43:M163. Winters SJ. Diurnal rhythm of testosterone and luteinizing hormone in hypogonadal men. J Androl 1991; 12:185. Vermeulen A, Verdonck G. Representativeness of a single point plasma testosterone level for the long term hormonal milieu in men. J Clin Endocrinol Metab 1992; 74:939. Boyar RN, Rosenfeld RS, Kapen S, et al. Human puberty: simultaneous augmented secretion of luteinizing hormone and testosterone during sleep. J Clin Invest 1974; 54:609. Wu F, Brown DC, Butler GE, et al. Early morning plasma testosterone is an accurate predictor of imminent pubertal development in prepubertal boys. J Clin Endocrinol Metab 1993; 76:26. Zmuda JM, Thompson PD, Winters SJ. Exercise increases serum testosterone and sex hormone-binding globulin levels in older men. Metabolism 1996; 45:935. Kujala H, Alem M, Huhtaniemi IT. Gonadotrophin-releasing hormone and human chorionic gonadotropin tests reveal that both hypothalamic and testicular endocrine functions are suppressed during acute prolonged exercise. Clin Endocrinol 1990; 33:219. Turner HE, Wass JAH. Gonadal function in men with chronic illness. Clin Endocrinol 1997; 47:379. Dunn JF, Nisula BC, Rodbard D. Transport of steroid hormones: binding of 21 endogenous steroids to both testosterone-binding globulin and corticosteroid-binding globulin in human plasma. J Clin Endocrinol Metab 1981; 53:58. Petra PH. The plasma sex steroid binding protein (SBP or SHBG). A critical review of recent developments on the structure, molecular biology and function. J Steroid Biochem Molec Biol 1991; 40:735. Terasaki T, Nowlin DM, Pardridge WM. Differential binding of testosterone and estradiol to isoforms of sex hormone-binding globulin: selective alteration of estradiol binding in cirrhosis. J Clin Endocrinol Metab 1988; 67:639. Joseph DR. Structure, function, and regulation of androgen-binding protein/sex hormone-binding globulin. Vitam Horm 1994; 49:197. Wheeler MJ. The determination of bioavailable testosterone. Ann Clin Biochem 1995; 32:345. Masters AM, Hahnel R. Investigation of sex-hormone binding globulin interference in direct radioimmunoassays for testosterone and estradiol. Clin Chem 1989; 35:979. Wheeler MJ, D'Souza A, Matadeen J, et al. Ciba Corning ACS:180 testosterone assay evaluated. Clin Chem 1996; 42:1445. Sodergard R, Backstrom T, Shanbhag V, Carstensen H. Calculation of free and bound fractions of testosterone and estradiol 17-b to human plasma proteins at body temperature. J Steroid Biochem 1982; 16:810. Kapoor P, Luttrell BM, Williams D. The free androgen index is not valid for adult males. J Steroid Biochem 1993; 45:325.

24a.Rosner W. Errors in the measurement of plasma free testosterone. J Clin Endrocrinol Metab 1997; 82:2014. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.

Winters SJ, Kelley DE, Goodpaster B. The analog free testosterone assay: are the results in men clinically useful? Clin Chem 1998; 44:2178. Manni A, Pardridge WM, Cefalu W, et al. Bioavailability of albumin-bound testosterone. J Clin Endocrinol Metab 1985; 61:705. Vermeulen A, Kaufman JM, Giagulli. Influence of some biological indexes on sex hormone-binding globulin and androgen levels in aging or obese males. J Clin Endocrinol Metab 1996; 81:1821. Cox C, Caulier C, Havelange G, et al. Two-sites immunoradiometric assay using monoclonal antibodies for the determination of serum human sex hormone binding globulin. J Immunoassay 1992; 13:355. Nisula BC, Loriaux DL, Wilson YA. Solid phase method for measurement of the binding capacity of testosterone-estradiol binding globulin in human serum. Steroids 1979; 31:681. Pugeat M, Crave JC, Tournaire J, Forest MG. Clinical utility of sex hormone-binding globulin measurement. Horm Res 1996; 45:148. Chin WW, Boime I, eds. Glycoprotein hormones. Norwell, MA: Serono Symposia, 1990. Shupnik MA. Gonadotropin gene modulation by steroids and gonadotropin-releasing hormone. Biol Reprod 1996; 54:279. Nankin HR, Troen P. Repetitive luteinizing hormone elevations in serum in normal men. J Clin Endocrinol Metab 1971; 33:558. Veldhuis JD, King JC, Urban RJ, et al. Operating characteristics of the male hypothalamo-pituitary-gonadal axis: pulsatile release of testosterone and follicle-stimulating hormone and their temporal coupling with luteinizing hormone. J Clin Endocrinol Metab 1987; 65:929. Haavisto A-M, Pettersson K, Bergendahl M, et al. Occurrence and biological properties of a common genetic variant of luteinizing hormone. J Clin Endocrinol Metab 1995; 80:1257. Wu FCW, Butler GE, Kelnar CJH, et al. Patterns of pulsatile luteinizing hormone and follicle-stimulating hormone secretion in prepubertal (midchildhood) boys and girls and patients with idiopathic hypogonadotropic hypogonadism (Kallmann's syndrome): a study using an ultrasensitive time-resolved immunofluorometric assay. J Clin Endocrinol Metab 1991; 72:1229. Odink RJ, Schoemaker J, Schoute E, et al. Predictive value of serum follicle-stimulating hormone levels in the differentiation between hypogonadotropic hypogonadism and constitutional delay of puberty. Horm Res 1998; 49:279. Anawalt BD, Bebb RA, Matsumoto AM, et al. Serum inhibin B levels reflect Sertoli cell function in normal men and men with testicular dysfunction. J Clin Endocrinol Metab 1996; 81:3341. Rosen SW, Weintraub BD. Monotropic increase of serum FSH correlated with low sperm count in young men with idiopathic oligospermia and aspermia. J Clin Endocrinol Metab 1972; 32:410. Snyder PJ. Gonadotroph cell adenomas of the pituitary. Endocr Rev 1985; 6:552. Vos P, Croughs RJM, Thijssen JHH, et al. Response of luteinizing hormone secreting pituitary adenoma to a long-acting somatostatin analog. Acta Endocrinol 1988; 118:587. Ito T, Horton R. The source of plasma dihydrotestosterone in man. J Clin Invest 1971; 50:1621. McConnell JD, Wilson JD, George FW, et al. Finasteride, an inhibitor of 5a-reductase, suppresses prostatic dihydrotestosterone in men with benign prostatic hyperplasia. J Clin Endocrinol Metab 1992; 74:505. Thigpen AE, Silver RI, Guileyardo JM, et al. Tissue distribution and ontogeny of steroid 5a-reductase isozyme expression. J Clin Invest 1993; 92:903. Belanger A, Candas B, Dupont A, et al. Changes in serum concentrations of conjugated and unconjugated steroids in 40- to 80-year old men. J Clin Endocrinol Metab 1994; 79:1086. Horton R. Dihydrotestosterone is a peripheral paracrine hormone. J Androl 1992; 13:23. Lookingbill DP, Egan N, Santen RJ, Demers LM. Correlation of serum 3a-androstenediol glucuronide with acne and chest hair density in men. J Clin Endocrinol Metab 1988; 67:986. Perryman RL, Thorner MO. The effects of hyperprolactinemia on sexual and reproductive function in men. J Androl 1981; 2:233. Spitz IM, LeRoith D, Livshin J, et al. Exaggerated prolactin response to TRH and metoclopramide in primary testicular failure. Fertil Steril 1980; 34:573. Winters SJ, Johnsonbaugh RE, Sherins RJ. The response of prolactin to chlorpromazine stimulation in men with hypogonadotropic hypogonadism and early pubertal boys: relationship to sex steroid exposure. Clin Endocrinol (Oxf) 1982; 10:321. Sigman M, Jarow JP. Endocrine evaluation of infertile men. Urology 1997; 50:659. Nagel SC, vom Saal FS, Welshons WV. The effective free fraction of estradiol and xenoestrogens in human serum measured by whole cell uptake assays: physiology of delivery modifies estrogenic activity. Proc Soc Exp Biol Med 1998; 217:300. Khosla S, Melton LJ, Atkinson EJ, et al. Relation of serum sex steroid levels and bone turnover markers with bone mineral density in men and women: a key role for bioavailable estrogens. J Clin Endocrinol Metab 1998; 83:2266. Stanik S, Dornfeld LP, Maxwell MH, et al. The effect of weight loss on reproductive hormones in obese men. J Clin Endocrinol Metab 1981; 53:828. Schioler V, Thode J. Six direct radioimmunoassays of estradiol evaluated. Clin Chem 1988; 34:949. Winters SJ, Troen P. Pulsatile secretion of immunoreactive a-subunit in man. J Clin Endocrinol Metab 1985; 60:344. Winters SJ, Troen P. a-Subunit secretion in men with idiopathic hypogonadotropic hypogonadism. J Clin Endocrinol Metab 1988; 66:338. Pralong FP, Pavlou SN, Waldstreicher J, et al. Defective regulation of glyco-protein free a-subunit in males with isolated gonadotropin-releasing hormone deficiency—a clinical research center study. J Clin Endocrinol Metab 1995; 80:3682. Kourides IA, Weintraub BD, Ridgway EC, Maloof F. Pituitary secretion of free alpha and beta subunit of human thyrotropin in patients with thyroid disorders. J Clin Endocrinol Metab 1975; 40:872. Blackman MR, Weintraub BD, Kourides IA, et al. Discordant elevation of the common a-subunit of the glycoprotein hormones compared to b-subunits in serum of uremic patients. J Clin Endocrinol Metab 1981; 53:39. Somjen D, Tordjman K, Kohen F, et al. Combined beta FSH and beta LH response to TRH in patients with clinically non-functioning pituitary adenomas. Clin Endocrinol 1997; 46:555. Fein HG, Rosen SW, Weintraub BD. Increased glycosylation of serum human chorionic gonadotropin and subunits from eutopic and ectopic sources: comparison with placental and urinary forms. J Clin Endocrinol Metab 1980; 50:1111. Saller B, Clara R, Spottl G, et al. Testicular cancer secretes intact human choriogonadotropin (hCG) and its free b-subunit: evidence that hCG (+hCG-b) assays are the most reliable in diagnosis and follow-up. Clin Chem 1990; 36:234. Burger HG. Inhibin in the male: progress at last. Endocrinology 1997; 138:1361. Hayes FJ, Hall JE, Boepple PA, Crowley WF Jr. Differential control of gonadotropin secretion in the human: endocrine role of inhibin. J Clin Endocrinol Metab 1998; 83:1835.

65a.Anderson RA, Sharpe RM. Regulation of inhibin production in the human male and its clinical applications. Int J Androl 2000; 23:136. 66. Illingworth PJ, Groome NP, Byrd W, et al. Inhibin-B: a likely candidate for the physiologically important form of inhibin in men. J Clin Endocrinol Metab 1996; 81:1321. 67. Robertson DM, Cahir N, Findlay JK, et al. The biological and immunological characterization of inhibin A and B forms in human follicular fluid and plasma. J Clin Endocrinol Metab 1997; 82:889. 68. Groome NP, Illingworth PJ, O'Brien M, et al. Quantification of inhibin pro-aC-containing forms in human serum by a new ultrasensitive two-site enzyme-linked immunosorbent assay. J Clin Endocrinol Metab 1995; 80:2926. 69. Andersson AM, Toppari J, Haavisto AM, et al. Longitudinal reproductive hormone profiles in infants: peak of inhibin B levels in infant boys exceeds levels in adult men. J Clin Endocrinol Metab 1998; 83:675. 70. Nachtigall LB, Boepple PA, Seminara SB, et al. Inhibin B secretion in males with gonadotropin releasing hormone (GnRH) deficiency before and during long-term GnRH replacement: relationship to spontaneous puberty, testicular volume, and prior treatment—a clinical research center study. J Clin Endocrinol Metab 1996; 81:3520. 71. Wallace EM, Groome NP, Riley SC, et al. Effects of chemotherapy-induced testicular damage on inhibin, gonadotropin and testosterone secretion: a prospective longitudinal study. J Clin Endocrinol Metab 1997; 82:3111. 72. Petersen P, Andersson A-M, Rorth M, et al. Undetectable inhibin B serum levels in men after testicular irradiation. J Clin Endocrinol Metab 1999; 84:213. 73. Pierik FH, Vreeburg JTM, Stijnen T, et al. Serum inhibin B as a marker of spermatogenesis. J Clin Endocrinol Metab 1998; 83:3110. 74. Petersen PM, Skakkebaek NE, Vistisen K, et al. Semen quality and reproductive hormones before orchiectomy in men with testicular cancer. J Clin Oncol 1999; 17:941. 74a.Hiort O, Holterhus PM. The molecular basis of male sexual differentiation. Eur J Endocrinol 2000; 142:101. 75. Lee MM, Donahoe PK, Silverman BL, et al. Measurements of serum müllerian inhibiting substance in the evaluation of children with nonpalpable gonads. N Engl J Med 1997; 336:1480. 76. Pierce JG, Parsons TF. Glycoprotein hormones: structure and function. Annu Rev Biochem 1982; 50:465. 77. van Damme M-P, Robertson DM, Diczfalusy E. An improved in vitro bioassay method for measuring luteinizing hormone (LH) activity using mouse Leydig cell preparations. Acta Endocrinol 1974; 77:655. 78. Dahl KD, Stone MP. FSH isoforms, radioimmunoassays, bioassays, and their significance. J Androl 1992; 13:11. 79. Jaakkola T, Ding Y-Q, Kellokumpu-Lehtinen P, et al. The ratios of serum bioactive/immunoreactive luteinizing hormone and follicle-stimulating hormone in various clinical conditions with increased and decreased gonadotropin secretion: reevaluation by a highly sensitive immunometric assay. J Clin Endocrinol Metab 1990; 70:1496. 80. Christin-Maitre S, Bouchard P. Bioassays of gonadotropins based on cloned receptors. Mol Cell Endocrinol 1996; 125:151. 81. Beato M, Truss M, Chavez S. Control of transcription by steroid hormones. Ann NY Acad Sci 1996; 784:93. 82. Deslypere JP, Young M, Wilson JD, McPhaul MJ. Testosterone and 5 alpha-dihydrotestosterone interact differently with the androgen receptor to enhance transcription of the MMTV-CAT reporter gene. Mol Cell Endocrinol 1992; 88:15. 83. Quigley CA, de Bellis A, Marschke E, et al. Androgen receptor defects: historical, clinical and molecular perspectives. Endocr Rev 1995; 16:27. 84. Aiman J, Griffin JE. The frequency of androgen receptor deficiency in infertile men. J Clin Endocrinol Metab 1982; 54:725.

85. Wang Q, Ghadessy FJ, Yong EL. Analysis of the transactivation domain of the androgen receptor in patients with male infertility. Clin Genet 1998; 54:185. 86. MacLean HE, Warne GL, Zajac JD. Spinal and bulbar muscular atrophy: androgen receptor dysfunction caused by a trinucleotide repeat expansion. J Neurol Sci 1996; 135:149. 87. Tut TG, Ghadessy FJ, Trifiro MA, et al. Long polyglutamine tracts in the androgen receptor are associated with reduced trans-activation, impaired sperm production, and male infertility. J Clin Endocrinol Metab 1997; 82:3777. 88. Kantoff P, Giovannucci E, Brown M. The androgen receptor CAG repeat polymorphism and its relationship to prostate cancer. Biochim Biophys Acta 1998; 1378:C1. 89. Forest MG. Pattern of the response of testosterone and its precursors to human chorionic gonadotropin stimulation in relation to age in infants and children. J Clin Endocrinol Metab 1979; 49:132. 90. Aynsley-Green A, Zachmann M, Illig R, et al. Congenital bilateral anorchia in childhood: a clinical, endocrine and therapeutic evaluation of twenty-one cases. Clin Endocrinol (Oxf) 1976; 5:381. 91. Cisek LJ, Peters CA, Atala A, et al. Current findings in diagnostic laparoscopic evaluation of the nonpalpable testis. J Urol 1998; 160:1145. 92. Lee PA, Danish RK, Mazur T, Migeon CJ. Micropenis III. Primary hypogonadism, partial androgen insensitivity syndrome, and idiopathic disorders. Johns Hopkins Med J 1980; 147:175. 93. de Kretser DM, Burger HG, Hudson B, Keogh EJ. The hCG stimulation test in men with testicular disorders. Clin Endocrinol (Oxf) 1975; 4:591. 94. Vermeulen A. Decreased androgen levels and obesity in men. Ann Med 1996; 28:135. 95. Korenman SG, Morley JE, Mooradian AD, et al. Secondary hypogonadism in older men: its relation to impotence. J Clin Endocrinol Metab 1990; 71:963. 96. Tenover JS, Matsumoto AM, Plymate SR, Bremner WJ. The effects of aging in normal men on bioavailable testosterone and luteinizing hormone secretion: response to clomiphene citrate. J Clin Endocrinol Metab 1987; 65:1118. 97. Harman SM, Tsitouras PD, Costa PT, et al. Evaluation of pituitary gonadotropic function in men: value of luteinizing hormone-releasing hormone response versus basal luteinizing hormone level for discrimination of diagnosis. J Clin Endocrinol Metab 1982; 54:196. 98. Hudson RW. The endocrinology of varicoceles. Fertil Steril 1988; 49:199. 99. Ghai K, Cara JF, Rosenfield RL. Gonadotropin releasing hormone agonist (nafarelin) test to differentiate gonadotropin deficiency from constitutionally delayed puberty in teen-age boys—a clinical research center study. J Clin Endocrinol Metab 1995; 80:2980. 100. World Health Organization. WHO laboratory manual for the examination of human semen and sperm-cervical-mucus interaction, 3rd ed. Cambridge, England: Cambridge University Press, 1992. 101. Zuckerman Z, Rodriguez-Rigau L, Smith KD, Steinberger E. Frequency distribution of sperm counts in fertile and infertile males. Fertil Steril 1977; 28:1310. 102. Bonde JPE, Ernst E, Jensen EK, et al. Relation between semen quality and fertility: a population-based study of 430 first-pregnancy planners. Lancet 1998; 352:1172. 103. Critser JK, Noiles EE. Bioassays of sperm function. Semin Reprod Endocrinol 1993; 11:1. 103a.Carrell DT. Semen analysis at the turn of the century: an evaluation of potential uses of new sperm function assays. Arch Androl 2000; 44:65. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118.

Adeghe JH. Male subfertility due to sperm antibodies: a clinical overview. Obstet Gynecol Surv 1993; 48:1. Clarke GN, Bourne J, Baker HWG. Intracytoplasmic sperm injections for treating infertility associated with sperm immunity. Fertil Steril 1997; 68:112. Fujisawa M, Nakano Y, Matsui T, et al. Chlamydia trachomatis detected by ligase chain reaction in the semen of asymptomatic patients without pyospermia or pyuria. Arch Androl 1999; 42:41. Mann T, Lutwak-Mann C. Male reproductive function and semen. New York: Springer-Verlag, 1981. Garcia Diez LC, Gonzalez Buitrago IM, Corrales IJ, et al. Hormone levels in serum and seminal plasma of men with different types of azoospermia. J Reprod Fertil 1983; 67:208. Abou-Shakra FR, Ward NI, Everard DM. The role of trace elements in male infertility. Fertil Steril 1989; 52:307. Matalliotakis I, Kirakou D, Fragouli I, et al. Interleukin-6 in seminal plasma of fertile and infertile men. Arch Androl 1998; 41:43. Alexander BH, Checkoway H, Faustman EM, et al. Contrasting associations of blood and semen lead concentrations with semen quality among lead smelter workers. Am J Industr Med 1998; 34:464. Hovatta O, Venalainen ER, Kuusimaki L, et al. Aluminum, lead, and cadmium concentrations in seminal plasma and spermatozoa, and semen quality in Finnish men. Hum Reprod 1998; 13:115. Dajani YF, Kilani Z. Role of testicular fine needle aspiration in the diagnosis of azoospermia. Int J Androl 1998; 21:295. Kuligowska E, Baker CE, Oates RD. Male infertility: role of transrectal US in diagnosis and management. Radiology 1992; 185:353. Belker AM, Sherins RJ, Dennison-Lagos L, et al. Percutaneous testicular sperm aspiration: a convenient and effective office procedure to retrieve sperm for in vitro fertilization with intracytoplasmic sperm injection. J Urol 1998; 160:2058. Van Assche E, Bonduelle M, Tournaye H, et al. Cytogenetics of infertile men. Hum Reprod 1996; 11(Suppl 4):1. Roberts KP. Y-chromosome deletions and male infertility. State of the art and clinical implications. J Androl 1998; 19:255. Durieu I, Bey-Omar F, Rollet J, et al. Diagnostic criteria for cystic fibrosis in men with congenital absence of the vas deferens. Medicine 1995; 74:42.

CHAPTER 115 MALE HYPOGONADISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 115 MALE HYPOGONADISM STEPHEN R. PLYMATE Clinical Characteristics of Hypogonadism Primary Hypogonadism Klinefelter Syndrome XX Males XY/XO Mixed Gonadal Dysgenesis XYY Syndrome Ullrich-Noonan Syndrome Myotonic Dystrophy Sertoli-Cell-Only Syndrome Functional Prepubertal Castrate Syndrome Enzymatic Defects Involving Testosterone Biosynthesis 5a-Reductase Deficiency Luteinizing Hormone/Gonadotropin–Resistant Testis Persistent Müllerian Duct Syndrome Male Pseudohermaphroditism Involving Androgen-Receptor Defects Postpubertal Orchitis (Epidemic Parotitis) Cryptorchidism Leprosy Testicular Trauma Testicular Irradiation Effects Autoimmune Testicular Failure Chemotherapy Effects Secondary Hypogonadism Hypogonadotropic Hypogonadism Classic Hypogonadotropic Eunuchoidism Isolated Deficiency of Luteinizing Hormone or Follicle-Stimulating Hormone Acquired Forms of Gonadotropin Deficiency Prolactin-Secreting Pituitary Tumors Infections of the Hypothalamus and Pituitary Severe Systemic Illness Uremia Lymphocytic Hypophysitis Hemochromatosis Combined Primary and Secondary Hypogonadism Aging Hepatic Cirrhosis Sickle Cell Disease Chapter References

Male hypogonadism may be defined as a failure of the testes to produce testosterone, spermatozoa, or both (Table 115-1). This may be caused by a failure of the testes or of the anterior pituitary. Hypogonadism may also occur if a testicular product is unable to exert an effect, as in the androgen-resistance syndromes (Table 115-2).

TABLE 115-1. Classification of Male Hypogonadism

TABLE 115-2. Manifestations of Testicular Androgen Failure

CLINICAL CHARACTERISTICS OF HYPOGONADISM The clinical presentation of hypogonadism depends on whether the onset was in utero, prepubertal, or postpubertal. If hypogonadism is present because of a defect that occurred in utero, the individual will have ambiguous genitalia (see Chap. 77 and Chap. 90). The clinical pictures of testicular androgen failure of prepubertal and postpubertal onset are presented in Table 115-2. Although the findings on physical examination may be normal, a problem with seminiferous tubule function may manifest as infertility. With the development of sensitive assays for testosterone, an increasing number of circumstances have been described in which serum testosterone levels are lower than normal without any obvious end-organ deficiencies. Examples of this phenomenon are seen in aging or stressed men who have low levels of serum testosterone without definitive evidence of end-organ deficiency. In the case of stress, end-organ deficiency may not be seen, because the period of hypogonadism is transient. Deciding whether these men really are androgen deficient or whether they simply are displaying a physiologic response to stress or age may be difficult. These situations pose further difficulties for clinicians who must decide whether androgen replacement is needed. The classic states of androgen deficiency are discussed in this chapter; androgen replacement therapy, described in Chapter 119, is mentioned. In those situations in which obvious deficiency states are not present, however, the indication for replacement therapy

may not be clear-cut. In some of these patients, the finding of an exaggerated gonadotropin response to luteinizing hormone–releasing hormone (LHRH) may help to define the presence of testicular failure. Functionally, the hypogonadal states may be classified according to the level at which the hypothalamic–pituitary–testicular axis is defective. Briefly, the control of testicular function begins with the release, in a pulsatile fashion, of LHRH from the hypothalamus (see Chap. 16). LHRH, transported by the hypothalamic-pituitary portal system, then causes the release of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) from the anterior pituitary. An optimal rate of pulsation (3.8 pulses every 6 hours) appears to be necessary for adequate secretion of both LH and FSH for normal gonadal function. When the rate is slower than optimal, FSH may be preferentially released in greater amounts than normal and LH in lesser amounts. When the pulse frequency of LHRH is more rapid than normal, serum FSH levels may be suppressed and LH release preferentially stimulated. LH subsequently binds to the Leydig cell to initiate testosterone synthesis and secretion. FSH binds to the Sertoli cell and stimulates the production of several factors that, with testosterone from the Leydig cell, induce and maintain normal spermatogenesis (see Chap. 113). LH and FSH release also are regulated by a negative feedback system (i.e., serum testosterone and estradiol). Inhibin, an FSH-stimulated Sertoli cell peptide, can partially block FSH release from the pituitary without influencing LH release. Serum testosterone and inhibin also may affect the release of LHRH from the hypothalamus. However, the precise role of inhibin in the feedback process in men is not completely defined. In addition to the direct effects of testosterone on sexual tissues, the secretion of androgens is related to other endocrine systems that may affect body habitus during both puberty and adulthood. This is especially true for the somatotropin axis, in which androgens are necessary for normal growth hormone and insulin-like growth factor-I secretion. Furthermore, in normal men, components of the gonadal axis are regulated by other endocrine systems (e.g., insulin pulsation closely determines the blood levels of sex hormone–binding globulin [SHBG]). Therefore, hypogonadism may result from abnormalities in multiple systems, and its manifestations are evident in most physiologic systems.1,2 and 3 Based on this understanding of the hypothalamic–pituitary–gonadal axis, male hypogonadal disorders may be classified into two broad categories. The first is primary hypogonadism, in which the dysfunction is in the testis. Primary hypogonadism is manifested by a deficiency in the main testicular products, testosterone or sperm. The basis for the primary hypogonadism is a testicular defect. In states of primary hypogonadism, negative feedback by testicular products such as testosterone or inhibin on the hypothalamus and pituitary is lost, so that serum LH and FSH levels are elevated in the basal state or, despite normal basal gonadotropin levels, an exaggerated gonadotropin response to LHRH occurs. The presence of testosterone alone is not enough to achieve normal male development. A portion of the Y chromosome also is needed for normal testicular development and regression. In humans, this testis-determining factor has been mapped to a 35-kilobase (kb) segment on the short arm of the Y chromosome close to the pseudoautosomal region. The gene isolated from this locus that equates with the testis-determining factor has been called sex-determining region Y (SRY).4,5 Secondary hypogonadism (i.e., decreased gonadotropin stimulation of potentially normal testes) presents with low serum testosterone levels or decreased sperm production and low serum gonadotropin levels (or values inappropriately low for the level of serum testosterone, sperm production, or both). Since the discovery of LHRH, some authors have divided secondary hypogonadism into pituitary failure and hypothalamic failure (“tertiary hypogonadism”). In this chapter, the division into primary and secondary hypogonadism is used, with the latter including both pituitary and hypothalamic disorders. Table 115-1 lists the categories of diseases discussed.

PRIMARY HYPOGONADISM Including infertile men, primary testicular failure affects 5% to 10% of the male population.4 Although male infertility is commonly considered a problem that involves exclusively seminiferous tubular function and spermatogenesis, evidence from men with varicoceles has demonstrated the presence of an exaggerated response of both serum LH and FSH to LHRH.6 Although total serum testosterone levels are normal in these patients, the augmented gonadotropin response to LHRH indicates a failure in both Leydig cell and seminiferous tubular function (Fig. 115-1). Some compensation for the impaired Leydig cells must occur that is sufficient to return testosterone levels to normal under the influence of increased LH stimulation. These findings confirm that infertile men often possess primary gonadal failure of a subtle nature. Because the incidence of male infertility is fairly high, primary testicular dysfunction in the male population is an important issue.

FIGURE 115-1. Exaggerated luteinizing hormone (LH) and follicle-stimulating hormone (FSH) response to gonadotropin-releasing hormone in fertile and infertile men with a varicocele is compared with that of normal men. Results suggest impairment of both seminiferous tubule and Leydig cell function. Bars indicate standard error of the mean. (Fert. Varic., fertile men with varicocele; Infert. Varic., infertile men with varicocele.) (From Nagao RR, Plymate SR, Berger RE, et al. Comparison of gonadal function between fertile and infertile men with varicoceles. Fertil Steril 1986; 46:930.)

KLINEFELTER SYNDROME The chromosomal constitution of 47,XXY epitomizes the classic form of male primary testicular failure. This abnormality is present in ~1 in 400 men. Klinefelter syndrome was first described in 1942 in nine male patients who, at puberty, experienced the onset of bilateral gynecomastia, small testes with Leydig cell dysfunction and azoospermia, and increased urinary gonadotropin excretion.7 Later, Leydig cell failure was shown to be variable in its magnitude. In 1956, the X chromatin body (Barr body) was found in these individuals, and in 1959, the XXY chromosome constitution was first described, demonstrating that the disease was the result of an extra X chromosome (see Chap. 90).8,9 PHENOTYPIC MANIFESTATIONS The phenotypic manifestations of Klinefelter syndrome are characteristic for the classic form of the disease in which all cells carry the XXY karyotype. Many men with Klinefelter syndrome have a mosaic form, in which some cell lines are XXY and others are XY. In these mosaic individuals, all cell lines are from a single zygote, and the XXY cell lines arise from mitotic nondisjunction after fertilization. Manifestation of the disease may not be typical or consistent. Before puberty, the only physical findings are the small testes; in the classic form of Klinefelter syndrome, a gonadal volume of 45 years.31,32 Sixty-five percent of patients with Klinefelter syndrome have a maternal origin for their extra X chromosomes. Maternal age is thought to be a factor because of the longer diplotene stage of the ova in older women (see Chap. 94). The existence of mothers with an XX/XXX chromosome pattern (see earlier) and an increased incidence of twins with Klinefelter syndrome, as well as the finding of an increased frequency of Klinefelter syndrome in patients with Down syndrome, suggest that some families may have a factor predisposing to chromosomal nondisjunction. DISEASE ASSOCIATIONS AND MEDICAL COMPLICATIONS Numerous disease associations have been made with Kline-felter syndrome; especially prominent are malignancies and autoimmune diseases. Malignancies include breast carcinoma (with an incidence 20 times greater than that of men with an XY chromosome pattern and 20% that of women), nonlymphocytic leukemia, lymphomas, marrow dysplastic syndromes, and extragonadal germ cell neoplasms.33,34,35,36 and 37 Autoimmune diseases, especially chronic lymphocytic thyroiditis and rare syndromes such as Takayasu arteritis, have been reported in Klinefelter syndrome.38,39,40,41 and 42 Whether this results from an effect of decreased androgens on the OKT4/OKT8 lymphocyte ratios or from the combined effect of decreased immune surveillance and increased X chromosome material is unknown.43,44 Taurodontism (enlargement of the molar teeth by an extension of the pulp) is present in 40% of men with Klinefelter syndrome, compared with ~1% of men with an XY chromosome pattern.45 Varicose veins, with or without hypostatic ulceration, are seen in 40%; this, along with the 55% incidence of mitral valve prolapse, suggests a connective tissue defect.46,47 Abnormal glucose tolerance due to a postreceptor defect in insulin action is seen in 10% of patients with Klinefelter syndrome.48 The incidence of asthma and chronic bronchitis is also increased. Although osteoporosis is increased, the frequency is no greater than in other male hypogonadal states.49,50,51 and 52 As with most hypogonadal syndromes, sexual activity or sexual orientation is of concern to these patients and their family members or spouses. The patients with marked hypogonadism tend to associate less in adolescence with male peer groups because of their physical inability to compete. However, patients with Klinefelter syndrome do not have any greater homosexual tendencies than do men with an XY chromosome pattern. Usually, heterosexual activity is less frequent among patients with Klinefelter syndrome; many do not attain orgasm during sexual activity.53 VARIANT FORMS In certain patients, the sex chromosome configuration may show two or more stem cell lines (mosaicism). For example, in a testicular biopsy prepared for chromosome analysis, the predominant cell line may be XXY but another line of normal XY cells also may be present. Thus, on metaphase analysis, an XXY/XY cell pattern is found. This lack of a pure XXY cell line may result in a modification of the phenotypic characteristics of the patient such that the only clinical manifestation of classic Klinefelter syndrome is infertility (Table 115-3). On the other hand, if most cells are XXY, the patient may appear to have the classic Klinefelter syndrome, yet because of a normal XY cell line in some testicular germ cells, spermatogenesis and fertility may occur54 (see Fig. 115-5). Approximately 10% of patients with Klinefelter syndrome have a mosaic chromosomal constitution.

TABLE 115-3. Karyotype and Clinical Features of Classic and Variant Forms of Klinefelter Syndrome

The existence of mosaic forms may be suspected in men who are infertile with small testes or who have primary hypogonadism and normal buccal smears. In these patients, if more than one tissue specimen is examined for the presence of an XXY chromosome constitution (e.g., leukocytes and testicular tissues), and an extra X chromosome is found in one cell line but not the other, evidence for mosaicism exists. Importantly, the examination for chromatin in tissues often is poorly performed and may lead to erroneous conclusions. For example, many laboratories allow “normal” men to have an occasional X-chromatin body in a buccal smear. These are not true X-chromatin bodies but folds in cells (or bacteria). Therefore, when Klinefelter syndrome, especially the mosaic form, is suspected, confirmation should be made by a formal karyotyping (Fig. 115-6).

FIGURE 115-6. Karyotype with Giemsa banding from a peripheral lymphocyte of a patient with Klinefelter syndrome. Note the similar banding patterns of the duplicated X chromosome. (Courtesy of C. M. Disteche, Department of Pathology, University of Washington, Seattle, WA.)

Several additional variant forms of Klinefelter syndrome have been reported (see Table 115-3). These usually arise from consecutive nondisjunctional events in oogenesis or spermato-genesis. Individuals with the poly-X chromosomal syndromes (e.g., XXXY, XXXXY) display more severe abnormalities than do those with the classic Klinefelter syndrome55 (Fig. 115-7). A clue to these disorders is the presence of more than one chromatin body on buccal smear preparations. Often, marked mental retardation is present. The seminiferous tubules may have undergone hyalinization before puberty, and the testes commonly are undescended. Skeletal abnormalities, including proximal radioulnar synostosis and overgrowth of the radioulnar head, are characteristic of these disorders. As evidence of more severe androgen deficiency, the development of the genitalia may be retarded and a bifid scrotum or hypoplastic penis may be present.

FIGURE 115-7. Youth with the XXXXY syndrome. Epicanthal folds, a hypoplastic midface, and prognathism are seen. The neck is short, the penis is small, and the scrotum is hypoplastic. (From Goodman RM, Gorlin RJ. The face in genetic disorders. St. Louis: CV Mosby, 1977:722.)

The XXYY syndromes can combine the gonadal and associated features of Klinefelter syndrome with the skeletal height and social aggressiveness of the XYY syndrome.56 All patients in these categories may have social problems, which usually are related to the decreased IQ. Indeed, when these patients have normal IQs, they may be socially sensitive and model citizens. The specific skeletal feature is their mean height of 190 cm, which is 6 cm taller than the mean height of 184 cm of those with XXY Klinefelter syndrome. Both these mean heights are significantly greater than the mean height of normal men. This probably is due to the excessive growth of the lower extremities. THERAPY The treatment of patients with Klinefelter syndrome must address three major facets of the disease: hypogonadism, gynecomastia, and psychosocial problems. Androgen therapy is the most important aspect of treatment. Androgens are necessary to prevent the physical consequences of androgen deficiency, such as diminished libido and physical endurance, decreased muscle strength, and osteoporosis.57,58,59,60 and 61 Maintaining or increasing muscle strength and improving libido usually result in an enhanced self-image and an improved ability to cope with life (see Fig. 115-3). Treatment with testosterone should begin at puberty or as soon after puberty as the diagnosis is made. Androgen therapy should not be used in cases of severe mental retardation in which management of a more aggressive individual could be a problem. The presence of breast cancer or prostatic carcinoma demands special consideration before androgen therapy is maintained or initiated. Patients and their sexual partners should be counseled before treatment with androgens is begun, because a sudden increase in libido may present the couple with adjustment problems. Because the infertility does not appear to be related directly to the decreased intratesticular androgen levels, androgen therapy does not improve infertility, and because of the suppression of gonadotropins, this therapy could further diminish any spermatogenesis that is taking place. Specific information on androgen replacement therapy is found in Chapter 119. The gynecomastia of Klinefelter syndrome may be disfiguring and can cause significant social problems. Treatment with testosterone does not significantly diminish or worsen the condition, and no medical therapy is available. Therefore, if the breast enlargement concerns the patient, cosmetic surgery is required. Correction of the gynecomastia should help improve the patient's self-image. No long-term prospective studies are available regarding therapy for the learning disabilities and psychological problems associated with Klinefelter syndrome.62,63 With the increased use of amniocentesis, more cases with this disease are being identified, and the development of prepubertal support programs should be possible. Screening for earlier diagnosis may be considered in certain groups of young boys, such as those with learning disabilities who have testes smaller than 1.5 mL. Most patients have only mild problems in their verbal IQ, and early educational support may help prevent frustration and isolation. XX MALES Patients with XX chromosomal configurations may appear as normal females, females with gonadal dysgenesis, true hermaphrodites, or males with gonadal dysgenesis. Only those manifesting as phenotypic males are discussed here. These men proceed through puberty and subsequently present with small testes, infertility, and gynecomastia. As in patients with Klinefelter syndrome, serum testosterone levels may be low to normal, but serum gonadotropin levels are invariably elevated.63,64 These patients tend to be shorter than normal men. A high incidence of hypospadias is found, but no mental retardation. The abnormal skeletal proportions and many of the other associated features of Klinefelter syndrome are absent. The incidence of the syndrome ranges from 1 in 9000 to 1 in 20,000 live

births. In approximately two-thirds of cases, XX males result from the translocation of the SRY gene to the X chromosome. In the remaining cases, no SRY has been detected. In the latter individuals, a mutation is thought to have occurred in an autosome that triggered the same series of events as SRY, leading to testicular development. Evidence for this is found in the fact that SRY-negative men with the XX chromosome pattern often have other congenital abnormalities, especially cardiac problems. These abnormalities suggest autosomal mutations close to an SRY-like region of an autosome, although this region has not been identified.65 The treatment of XX males includes surgery for gynecomastia, androgen replacement therapy, and psychological support, as described for Klinefelter syndrome. XY/XO MIXED GONADAL DYSGENESIS Patients with XY/XO mixed gonadal dysgenesis who have the 45,XO/46,XY genotype may appear as phenotypic males, although most patients with this chromosomal constitution are phenotypic females.66 They have been considered to be H-Y antigen positive. Before the report of the translocation in XX males, the suggestion was made that, because such men were H-Y antigen positive but had gonadal dysgenesis, they must have lost the Y chromosome from the cell in the zygote. This issue is now somewhat clouded, and these patients need further study. Usually, the gonads are located within the abdomen. Both testes may be defective, or one may be a streak gonad. Depending on the gestational timing of the arrested gonadal development of the ipsilateral gonad, a paramesonephric duct may be present. In addition, a rudimentary uterus may be present. At birth, the external genitalia may range from female appearing, with clitoral enlargement, to male appearing, with some degree of hypospadias. These patients usually have been raised during their prepubertal years as females; they are discovered at the time of puberty, when primary amenorrhea is noted, or when, because of the increased pubertal stimulation of the defective testis and subsequent androgen production, marked virilization occurs. Treatment consists of supporting the sex of rearing with appropriate hormone replacement and castration. Castration is done to prevent the development of a malignancy in the defective gonad (20% incidence) and, if the individual has been raised as a female, to prevent the virilization that may occur after puberty. Fertility is extremely rare and should not be considered in the decision to remove an intraabdominal testis. The tumors that develop may be either dysgerminomas, gonado-blastomas, or embryonal cell carcinomas (see Chap. 122). XYY SYNDROME Individuals with the XYY syndrome have erroneously been called “supermales”; this is a misnomer, because testicular function may be normal or associated with varying degrees of impaired spermatogenesis.67,68 and 69 These individuals, with a mean height of 189 cm, are markedly taller than normal men and are more prone to antisocial behavior. Serum testosterone levels may be normal or elevated. Serum LH and FSH levels are normal unless spermatogenesis is markedly impaired, in which case FSH levels are elevated. Findings in testicular biopsy specimens have ranged from normal to markedly impaired spermatogenesis with seminiferous tubules that have undergone hyalinization. The sex chromosomal abnormalities arise from meiotic nondisjunction in the male or from zygotic nondisjunction. The impaired spermatogenesis alone, therefore, may be the result of diploid YY or XY spermatogonia, as has been described for Klinefelter syndrome. No specific therapy exists for these patients unless they have decreased testosterone levels. ULLRICH-NOONAN SYNDROME Ullrich-Noonan syndrome (Noonan syndrome) often has been referred to as “male Turner syndrome.”70 This is because of the phenotypic characteristics commonly shared with women who have Turner syndrome: webbed neck, low hairline, short stature, shield chest, and cubitus valgus (see Chap. 90).71 However, these men have a normal chromosome pattern. In Noonan syndrome, commonly encountered cardiac abnormalities include pulmonary artery stenosis and atrial septal defects. Although testicular function may be normal, primary gonadal failure usually is present. Infertility, if present, generally is associated with cryptorchidism. Serum levels of testosterone, LH, and FSH may be normal or consistent with primary testicular failure (i.e., decreased serum testosterone and increased LH and FSH). Mental retardation, ptosis, hypertelorism, and low-set ears also are prominent in this disorder. Rarely, autosomal dominant transmission may occur. Treatment is directed toward any cardiac abnormalities; if androgen deficiency exists, testosterone replacement therapy should be instituted. MYOTONIC DYSTROPHY Myotonic dystrophy is an autosomal dominant disorder characterized by an inability to relax the striated muscles after contraction (myotonia). The disorder results in muscle atrophy and eventual death. In addition to the myotonia, frontal balding, lenticular opacities, and primary testicular failure are commonly associated abnormalities72 (Fig. 115-8). Because the disease usually is not manifest until after puberty, most commonly in the late 30s and mid-40s, these men do not have an eunuchoid body habitus. The most prominent physical manifestation of their gonadal failure is the small testes and sometimes the other signs of postpubertal gonadal failure. Many cases demonstrate low serum levels of testosterone and azoospermia. In such patients, testicular biopsy specimens reveal complete hyalinization of the seminiferous tubules. Although serum testosterone levels usually are decreased, LH is increased in only 50% of the patients.73 Occasionally, only seminiferous tubule failure is present, which is characterized by a monotropic increase in serum FSH with normal testosterone levels.

FIGURE 115-8. Patient with myotonic dystrophy (myotonia dystrophica, Steinert disease). Weakness of the muscles, wasting of the limbs, and slumped posture are present. Note the expressionless facies, the frontal balding, and the marked atrophy of the temporal and masseter musculature (arrows).

The pathogenesis of the testicular failure is unknown; however, the (CTG)n amplification in the myodystrophy (MD) and MT-PK gene mutations varies from 70 to 1520. The length of the repeat correlated significantly (p 30 rad, and the return of serum FSH levels to normal paralleled the return to normal spermatogenesis. No significant change in the levels of serum LH or testosterone occurred at any of these doses.

FIGURE 115-14. Testicular biopsy before (A) and 228 days after exposure to 100 rad of x-ray irradiation (B). Note the loss of germ cells except for a few type A spermatogonia.

FIGURE 115-15. The effects of various doses of testicular x-ray irradiation on type A spermatogonia over time. Lines represent least squares regression over data points for each dose group. The return of spermatogenesis depends on the degree of loss of spermatogonia. (R, rads.) (Modified from Clifton DK, Bremner WJ. The effect of testicular x-irradiation on spermatogenesis in man. J Androl 1983; 4:387.)

If a patient is scheduled to receive therapeutic irradiation, suppression of spermatogenesis with testosterone or testosterone and an LHRH agonist or antagonist has been suggested to protect against infertility, because dividing cells are more sensitive to the effects of irradiation. This preventive treatment has not been tested. The only acceptable methods of preventing the damage are testicular shielding during irradiation or sperm banking before treatment. Consideration must be given to the increased numbers of chromosomal breaks that will occur during the irradiation, which may lead to an increase in fetal anomalies. Leydig cell dysfunction is not seen until doses of >800 rad are administered. However, these are common doses of therapeutic irradiation.124 In studies performed on men who had received therapeutic irradiation for prostate cancer without testicular shielding (the usual procedure), a significant decrease in basal serum testosterone levels, compared with age-matched men, was noted. A single, large dose was more deleterious than were fractional doses of similar total magnitude. Serum LH and FSH levels were found to be significantly increased, which indicates primary testicular damage. Treatment of the testosterone deficiency should be considered, because this should improve the anabolic state and enable patients to tolerate chemotherapy. However, in the case of prostate carcinoma, an androgen-responsive tumor, replacement therapy should not be given. Radioiodine therapy for thyroid cancer also has been associated with testicular dysfunction, primarily a loss of spermatogenesis.125 Because comprehensive studies have not been done, however, the frequency of significant damage is unknown (see Chap. 226). AUTOIMMUNE TESTICULAR FAILURE Two types of autoimmune testicular failure have been described. The most common of these is infertility resulting from the production of antibodies to sperm.126 This situation has been described most convincingly after vasovasostomy in men who have had a vasectomy. The antibody reaction may be responsible for the persistence of infertility in spite of normal sperm counts after vasovasostomy. The presence of antisperm antibodies of the immunoglobulin A (IgA) class in semen and of the immunoglobulin G (IgG) class in the peripheral blood may be a cause of idiopathic infertility. The mechanism for the appearance of these antibodies in infertile men without histories of testicular damage or genital infections is unknown. Because of the high incidence of anti-sperm antibodies in the normal fertile population, however, this hypothesis may not be valid (see Chap. 118). Treatment of this form of infertility with high-dose glucocorticoids has been attempted. The second, and less common, form of autoimmune testicular failure is that which occurs with steroid cell antibodies and subsequent loss of testosterone production.127 This type of steroid cell antibody has been recognized most commonly in women with ovarian failure but is found in men in rare cases. Usually, this condition is seen with other endocrine gland autoimmune diseases, especially Addison disease. The antibody is directed against the microsomal portion of the cell and not against the steroid hormone. Thyroid disease (e.g., Graves disease, hypothyroidism, or Hashimoto thyroiditis); hypopara-thyroidism; pernicious anemia; diabetes mellitus; and alopecia totalis also have been associated with Leydig cell failure due to steroid cell antibodies (see Chap. 197). This form of autoimmune disease has been linked with the HLA-B8 major histo-compatibility locus on chromosome 6. Because of the small number of men described with autoimmune Leydig cell failure, specific types of HLA associations are unknown. Androgen replacement treatment is required for these patients. CHEMOTHERAPY EFFECTS Similar to radiation exposure, chemotherapy also may damage the testes (see Chap. 226). Alkylating agents, such as nitrogen mustards, cyclophosphamide, and chlorambucil, consistently harm spermatogenesis in a dose-related fashion.128,129 Other agents, such as procarbazine and various combinations of drugs, also affect spermatogenesis in most individuals.130 The appearance of gynecomastia provides a clue that Leydig cell function may be affected.131,132 The same recommendations for prevention of damage described in the section on irradiation also apply to the use of chemotherapeutic agents. Sperm banking may be considered in young men who undergo curative chemotherapy for malignancies such as Hodgkin disease or germ cell carcinoma. In some of these tumors, especially Hodgkin disease, testicular function may be affected by the disease itself. The effects of chemotherapy on Leydig cell function have not been well described; however, the association of gynecomastia with chemotherapy suggests a defect in testosterone production.

SECONDARY HYPOGONADISM HYPOGONADOTROPIC HYPOGONADISM Hypogonadotropic hypogonadism may be caused by acquired or congenital defects. The presentation of the acquired forms of the disorder varies depending on whether the individual has gone through puberty and whether other anterior or posterior pituitary hormone deficiencies are involved. If the individual acquires the disorder before the onset of puberty, the presentation is that of a prepubertal male, as noted in Table 115-2. If puberty has occurred before the onset of the disorder, the presentation is with signs and symptoms of postpubertal testicular failure. The acquired forms of the disorder and some congenital forms have additional anterior pituitary deficiencies. If growth hormone is deficient, it is of clinical significance only if the individual is prepubertal and has not reached his adult height. After puberty, when the epiphyses of the long bones have fused and linear growth has ceased, growth hormone deficiency will no longer be apparent. CLASSIC HYPOGONADOTROPIC EUNUCHOIDISM Just as Klinefelter syndrome has become the prime example of hypergonadotropic hypogonadism, congenital hypogonadotropic eunuchoidism is the classic example of hypogonadotropic hypogonadism. As originally described by Kallmann and colleagues, this syndrome (Kallmann syndrome) is characterized by isolated gonadotropin deficiency and anosmia or hyposmia due to defective development of the olfactory bulbs.132,133 Other findings occasionally described with this disorder include midline cleft palate and lip, congenital deafness, cerebellar seizures, a short fourth metacarpal, and cardiac abnormalities.134,135 Kallmann syndrome is most commonly associated with autosomal dominant inheritance.134 The associated defects, especially anosmia, have enabled tracing of father-to-son transmission.136 However, kindreds have been reported that suggest an autosomal recessive or X-linked form of inheritance.137 The X-linked form of Kallmann syndrome is due to point mutations and exon deletions of the KAL gene located on the X chromosome at location Xp22.3.138,139 and 140 The gene consists of 14 exons and produces a 680-amino-acid, 76-kDa protein (anosmin-1) that can be glycosylated to form an 85-kDa protein product. The protein contains four fibronectin type III repeats associated with cellular adhesion functions, and a four-disulfide core motif that is commonly associated with antiprotease activity. The question of other forms of inheritance will be further defined as more men and women with the syndrome become fertile as a result of newer modes of therapy, and their progeny are studied. The incidence of this syndrome is ~1 in 10,000 male births.134 These men present as prepubertal eunuchs. They usually have no evidence of pubertal physical findings and their skeletal proportions are eunuchoid, with the ratio of upper body segment (pubis to vertex) to lower body segment (pubis to floor) decreased; the arm span is at least 6 cm greater than the height. These body proportions result from a failure of epiphyseal fusion and continued long bone growth. Beard and pubic hair growth are absent or minimal. Most patients retain their prepubertal subcutaneous body fat (Fig. 115-16). Hyposmia is present in most cases but may be missed unless specifically tested using appropriate olfactory sensation materials141 (Fig. 115-17). Muscle mass and strength remain at prepubertal levels. Although some CNS problems have been described,142 mental retardation is not one of them. A small phallus and temporal facial wrinkling caused by hypogonadism are common in this disorder. Gynecomastia may occur; peripheral aromatization of adrenal androgens may be contributory. Testicular development remains at prepubertal levels, with a few patients showing some testicular enlargement at puberty.143 However, unlike in Klinefelter syndrome, the testes are of normal prepubertal size because no tubular scarring or loss of germ cells is present. The Leydig cells are immature (as would be seen without LH stimulation), although normal numbers of interstitial cells are present, and the development into mature testosterone-producing Leydig cells occurs with gonadotropin stimulation (Fig. 115-18).

FIGURE 115-16. A, Patient with hypogonadotropic hypogonadism (Kallmann syndrome). Note the long legs and the increased arm span. Body hair is sparse, and the penis is small (B). C, Note the increase in penile size and pubic hair after 6 months of therapy with testosterone.

FIGURE 115-17. Olfactory thresholds for various molar concentrations of aromatic solutions in patients with hypogonadotropic hypogonadism (Kallmann syndrome). Transverse lines indicate the olfactory thresholds for normal adults. This test is performed by making separate molar solutions of the three substances indicated. The subjects are blinded to the contents of the containers and, after they are presented in random order, are asked to indicate whether a solution has an identifiable odor. Solutions should be made fresh weekly and the consistency of the investigator's threshold should be noted to determine loss of aroma of the test solutions. The molar concentration at which the test subject can no longer consistently detect an aroma is noted. This threshold then is compared with that of a group of normal controls and to the threshold of the normal investigator. Patients with anosmia consistently fall below the normal threshold, but not all patients with hypogonadotropic hypogonadism have anosmia. (From Santen RJ, Paulsen CA. Hypogonadotropic eunuchoidism. I. Clinical study of the mode of inheritance. J Clin Endocrinol Metab 1973; 36:47.)

FIGURE 115-18. Initiation of spermatogenesis with administration of human chorionic gonadotropin (hCG) and human menopausal gonadotropin (hMG) in a patient with hypogonadotropic hypogonadism as demonstrated by serial testicular biopsy specimens. A, Initial state. B, After 6 months of therapy with hCG plus hMG. (From Paulsen CA. The testis. In: Williams RH, ed. Textbook of endocrinology. Philadelphia: WB Saunders, 1981:333.)

Basal serum gonadotropin levels are low normal or undetectable in these men. When multiple samples are taken over a 24-hour period, however, some subjects demonstrate occasional small pulses of LH.144 When a single bolus dose of gonadotropin-releasing hormone is administered, the response of the pituitary gonadotropins usually is minimal. However, if repeated pulses of gonadotropin-releasing hormone are given, a normal rise in serum LH and FSH levels eventually occurs.143,145,146 These studies indicate that the most likely defect is a deficiency in LHRH secretion by the hypothalamus. Furthermore, autopsy studies have demonstrated anatomically normal pituitary glands.147 Serum testosterone levels remain in the middle of the female range in most patients with this condition before treatment, although they may have shown partial progression through puberty. No other pituitary hormonal defect has been documented in these patients. Approximately 10% of men with idiopathic hypogonadotropic hypogonadism provide a history of some pubertal development with subsequent regression. This diagnosis should be suspected in any man who has passed the normal pubertal age and remains prepubertal.144 Serum gonadotropin levels are low, testosterone levels are low, and prolactin levels are normal.148,149 If one of the associated abnormalities, such as anosmia, cleft palate, and cleft lip, is present, the diagnosis can be made with some assurance. If these findings are absent, however, differentiating between delayed puberty and idiopathic hypogonadotropic eunuchoidism can be extremely difficult, because puberty does not occur until 18 or 19 years of age in some normal men. Several maneuvers have been described that purport to differentiate this condition from delayed puberty.147,150,151 These include a subnormal serum prolactin response to thyrotropin-releasing hormone or to a phenothiazine, a decreased serum testosterone response to exogenous hCG, and a normal LH response to pulsatile stimulation by LHRH. None of these tests has been consistently reliable. If the differentiation cannot be made and the circumstances warrant, treatment with testosterone or hCG for 3 or 6 months may be indicated. After this, therapy can be discontinued and the patient observed to see whether spontaneous puberty progresses. The treatment of idiopathic hypogonadotropic eunuchoidism has two goals: to provide adequate androgen replacement and to achieve fertility.151a Androgen replacement may be accomplished as indicated in Chapter 119. Androgen replacement therapy does not impair subsequent therapeutic stimulation of spermatogenesis. In contrast to most of the causes of primary testicular failure in which improvement of fertility is not possible, in secondary hypogonadism fertility may be initiated. Several methods are available. One method of restoring spermatogenesis involves the use of hCG and then FSH in the form of human menopausal gonadotropins (hMGs).152,153,154 and 155 Initially, hCG is administered at a dosage of 1000 to 2000 IU intramuscularly three times a week to stimulate Leydig cell function; this increases intratesticular testosterone levels. After the dosage of hCG that achieves normal serum testosterone levels has been determined and the patient has been treated for 8 to 12 weeks, hMG may be added (see Fig. 115-18). The initial regimen is 75 IU intramuscularly three times a week while continuing hCG therapy. Once spermatogenesis is initiated, the dosage of hMGs can be decreased to 12 to 25 IU three times weekly. In some men, especially those with an initial testicular volume >5 mL, spermatogenesis can be maintained without continuing hMG. A history of undescended testes or of multiple pituitary hormone defects in syndromes other than hypogonadotropic eunuchoidism may result in treatment failure. A second method, which takes advantage of the knowledge of the pulsatile release of LH and FSH in these patients, has been used to stimulate spermatogenesis and Leydig cell function.144,156,157 By administration of a subcutaneous pulse of LHRH every 90 to 120 minutes through an infusion pump worn by the patient, a normal pattern of gonadotropin pulsation can be induced and spermatogenesis subsequently occurs.144,145 The dose of LHRH given with each pulse depends on the number of pulses administered each day. Usually, a pulse is given every 90 minutes. Thus, the average dose per pulse is between 1.5 and 10 ng/kg, assuming an average of 16 pulses each day. The dosage generally is determined by what is needed to produce a normal serum testosterone level. Spermatogenesis can be expected to occur after 2 to 3 months. The administration of LHRH at intervals faster or slower than 90 to 120 minutes produces inappropriate secretion of LH and FSH. The cost of pulsatile LHRH administration and hCG/hMG therapy is similar. A clear reason for one therapy or the other is not apparent. In addition to inducing fertility in these men, androgen replacement therapy is needed to develop muscle mass, prevent osteoporosis, provide psychological support, and produce full pubertal development. Any of the methods for inducing spermatogenesis also replace androgens by stimulating the Leydig cells. However, once fertility has been achieved, exogenous androgen therapy probably is the most satisfactory and economic choice for the patient. Then, if another pregnancy is desired, the patient can be switched back to gonadotropin or LHRH therapy. Idiopathic hypogonadotropic eunuchoidism may occur in women. The condition may be isolated or inherited as an autosomal dominant or autosomal recessive trait. It presents similarly in women and men, with a failure to proceed through puberty. These patients usually appear as normal prepubertal females and only rarely have had any signs of pubertal development. In addition, they may have other characteristic somatic features, including anosmia, short fourth metacarpals, and midline defects. Other syndromes manifest as isolated gonadotropin deficiency. Most of these are associated with severe neurologic damage and mental retardation (Table 115-5).

TABLE 115-5. Syndromes Associated with Congenital Gonadotropin Deficiency

ISOLATED DEFICIENCY OF LUTEINIZING HORMONE OR FOLLICLE-STIMULATING HORMONE The original reports of isolated LH deficiency occurred in eunuchoid men with testicular volumes, suggesting some development greater than that seen in patients with idiopathic hypogonadotropic hypogonadism.158,159 The term fertile eunuch was applied to these men, although most are infertile because the low intratesticular concentrations of testosterone do not support complete spermatogenesis.160 In these patients, testicular biopsy specimens reveal some progression of spermatogenesis. The serum testosterone levels are also low, accounting for the eunuchoid body habitus (Fig. 115-19). Therapy with hCG restores normal serum testosterone levels and completes spermatogenesis.161 Because data have shown that some men with classic idiopathic hypogonadotropic hypogonadism and low levels of both serum LH and FSH may have testicular volumes >5 mL and may demonstrate complete spermatogenesis with hCG alone, these syndromes most likely are a part of the spectrum of the same disease. A group of older men who manifest impotence, low serum testosterone levels, low LH levels, and normal FSH levels has been described (see Chap. 114). These men have gone through normal puberty and fathered children, and whether their LH deficiency is acquired or is a late expression of a congenital disorder has not been determined.

FIGURE 115-19. Patient with isolated deficiency of luteinizing hormone (“fertile eunuch”). Note the eunuchoidal body proportions but pubertal development and partial masculinization. Serum testosterone levels ranged from 1 to 2 ng/mL, which are well above the normal female range and those seen in hypogonadotropic hypogonadism, but below those of a normal man. (From Killinger DW. The testis. In: Ezrin C, Godden JO, Volpé R, eds. Systematic endocrinology, 2nd ed.

Hagerstown, MD: Harper & Row, 1979:232.)

Isolated deficiency of FSH also has been reported.162,163 Men with this disorder are seen for infertility and are normally androgenized. They are not eunuchoid in appearance, have normal serum testosterone levels, and have sperm counts ranging from azoospermia to oligospermia. In addition to syndromes of isolated FSH deficiency, male hypogonadism has been reported to be due to mutations in the b subunit of FSH that result in an inactive protein.164 ACQUIRED FORMS OF GONADOTROPIN DEFICIENCY Gonadotropin deficiency may result from numerous acquired disorders, as noted in Table 115-6. These lesions had been thought to be caused by the loss of gonadotropin secretion by compression and necrosis of pituitary tissue. This especially appears to be true in the case of tumors or granulomatous disease. However, some pituitary lesions can cause hypogonadism without anatomically affecting the pituitary tissue. For instance, tumors that produce corticotropin-releasing hormone, cortisol, or prolactin may directly inhibit gonadotropin secretion. Clinically, the appearance of these patients depends on whether or not they have gone through puberty (see Table 115-2).

TABLE 115-6. Causes of Acquired Gonadotropin Deficiency

Patients who have space-occupying lesions of the sella may lose secretion of other pituitary hormones and have symptoms of multiple hormone deficiencies (i.e., hypothyroidism due to thyroid-stimulating hormone deficiency, hypoadrenalism due to adrenocorticotropic hormone deficiency). Before puberty, manifestations of growth hormone deficiency may be present. However, clinical growth hormone deficiency is not seen after puberty because maximum growth has already occurred. Some of the causes of acquired gonadotropin deficiency are listed in Table 115-5. PROLACTIN-SECRETING PITUITARY TUMORS Prolactin-producing tumors of the pituitary have been divided into microadenomas (10 mm in diameter) (see Chap. 13). In women, >80% of the tumors are microadenomas; in men, 80% are macroadenomas.165,166 This difference may result from earlier detection in women (due to the occurrence of galactorrhea or of irregular menstrual cycles) or from a difference between the two sexes in the character of tumor growth. Because galactorrhea occurs so rarely in men, even in men with high serum prolactin levels, early clinical signs of the tumor are not available. Therefore, the tumor may not be manifest until symptoms caused by the mass of the lesion are present.167 In some men, temporal lobe epilepsy has been associated with hyperprolactinemia.168 The mechanism by which prolactin-producing pituitary adenomas cause hypogonadism has not been well defined. Prolactin had been thought to lower basal serum LH and FSH levels, but this has been an inconsistent finding. A change in LH pulse amplitude or frequency has occurred in some men.169 Another possibility is that a single neurotransmitter, such as dopamine (see Chap. 16) or g-aminobutyrate, may be involved. The latter may both stimulate serum prolactin and suppress serum LH, in which case the prolactin itself may have little direct effect on gonadal function.170 The diagnosis of a prolactinoma requires the finding of an elevated serum prolactin level. Most men with demonstrable prolactinomas have serum prolactin levels of >50 ng/mL. Computerized tomography of the sella is the current standard for demonstrating the tumor mass and should be performed in men with elevated serum prolactin levels. Although magnetic resonance imaging (MRI) of the sella may be useful, instruments of 160/95), a 2.1-fold greater likelihood of hypercholesterolemia (>250 mg/dL), and a 3.8-fold greater likelihood of diabetes than the non-overweight sample. Admittedly, it is an arbitrary judgment, but at this stage of understanding, the Expert Panel on the Identification, Evaluation and Treatment of Overweight and Obesity in Adults convened by the NIH, in concordance with the World Health Organization, defines overweight as a BMI between 25 and 29.9 and obesity as a BMI >30.2 While return to a BMI in the normal range is probably optimal for an individual losing weight, evidence suggests that weight loss of a lesser degree, in some cases as little as 10% to 20% of total body weight, may suffice to lessen the morbidity associated with diabetes6 and hypertension.7 It has been suggested that this finding is due to a disproportionate loss of visceral fat with small amounts of weight loss. Thus, in any evaluation of the hazards of obesity, the distribution of adipose tissue can be an important consideration in addition to the BMI. DISTRIBUTION OF FAT An abdominal or “potbelly” distribution in either sex must elicit more concern than fat in other distributions. Central distribution of adipose tissue, characterized by an increase in the ratio of the waist to hip (W/H) circumference, or in the abdominal sagittal diameter, has been associated with an increased risk of medical complications, including angina pectoris, stroke, and type 2 diabetes.8,9 and 10 As a result, it has been suggested that waist circumference should be part of the evaluation of the overweight and obese individual. Between a BMI of 25 and 35, a waist circumference >102 cm (40 inches) in men and 88 cm (35 inches) in women represents a significant and probably independent risk to health (Table 126-1). The possible reasons for this unusual association of regional adiposity with diabetes, hypertension,

and plasma lipoproteins are dealt with in the Obesity and Diabetes Mellitus section.

TABLE 126-1. Classification of Overweight and Obesity by Body Mass Index, Waist Circumference, and Associated Disease Risk*

The list of health hazards of obesity unfortunately is not restricted to those risk factors already enumerated but also includes hiatal hernia, gastroesophageal reflux, pickwickian syndrome, and gout. The specific contribution of obesity to coronary artery disease and to reduction in longevity has been more difficult to demonstrate because it is mediated via its impact on other risk factors such as dyslipidemia, type 2 diabetes, and hypertension, but is now clear enough for the American Heart Association to have declared obesity a “major, modifiable risk factor” for coronary artery disease. Finally, the obvious adverse psychosocial consequences of obesity in society are by no means trivial.

PATHOGENESIS THERMODYNAMIC CONSIDERATIONS It has become customary to remind all who deal with obesity that it is a disease of energy storage and as such must obey the laws of thermodynamics. The first law mandates that the amount of energy in any system cannot vary unless the inflow or egress of energy from the system becomes unequal. In mathematical terms, DE = Q – W. For our purposes this formula translates into the following: Any change in adipose tissue mass (E) cannot come about unless there is a change of Q (food intake) or W (caloric expenditure). A brief perspective on the history of thermodynamics may be relevant.11 Engineers, mechanics, and scientists before the nineteenth century earnestly labored to develop a perpetual motion machine. In fact, the history of science is strewn with the wrecks of plans and models for such machines. The second law of thermodynamics, enunciated in the nineteenth century, ordained that entropy rises with the operation of these machines, and, thus, the machines must stop unless more energy is supplied; however, this did not curtail efforts to create engines that might subdue entropy. The history of obesity research and the first law of thermodynamics are similar.12 At the end of the eighteenth century, Lavoisier had shown that respiration and oxidation are not different from combustion. There is no vital heat production in living organisms, different from the production of heat by the burning of fuels in ovens or bonfires. Nevertheless, the search for ways to extract more energy from ingested food or burn away more energy than can be explained by simple chemical considerations remains active. Periodically, new diets allege to “burn away fat” or to enable one to “eat all you want without weight gain.” Consider for a moment the formula, DE = Q – W. It should be evident that, over long periods of time, DE = 0. This means that weight is generally stable whether one is obese or lean. With weight stability, however, passage of calories through the system is high; during an average lifetime one can expect that >50 million calories will come in as Q and disappear as W. Hence, very small alterations in the system can make for great obesity or remarkable leanness. The most interesting phenomenon from the standpoint of energy storage is the reason why Q and W are so often precisely equal. Most clinicians have chosen to focus on the possibility that some individuals eat a lot and remain thin whereas others eat very little yet become obese. These observations that appear to violate thermodynamic principles have carried the name endogenous obesity. Other observers have focused on the fact that obese individuals are driven by remarkable psychological or hedonic drives and supposedly do this in secret. This would be considered exogenous obesity. Classic studies nearly 70 years ago demonstrated clearly that all individuals, obese or not, respond to diets in lawful ways.13 One might have hoped that bizarre notions about diets and obesity would have vanished. However, perusals of studies of nutrition and obesity done in recent years still level accusing fingers at either the sedentary nature of our lives or our propensity to overeat as though in some way these two behaviors can become completely disarticulated from each other. To be sure, the presence of obesity means that at some point DE rose above 0. It is not often realized, however, that such changes in DE must have been quite small since the acquisition of obesity is measured in months or, more frequently, many years. The extra bites of food eaten each day or the slight diminutions in physical activity that create obesity have been difficult to measure. When precise measures of total 24-hour energy expenditure were made in obese and nonobese individuals in a hospital setting, utilizing various techniques for measuring energy expenditure, it was found that differences between the obese and nonobese, when expressed per kilogram fat-free mass, were small.14 For example, the obese expended 51 ± 7 kcal/kg fat-free mass, as compared to 47 ± 7 kcal/kg fat-free mass in the nonobese. Since these studies were done in a circumstance in which weight was kept stable by giving a sufficient amount of food intake to maintain stability, clearly Q = W; therefore, the food intake of the obese is only very slightly larger than that of the nonobese and is matched by the slight increase in energy expenditure. This increased caloric exchange of the obese may well be the necessary additional energy required for carrying a larger fat mass or the need for more cardiopulmonary activity than in the lean. The previous data do not speak to any deficiency of caloric expenditure in the obese. One theory, which has been prominent since early in this century, is that all individuals have “luxus konsumption,” which permits them to overeat and burn more calories in response to overeating, as a control against becoming obese. The fault of obesity was believed by some to be found in an inability to exercise this special control mechanism with precision. Extrapolating from brown adipose tissue, which in some animals and in the premature human infant create nonshivering thermogenesis by excess caloric burning, it has been thought that the obese experience a dysfunction of brown adipose tissue or a deficiency in amount. Yet, the data clearly show no deficiency in energy production in the obese, but rather a slight increase. Since the obese usually maintain obese weight at a constant level, energy outgo equals intake and, thus, the data do not implicate a sedentary nature of the obese as the cause of this unfortunate state. What may, however, be observed is a high intake of calories in the obese, when they have dieted and lost weight and are “breaking away” from the diet and regaining. Such hyperphagia can easily be misinterpreted as a natural state of hyperphagia in the obese. The more usual slight differences in food intake of the obese required to maintain the additional caloric needs of their obesity do not suggest that they have unusual hedonic or psychological needs that have to be met by large amounts of food intake. With precise measures of energy homeostasis, the effect of artificial increases in body weight of 10% above “usual” or 10% below showed unanticipated changes in caloric expenditure.14 The increase in caloric expenditure with weight increase is disproportionately high for the small accompanying changes in lean body mass. A reverse situation is the decline in caloric need per unit of lean body mass when weight is lowered 10%. The most likely explanation for these findings is that usual body weight is arrived at by a complex of forces, which detect body fatness and equalize the rate of energy expenditure and food intake. When alterations are made in weight by increasing or decreasing the level of fat storage, the same homeostatic mechanisms act to restore body fat to its previous state. When DE = 0, however, there is a reduction in caloric expenditure (and intake) when weight is lowered and an increase in caloric expenditure (and intake) when weight is elevated. The previous energy changes with weight change are largely in nonresting energy expenditure. There are smaller changes in the thermic effect of food and resting metabolic rate when weight declines. These changes in the efficiency of caloric expenditure as weight is changed may be due to alterations in metabolism, substituting oxidative for glycolytic metabolism, or as yet unknown mechanisms whereby muscle metabolism can vary in the efficiency of caloric disposal. Whatever the mechanism, it has become clear that there is a complex system maintaining constant body weight, which functions well in the obese but at a higher level of fat storage than in the nonobese. After obese or nonobese is changed from “usual” weight, Q and W must change coordinately if DE remains 0. This can be considered as a setpoint mechanism, which maintains the lean at their “set” and the obese at a higher set. Efforts to change body weight are resisted by the system. Thus, no laws of thermodynamics are contravened in human obesity. However, the system involved in the maintenance of a constant level of energy (fat) storage in adults is complex. How the system adjusts during growth and development and how the adult “set” may be altered by events in infancy, childhood, or even adolescence remain important topics for investigation. OPERATION OF THE SYSTEM MAINTAINING ENERGY STORAGE

The above description of energy metabolism in humans implicates the existence of a coordinated and complex mechanism controlling body fat. There is no reason to believe that a system of such great importance (i.e., the storage of energy for the prevention of starvation) would be any less complex than systems for the control of other physiologic variables such as body sodium, osmotic pressure, or blood pressure.15 Short-term regulators involve cognitive functions as well as gastrointestinal signals. Endocrine signals, signals from adipose tissue, and possibly signals from the liver are involved in both short- and long-term controls of fat storage in adipose tissue. The hypothalamus is considered the central signal-processing center with afferents from the above and efferents, which utilize the endocrine and autonomic nervous systems for control of both energy intake and outgo. An interesting aspect of this system is the relative ease with which it can be bypassed in the short run by the overeating of particularly attractive foods, by starvation imposed by illness, or by impeded access to food. It is generally observed that overriding of the system in the short term is readjusted over time to the original level of fat storage by operation of the central control center. It is also worth noting that the complexity of the system makes it likely that there are redundant loops so that the administration of a novel diet or a new drug for the treatment of obesity may affect one loop and be effective in the short run, but is often overridden in time by operation of other loops. HYPOTHALAMIC OBESITY Damage to the ventromedial hypothalamus has been shown to produce obesity in experimental animals. Hypothalamic injury is a rare cause of human obesity, but it can occur, after surgical removal of craniopharyngiomas or other tumors impinging upon the hypothalamus. In this event there is a marked increase in appetite and the rapid accession of fat along with hyperinsulinemia, similar to what is found in the experimental production of ventromedial lesions in animals. More subtle hypothalamic injuries may also produce obesity. In experimental animals it has been shown that canine distemper virus can infect the central nervous system early in life and, with little residual evidence of neural damage, produce obesity at a later age. Whether viral infections can produce obesity in humans remains uncertain. HORMONAL OBESITY Other chapters in this textbook describe the effects of growth hormone, thyroid hormone, and adrenal steroids on general metabolism and fat storage, and it must be remembered that these are rare causes of obesity. The central obesity of hyper-adrenocorticism, with striae, hypertension, acne, and hirsutism, produces a recognizable syndrome of obesity. Subtle changes in the “set” for fat storage independent of specific endocrine abnormalities are more usual causes for obesity than any antecedent hormonal change. DRUGS A number of drugs can produce a weight increase. Most notable are some antidepressants. Other drugs that may lead to weight gain include progestational agents, cyproheptadine, and valproate. GENETIC OBESITY For many years it has been recognized that a number of Mendelian disorders are associated with marked obesity. The most commonly considered are the Prader-Willi and the Bardet-Biedl syndromes. The former is related to a deletion of 15q 11.2q-12 in the father of affected individuals or to uniparental disomy for maternal chromosome, and separate cases of the latter have been linked with markers on chromosomes 3, 11, and 15. There are many uncertainties as to precisely how these chromosomal aberrations lead to obesity. The small nuclear ribonucleoprotein-associated polypeptide N (SNRPN) gene has been implicated in the Prader-Willi syndrome. Clinical descriptions of other Mendelian disorders with obesity are numerous; one review listed no fewer than 24.1 It has become clear that “ordinary” human obesity occurring in the absence of known endocrine states or specific syndromes also shows a high degree of heritability. Furthermore, precise molecular lesions have been found in experimental rodent obesities that have been studied for many years by traditional physiologic and biochemical techniques.16 Genes responsible for rodent genetic obesities and their protein products are: Yellow Obesity. Yellow obesity occurs in the yellow mouse as the result of a dominant mutation in the agouti locus on mouse chromosome 2 that leads to widespread expression of a protein known as the agouti signaling protein. The relevant site in humans is 20q13, although examples of this obesity in humans have not been described. Ob. The ob mouse has been studied for many years, but it was not until recently that the nature of a specific recessive mutation on mouse chromosome 6 was found. A peptide termed leptin cannot be elaborated or secreted from adipose tissue. It is believed that leptin is sensed in the central nervous system to modulate the complex mechanism for the control of body fat. The homologous site for this gene in humans is 7q31.3. Although a few families have been found with leptin deficiency, most obese humans have a large increase in plasma leptin levels (proportionate to body fat storage) rather than a deficiency of leptin (see Chap. 186). Db. In this animal there are abnormalities of the central leptin receptor, leading to obesity. The genetic lesion is found on mouse chromosome 4. The locus for the production of leptin receptor in humans is 1p21–p31. Whether leptin receptor abnormalities are present in humans remains uncertain. Lesions downstream to the receptor may also be a factor in human obesity. Leptin detected by the leptin receptor affects the JAK/STAT system. Neuropeptide Y figures prominently in this pathway and is a current focus of research (see Chap. 125). Zucker Rat. The Zucker obese rat has a leptin-receptor lesion on chromosome 5 similar to that of the db mouse. Tubby. The tubby mouse obesity (tub) leads to maturity-onset obesity. It is due to a recessive mutation found on chromosome 7. The mechanism of action of the gene product remains uncertain. Fat. This is due to a recessive mutation on mouse chromosome 8 with a related site on human 4q32. It is due to a mutation in the carboxypeptidase E gene. A similar abnormality of prohormone convertase in humans has been described. An inability to cleave large prohormones leads to hyperproin-sulinemia, associated with obesity. The obesity may be related to the inability to cleave prohormones other than insulin. Each of the above genetic abnormalities in mice has been important in enriching the study of the control mechanisms for fat storage. As of this time, it appears unlikely that single mutations, as found in mice, will be frequent causes for human obesity. However, genomic scans in families with obesity or in special subgroups of obese individuals are uncovering molecular lesions that may be related to obesity. A linkage of obesity to a peroxisome proliferator-activated receptor (PPARg) gene has been uncovered17; likewise, the possibility that uncoupling proteins (UCPs) may be related to obesity is under active study.18 It was originally thought that these proteins existed only in brown adipose tissue and were of importance only for thermoregulation during hibernation and infancy. Currently, UCPs found in white adipose tissue and muscle are being evaluated. Thus, genetic studies have been more revealing for uncovering the details of the mechanisms controlling fat storage than in discovering major causes for human obesity. It is likely that human obesity comes about by virtue of early life adjustment of the “set” for fat storage, which involves many genes. Given the complex nature of the system responsible for the “set,” there is abundant opportunity for early environmental influences to fashion lifelong changes in fat storage.

ADIPOSE TISSUE GROWTH AND DEVELOPMENT EARLY GROWTH OF ADIPOSE TISSUE Normal growth of human adipose tissue is achieved through an increase in adipocyte size and number. Histologic studies, as well as studies of adipocyte development in tissue culture, strongly suggest that the precursor cell, or adipoblast, is a cell that is indistinguishable from the fibroblast. The replication of preadipocytes and their differentiation into mature adipocytes have been shown to be affected by epidermal growth factor, insulin, glucocorticoids, and PPARg.19 This gene encodes two proteins, PPARg1 and PPARg2. The latter is found in adipose tissue and is believed to play a role in adipocyte differentiation. Adipose cells first appear at approximately the 15th week of human gestation, continue to appear in large numbers until approximately 23 weeks of gestation, and then appear more slowly throughout the remainder of gestation. During the first 2 years of life, adipose tissue mass grows by an increase in both cell size and number. Cell number rises slowly from 2 years of age until close to the onset of puberty, and during adolescence, there is another sharp elevation in cell number that accounts for a second spurt in the growth of the adipose tissue mass. Thereafter, the size of adipose tissue of nonobese adults maintaining constant body weight remains stable, as do adipose cell size and number. The size of adipose cells within nonobese individuals can differ considerably from one fat depot to another. In a large group of nonobese healthy subjects, cell size in six separate fat depots ranged from 0.14 to 0.68 µg of lipid per cell, with a mean size of 0.41 µg lipid per cell.20 The average total cell number in this

nonobese population was 30 × 109. ADIPOSE CELL NUMBER AND SIZE Modest changes in fat cell size occur with small changes in body weight. Greater expansion of the adipose tissue can be achieved either by an alteration in adipose cell size (hypertrophic, normal cell-number obesity) or in both adipose cell number and size (hyperplastic-hypertrophic obesity). Hyperplastic obesity in humans usually has its onset in early life, often before the age of 20 years. Obesity of later onset usually is accompanied by adipose cellular enlargement and normal cell number, but there is evidence that adult rodents can, under some circumstances, markedly increase their cell number. When adult body weight exceeds 170% of ideal, a maximum cell size (1.0–1.2 µg lipid per cell) is reached, after which cell number and obesity are highly correlated.21 Both the age of onset and severity of obesity appear to contribute to adipose hyperplasia. Although hypercellularity most often is found in those with early-onset obesity, patients have been observed with increased cell number in whom obesity apparently developed during adult life. These individuals usually are massively obese. Weight loss and reduction in the adipose tissue mass of all adult obese patients and of obese children, regardless of the age of onset or the degree and duration of obesity, are accompanied by a change in adipose cell size alone. Cell number remains constant, even in the face of dramatic weight loss. Thus, adipose hypercellularity appears to inflict a permanent, irreversible abnormality on the patient suffering from hyperplastic obesity. It has been hypothesized for nearly 30 years that fat-cell size and number may play a role in the control of human energy metabolism. The study of substances secreted by adipocytes has led to the addition of a new chapter in this textbook on the endocrine adipocyte (see Chap. 186). The contributions of fat-cell size and number to the newly discussed endocrine functions of adipose tissue are yet to be elucidated.

ENDOCRINE AND METABOLIC CONSEQUENCES OF OBESITY Various alterations in endocrine and metabolic function can be observed in obese individuals, including disturbances of growth hormone, changes of thyroid function, adrenal hormone metabolism, and insulin secretion and insulin action and alterations in lipid metabolism that may lead to hyperlipidemia. These are thought to be secondary to, rather than a primary characteristic of, obesity. OBESITY AND GROWTH HORMONE Growth hormone secretion in response to hypoglycemia or arginine is decreased in obese compared with lean patients, but weight reduction restores normal responses and plasma levels. OBESITY AND THYROID FUNCTION Thyroid function is normal in most obese patients, but subtle changes in thyroid hormone metabolism may occur. Plasma triiodothyronine (T3) levels are increased and reverse triiodothyronine levels (rT3) are decreased in many obese patients, whereas plasma thyroxine (T4) and thyroid-stimulating hormone levels are normal (see Chap. 33). The plasma T3 and rT3 changes are reversed by weight loss and T3 may even become subnormal although thyroid-stimulating hormone (TSH) remains normal. Hypothyroidism is rarely a cause of obesity, but it should be considered in the differential diagnosis. The obesity caused by hypothyroidism usually is mild, and restoration of euthyroidism eliminates the problem. OBESITY AND ADRENAL FUNCTION Cortisol secretion rates are increased in obesity, but normal plasma cortisol concentration, normal diurnal variation in secretion, and normal response to dexamethasone suppression distinguish obesity-associated alterations in adrenal function from Cushing syndrome (see Chap. 75). Total 24-hour urinary 17-hydroxycorticosteroids generally are increased in obese patients but are normal when expressed per kilogram of body weight. Twenty-four–hour urinary 17-ketosteroids may be normal or slightly increased in obesity. Weight reduction normalizes the cortisol secretion rate and 24-hour urinary 17-hydroxysteroid and 17-ketosteroid levels, indicating that these changes are secondary to obesity. However, hyperadrenocorticism (Cushing syndrome) can cause obesity; thus, it always should be considered and ruled out in an obese person (see Chap. 74). When Cushing syndrome is the cause of obesity, it tends to be mild and completely reversible after definitive treatment. OBESITY AND SEX HORMONES The onset of menarche is earlier in obese than in nonobese girls, and obese women often suffer from a variety of menstrual cycle abnormalities, including hypermenorrhea, oligomenorrhea, amenorrhea, irregular or anovulatory cycles, infertility, and premature menopause. Obesity is often associated with polycystic ovarian disease, hyperandrogenism, and hirsutism (see Chap. 96 and Chap. 101). Although the mechanism of these disorders in obese women is unknown, the disorders could be secondary to alterations in hypothalamic-pituitary function, to abnormalities in the synthesis or secretion of estrogens or androgens from the ovaries or adrenals, or to alterations in the metabolism of sex steroids by non–steroid-producing peripheral tissues. Elevation of plasma luteinizing hormone, reduction of plasma follicle-stimulating hormone, and increased luteinizing hormone/follicle-stimulating hormone ratios have been reported in some obese women. Elevated circulating levels of estrogens and androgens also have been reported in some obese women. These hormones may be of either ovarian or adrenal origin. Alterations in sex steroid metabolism, specifically increased conversion of androstenedione of adrenal origin to estrone and its metabolites by non–steroid-producing peripheral tissues, also may play a role in some of the menstrual cycle abnormalities and in the increase in prevalence of breast cancer in obese women.22 Conversely, an increased conversion of estrogen to androgen, which has been postulated to occur, may be a factor in creating hirsutism or abnormal menstruation. Most of these changes are reversible by weight loss. In men with severe obesity, both serum total and free testosterone levels may be mildly decreased. Increased conversion of androgens to estrogens may result in increased serum estrogen levels. Usually, libido and potency remain normal. Amenorrhea of hypothalamic origin, such as is found in starvation, occurs frequently in obese women who lose a great deal of weight. This is often accompanied by low T3 levels, leukopenia, and cold sensitivity. This may persist for long periods of time after weight loss and cannot be safely repaired by hormone therapy. OBESITY AND DIABETES MELLITUS Obesity is a major risk factor for type 2 diabetes. On average, >70% of type 2 diabetics studied in various populations throughout the world are obese, with prevalence variation dependent upon the population studied. The alarming increase in the worldwide prevalence of type 2 diabetes, especially in recently industrialized developing countries, in minority groups, and in children, appears, in large part, to be attributable to overnutrition and overweight. Other lifestyle factors—such as decreased physical activity and a general increase in longevity—all contribute to the emergence of type 2 diabetes as a serious worldwide medical and public health problem. INTERACTION OF GENES AND ENVIRONMENT For obesity to induce overt type 2 diabetes in a given individual, a genetic susceptibility must exist. Multiple genes are likely to be involved including those inducing insulin resistance and defective pancreatic B-cell function. It is upon this genetic background that obesity produces its “diabetogenic” effect. ADIPOSE TISSUE AND INSULIN RESISTANCE There is a close relationship between increased adipose-cell size, and insulin resistance or plasma insulin levels in both rodents and humans.23 The relationship is stronger than that between total body fat mass and either insulin resistance or plasma insulin levels. With weight loss and reduction in fat-cell size, insulin sensitivity usually improves, and there is almost always a decline in plasma insulin levels. The mechanism by which an enlarged adipocyte causes insulin resistance, increases plasma insulin levels, and impairs glucose metabolism is not known. Several factors are postulated to contribute. For example, release of increased amounts of free fatty acids (FFA) from enlarged, insulin-resistant adipocytes results in increased plasma FFA levels, and flux of FFA to peripheral tissues, the liver, and pancreatic B cells. Increased oxidation of FFA decreases insulin-stimulated glucose uptake into skeletal muscle due to inhibition of glucose transport.24 This adds to already existing

insulin resistance in type 2 diabetes and can precipitate hyperglycemia. Decreased insulin action caused by increased FFA oxidation leads to an even greater rate of lipolysis in adipose tissue and further elevation of plasma FFA levels, creating a vicious cycle. In addition, increased flux of FFA to the liver increases hepatic gluconeogenesis and hepatic glucose output, adding to the already existing abnormalities in hepatic glucose metabolism in type 2 diabetes.25 Elevated levels of FFA also stimulate insulin secretion, placing a further burden on genetically compromised and stressed B-cell function in individuals susceptible to type 2 diabetes. Furthermore, elevated FFAs have been reported to impair early insulin secretion, a characteristic B-cell abnormality in type 2 diabetes.26 Other factors associated with the enlarged adipose depot of the obese have been postulated to contribute to the link between obesity and diabetes. Release of increased amounts of the peptide leptin from the expanded adipose depot may play some role. Plasma leptin levels are elevated in most obese patients, and an inverse relationship between plasma leptin levels and insulin sensitivity has been reported.27 Studies in laboratory animals and humans, however, report conflicting data.28 In both db/db and ob/ob mice, roughly equivalent degrees of carbohydrate intolerance can develop in the presence of obesity, but in the former, plasma leptin levels are increased while in the latter they are reduced. In obese humans, plasma leptin levels are almost always elevated, whether diabetes is present or not; thus, a role for plasma leptin in the obesity-diabetes connection remains to be established. The enlarged adipose depot also produces increased amounts of tumor necrosis factor-a (TNF-a), which has been shown to impair insulin action and is postulated to play a role in the link between obesity and diabetes.29 ROLE OF BODY-FAT DISTRIBUTION The “diabetogenic” effects of obesity depend not only on an increased total body adipose tissue mass or fat-cell size, but also on the distribution of the excess fat. Excessive accumulation of abdominal fat (central obesity, android or “apple shape” body type) is associated with a much greater likelihood of diabetes and other metabolic derangements than equal amounts of fat in the gluteal-femoral region (peripheral obesity, gynoid or “pear shape” body type).9 Increased values of the W/H ratios are associated with increased plasma insulin levels, decreased glucose disposal, increased hepatic glucose output, glucose intolerance, dyslipidemias, especially hypertriglyceridemia, increased small dense low-density lipoprotein (LDL) cholesterol particles, decreased high-density lipoprotein (HDL), atherosclerosis, hypertension, and type 2 diabetes.8 It is postulated that the delivery to the liver of high levels of FFA from the enlarged, lipolytically active visceral adipose mass causes hepatic insulin resistance and hepatic triglyceride and very low-density lipo-protein (VLDL) overproduction; increased FFA delivery to the B cell causes hyperinsulinemia, secondarily leading to peripheral insulin resistance. Insulin resistance, whether due to increased rates of FFA oxidation, hyperleptinemia, excess TNF-a, or other factors, provokes increased insulin secretion to overcome impaired insulin action and to maintain normal blood glucose levels. Chronic hyperinsulinemia leads to down-regulation of insulin receptors, further worsening insulin resistance. Chronic exposure to excess insulin levels is thought to be a major factor in the development of the metabolic syndrome described below. OVERNUTRITION, EXCESSIVE CALORIC INTAKE, AND DECREASED PHYSICAL ACTIVITY Increased adiposity alone may not fully account for the “diabetogenic” effect of obesity. Chronic overeating—stimulating insulin secretion from defective pancreatic B cells in individuals who are genetically susceptible to type 2 diabetes—could produce hyperinsulinemia, which down-regulates insulin receptors, thereby contributing to insulin resistance. Chronic demand for increased insulin secretion from genetically defective B cells may lead to B-cell decompensation. Hyperinsulinemia triggers counterregulatory glucagon, catecholamine, and growth hormone responses known to be hyperglycemic. For these reasons, caloric restriction alone, independent of weight loss, significantly improves glycemic control. Decreased physical activity, characteristic of sedentary lifestyles in economically developed and “modernized” societies, is also thought to play an important role in the link between diabetes and obesity. OTHER POTENTIAL “DIABETOGENIC” EFFECTS OF OBESITY Hypothalamic controls mediated by the activity of various neuropeptides, hormones, and the autonomic nervous system, acting in concert to maintain a given level of adipose tissue and normal blood glucose concentration, may be deranged and may contribute to the relationship of obesity to diabetes.15 Adipose tissue UCP—reported to play an important role in regulating energy homeostasis and, therefore, in intermediary carbohydrate and lipid metabolism—has been implicated in the coexistence of obesity and diabetes.18 It should be noted that activation of PPARg by the thiazolidinedione class of drugs increases insulin sensitivity and decreases insulin resistance in type 2 diabetics. Finally, membrane glycoprotein PC-1 is expressed in tissues with insulin resistance, and PC-1 expression in skeletal muscle and adipose tissue has been reported to correlate inversely with insulin sensitivity.30 METABOLIC SYNDROME Abdominal obesity is associated not only with an increased risk of type 2 diabetes, but also with a number of other disorders with serious medical consequences, including atherosclerosis, dyslipidemia (most commonly hypertriglyceridemia, decreased HDL, increased small dense LDL, and increased FFA), increased plasminogen activator inhibitor and fibrinogen levels, hypertension, and vascular smooth muscle cell proliferation. This complex of disorders has been variously referred to as syndrome X, Reaven syndrome, metabolic syndrome, or the New World syndrome. This combination of disorders poses a major public health problem (see Chap. 145). CONTROL OF OVERNUTRITION AND OBESITY It seems highly likely that avoidance of overnutrition, overweight, or obesity, or successful long-term weight reduction in obese individuals, will significantly reduce the incidence and prevalence of type 2 diabetes and associated disorders. Indeed, a diet and/or exercise intervention in a population with impaired glucose tolerance (IGT) has been found to decrease significantly the incidence of type 2 diabetes over a 6-year period.31 OBESITY AND LIPIDS Obesity is typically associated with hypertriglyceridemia, decreased HDL cholesterol, and, to a lesser extent, hypercholesterolemia, abnormalities considered to be important risk factors for coronary heart disease (see Chap. 162 and Chap. 163). Elevation of plasma triglycerides very likely arises from increased influx of nutrient substrate (glucose, glycerol, FFA) into the liver of obese patients and the consequent production of triglyceriderich VLDL at a rate greater than that of peripheral tissue utilization. Weight reduction often restores plasma triglycerides to normal or near-normal levels. There is an inverse relationship between HDL cholesterol and BMI in obese patients, especially if they also are diabetic. Low HDL, which constitutes a risk for atherosclerosis, is also negatively correlated with high W/H ratios. After weight loss, plasma HDL levels transiently decrease and then usually rise above pre–weight-reduction levels. The relationship between obesity and hypercholesterolemia is less clear-cut than is the relationship between obesity and hypertriglyceridemia and HDL. Total body cholesterol synthesis and plasma cholesterol levels increase with increasing body weight, but in general, there is a poor correlation between weight loss and reduction in total or LDL plasma cholesterol.

TREATMENT OF OBESITY WEIGHT-REDUCTION MEASURES In most obese individuals, specific aberrations of the weight-control mechanism described under Pathogenesis cannot be identified and, thus, treatment must be nonspecific. Short-term weight loss often can be achieved, but obese patients tend to regain their lost weight: Only ~20% of obese patients maintain their weight loss over a 5- to 15-year period after initial treatment.32 Greater success is likely to be achieved through programs that lead to permanent changes both in eating habits and in physical activity. Achievement of this goal requires new approaches that extend beyond the confines of traditional medical settings, involving not only the physician but also various nonphysician resources in the community. It is important to remember that obesity is a medical as well as a cosmetic problem. Treatment should be considered for all with a BMI in excess of 27 or with lesser degrees of obesity in those with associated disorders, such as type 2 diabetes, hyperlipidemia, cardiovascular disease, hypertension, gout, pulmonary disease, and severe osteoarthritis. Management of type 2 diabetes associated with obesity requires not only aggressive control of blood glucose with pharmacologic agents, but also healthy nutrition, weight reduction, and exercise. Weight loss in overweight, obese individuals with type 2 diabetes (even if only partially successful) is imperative in the treatment of type 2 diabetes. As little as 5% weight loss has been shown to improve blood glucose and lipid levels in overweight and obese type 2 diabetics.

DIETARY THERAPY The cornerstone of treatment is dietary therapy (see Chap. 124). Weight loss depends on caloric deficit, and caloric deficit depends on the total number, not the kind, of calories consumed relative to calories expended. Many different types of calorie-restricted diets have been advanced for the treatment of obesity, some of which are in the form of liquid formulas, and others that use bizarre combinations of foods or nutrients based on unfounded theories that may be hazardous. None of these fad diets offers advantages over a calorie-restricted diet of normal foods that is balanced to provide the conventional distribution of carbohydrate (45–50% of total calories), fat (35% of calories), and protein (15% of calories). Low-carbohydrate, high-fat, ketogenic diets and diets very high in protein cannot be recommended for long-term treatment of obesity. The more rapid initial weight loss induced by some fad diets during the first few weeks of treatment is secondary to a loss of sodium, water, and body protein rather than fat, and after the first 2 or 3 weeks, the rate of weight loss may be disappointing. Furthermore, these diets can be associated with serious health hazards.33 Finally, the ingestion of unbalanced diets does nothing to modify eating habits or to develop healthy long-term nutritional practices, which are essential for successful long-term management of obesity. The extent to which calories should be restricted in a given patient depends on the degree of obesity and the age, sex, physical activity, and general health of the patient. A loss of ~2 pounds per week is a reasonable goal in most patients and can usually be accomplished by a daily deficit of ~1000 kcal. If the obesity is associated with severe, debilitating, or life-threatening disease, such as the pickwickian syndrome, severe congestive heart failure, or severe hyperglycemia, a more drastic reduction in calorie intake may be indicated. EXERCISE Daily exercise is an important adjunct to caloric restriction in achieving and maintaining reduced weight. However, exercise is not an effective means of inducing weight loss unless combined with the reduction of caloric intake. When combined with a hypocaloric diet, a daily exercise program, carefully planned and tailored to the patient's physical condition and ability, is important for long-term weight reduction. However, a sudden increase in physical activity in obese patients or a strenuous exercise program is hazardous and contraindicated. Exercise should begin gradually in the form of a low-level aerobic activity such as walking, swimming, or cycling. After normal or near-normal body weight is achieved, the degree of physical activity can be increased, depending on the patient's physical condition (see Chap. 132). LIFESTYLE CHANGES Successful long-term outcome of the treatment of obesity requires that an individual change or modify attitudes, beliefs, and behavior with respect to eating and physical activity. Although unusual dietary excesses and sedentary behavior may or may not be “causes” for obesity (see Thermodynamic Considerations, above), there is no way to combat obesity other than by increasing physical activity and diminishing caloric intake. These efforts may be opposed by hormonal and neural mechanisms; yet, a lifelong effort to combat any adverse feelings that may be associated with weight loss is essential. Therefore, a program of formal behavioral modification may be helpful. A behavioral modification program individualized to the person's needs and abilities, combined with a program of diet and exercise, can be advantageous. A wide variety of groups, such as Weight Watchers, TOPS, and Overeaters Anonymous, are available throughout the United States. A useful manual for the evaluation of treatment programs, entitled “Weighing the Options,” was published by the National Academy Press for the Institute of Medicine.34 Even the best of these programs has not been uniformly successful in achieving the long-term goal of maintaining reduced weight. Thus, an NIH Technology Assessment Panel concluded that, “For most weight loss methods, there are few scientific studies evaluating their effectiveness and safety. The available studies indicate that persons lose weight while participating in such programs, but after completing the program, tend to regain the weight over time.”3 WEIGHT CYCLING A number of investigators have examined the possibility that repeated bouts of weight loss and regain, termed cycling, may have deleterious consequences on energy metabolism and perhaps on the morbidity of obesity. In one study,35 it was found that rats subjected to cycling regained their lost weight more easily than did controls. However, careful review of the animal literature on this subject and the few relevant human experiments do not support this contention.36,37 Although there is continued concern about the relationship of weight variability to mortality in some epidemiologic studies,38 there is no reason to restrain a person's effort to lose weight, with its known attendant benefits, for alleged lesser, long-term hazards, for which there is no known pathogenesis and which have not been conclusively demonstrated. DRUG THERAPY Drug therapy in the treatment of obesity is of limited utility. Drugs are ineffective as the sole approach to the treatment of obesity, and it is unclear whether the modest benefits of adding drugs to caloric restriction, exercise, and behavioral modification outweigh their cost and untoward effects. Although some evidence suggests that medication can increase the weight loss achieved by caloric restriction, prolong weight maintenance, and decrease comorbidity in at least a portion of patients treated for years,39 larger, long-term studies must be performed before medication can be recommended as a primary treatment. After reviewing the available evidence, the Expert Panel on the Identification, Evaluation, and Treatment of Overweight and Obesity issued the following recommendation about the drug treatment of obesity in its report: “Weight loss drugs approved by the FDA [Food and Drug Administration] may only be used as part of a comprehensive weight loss program, including dietary therapy and physical activity, for patients with a BMI of ³30 with no concomitant obesity-related risk factors or diseases, and for patients with a BMI of ³27 with concomitant obesity-related risk factors or diseases. Weight loss drugs should never be used without concomitant lifestyle modifications. Continual assessment of drug therapy for efficacy and safety is necessary. If the drug is efficacious in helping the patient lose and/or maintain weight loss and there are no serious adverse effects, it can be continued. If not, it should be discontinued.”2 Two medications have been approved for long-term use by the FDA: sibutramine (Meridia), a norepinephrine and serotonin reuptake inhibitor that enhances satiety,40 and orlistat (Xenical), an inhibitor of pancreatic lipase.41 Thyroid hormone, diuretics, and digitalis should be used in the treatment of obesity only when specific indications exist (e.g., hypothyroidism, edema, hypertension, or congestive heart failure), and never for the achievement of weight reduction. Diabetes and hypertension occurring in obesity should be treated aggressively. Human chorionic gonadotropin and growth hormone have no place in the treatment of obesity. SURGICAL TECHNIQUES Surgical techniques, such as the vertical-banded gastroplasty and Roux en Y gastric bypass, have been used for the treatment of morbid obesity.42 An NIH Consensus Conference on Gastrointestinal Surgery for Severe Obesity had evaluated the need for this treatment.43 Surgery was only recommended for cooperative patients who have been at least twice their ideal body weight for at least 5 years, have failed all other treatment modalities, have no psychiatric, alcohol, or drug abuse problems, and have no medical contraindications to surgery. Substantial improvement in comorbid conditions has been reported; however, short- and long-term complications of surgery are common, and the risk/benefit ratio must be carefully examined for each patient. Referral of suitable patients to a medical center specializing in this type of procedure is recommended. Of note, laparoscopic techniques to accomplish stomach restriction are under investigation. However, their safety and efficacy await evaluation.

CONCLUSION The treatment of obesity is unsatisfactory and frequently unsuccessful. Nevertheless, the evidence is overwhelming that obesity is a significant risk factor for many illnesses. The physician and the obese patient must recognize that nothing short of a change in lifestyle (nutritional change, increase in physical activity, diminution of stressful life situations, etc.) is needed for treatment. No pill, formula, or unusual dietary manipulation is available to provide certain success. It may not be possible to correct all of the obesity; however, even some weight loss may ameliorate concurrent diabetes, hypertension, or other complicating diseases. The recent rapid growth of biomedical science and increasing attention to the understanding of the basics of energy metabolism may eventually provide further insights and better therapies for this unfortunate situation. CHAPTER REFERENCES

1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Bray GA, Bouchard C, James WP. Handbook of obesity. New York: Marcel Dekker Inc, 1995. Clinical guidelines on the identification, evaluation, and treatment of overweight and obesity in adults—the evidence report. Obes Res 1998; 6(Suppl 2):51S. NIH Technology Assessment Conference Panel. Methods for voluntary weight loss and control. Ann Intern Med 1993; 119:764. Brozek J, Henschel A, eds. Techniques for measuring body composition: proceedings of a conference, Quartermaster Research and Engineering Center, January 22–23, 1959. Washington, DC: National Academy of Sciences–National Research Council, 1961. Anonymous. Proceedings of a panel on the clinical uses of whole-body counting, Vienna, June 28–July 2, 1965. Vienna: International Atomic Energy Agency, 1966. Wing RR, Marcus MD, Epstein LH, et al. Long-term effects of modest weight loss in type II diabetic patients. Arch Intern Med 1987; 147:1749. Eliahou HE, Iana A, Gaon T. Body weight reduction necessary to attain normotension in the overweight hypertensive patient. Int J Obese 1981; 5(Suppl 1):157. Kalkhoff RK, Hartz AH, Rupley D, et al. Relationship of body fat distribution to blood pressure, carbohydrate tolerance and plasma lipids in healthy obese women. J Lab Clin Med 1983; 102:61. Björntorp P. Regional obesity. In: Björntorp P, Brodoff B, eds. Obesity. Philadelphia: JB Lippincott Co, 1992; 579–586. Pouliot MC, Despres JP, Lemieux S, et al. Waist circumference and abdominal sagittal diameter: best simple anthropometric indexes of abdominal visceral adipose tissue accumulation and related cardiovascular risk in men and women. Am J Cardiol 1994; 73(7):460. Van Baeyer HC. Maxwell's dream. New York: Random House Inc, 1998. Kleiber M. The fire of life: an introduction to animal energetics. New York: Robert E Krieger Publishing Co, 1975. Newburgh LH, Johnston MW. Endogenous obesity—a misconception. Ann Intern Med 1993; 3:815. Leibel RL, Rosenbaum M, Hirsch J. Changes in energy expenditure resulting from altered body weight. N Engl J Med 1995; 332:621. Rosenbaum M, Leibel RL, Hirsch J. Obesity. Medical progress. N Engl J Med 1997; 337:396. Hirsch J, Leibel RL. The genetics of obesity. Hosp Pract 1998; 33(3):55. Ristow M, Muller-Wieland D, Pfeffer A, et al. Obesity associated with a mutation in a genetic regulator of adipocyte differentiation. N Engl J Med 1998; 339:953. Ricquier D, Fleury C, LaRose M, et al. Contributions of studies on uncoupling proteins to research on metabolic disease. J Intern Med 1999; 245:637. Freake HC. A genetic mutation in PPARg is associated with enhanced fat cell differentiation: implications for human obesity. Nutr Rev 1999; 1:154. Salans LB, Cushman SW, Weismann RE. Studies of human adipose tissue: adipose cell size and number in nonobese and obese patients. J Clin Invest 1973; 52:929 Hirsch J, Batchelor BR. Adipose tissue cellularity in human obesity. Clin Endocrinol Metab 1976; 5(2):299. Nimrod A, Ryan KJ. Aromatization of androgens by human abdominal and breast fat tissue. J Clin Endocrinol Metab 1975; 40:367. Salans LB, Knittle JL, Hirsch J. Obesity, glucose intolerance and diabetes mellitus. In: Ellenberg M, Rifkin H, eds. Diabetes mellitus: theory and practice. New York: Medical Examination Publishing, 1983:469. Boden G. Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes 1990; 39:226. Rebrin K, Steil GM, Getty L, Bergman RN. Free fatty acids as a link in the regulation of hepatic glucose output by peripheral insulin. Diabetes 1995; 44:1038. Paolissa G, Gambardeira A, Amato L, et al. Effect of long-term fatty acid infusion on insulin secretion in healthy subjects. Diabetologia 1995; 38:1295. Segal KR, Landt M, Klein S. Relationship between insulin sensitivity and plasma leptin concentration in lean and obese man. Diabetes 1996; 45:988. Auwerx J, Staels B. Leptin. Lancet 1998; 351:737. Hotamisligil GS, Spiegelman BM. Tumor necrosis factor, as key component in the obesity–diabetes link. Diabetes 1994; 43:1271. Frittitta L, Youngren JF, Sbraccia P, et al. Increased adipose tissue PC-1 protein content, but not tumor necrosis factor-gene expression, is associated with a reduction of both whole body insulin sensitivity and insulin receptor tyrosine–kinase activity. Diabetologia 1997; 40:282. Pan XR, Li GW, Hu YH, et al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance: the Da Qing IGT and Diabetes Study. Diabetes Care 1997; 20:537. Drenick E. The prognosis of conventional treatment in severe obesity. In: Björntorp P, Cairella M, Howard A, eds. Recent advances in obesity research III. London: John Libbey, 1981:80. National Task Force on the Prevention and Treatment of Obesity. Very low calorie diets. JAMA 1993; 270:967. Thomas PR. Weighing the options—criteria for evaluating weight management programs. Washington, DC: National Academy Press, 1995. Brownell KD, Greenwood MRC, Stellar E, Shrager EC. The effects of repeated cycles of weight loss and regain in rats. Physiol Behav 1986; 38:459. Reed GW, Hill JO. Weight cycling: a critical review of the animal literature. Obes Res 1993; 1:392. Wing RR. Weight cycling in humans: a critical review of the literature. Ann Behav Med 1992; 14:113. Lissner L, Odell PA, D'Agostino RB, et al. Variability of body weight and health outcomes in the Framingham population. N Engl J Med 1991; 324:1839 Weintraub M. Long-term weight control: the National Heart, Lung, and Blood Institute funded multimodal intervention study. Clin Pharmacol Ther 1992; 51:581. Fanghanel G, Cortinas L, Sanchez-Reyes L, Berber A. A Clinical trial of the use of sibotramine for the treatment of patients suffering essential obesity. Int J Obes Related Metab Disord 2000; 24:144. Heymsfield SB, Segal KR, Hauptman J, et al. Effects of weight loss with orlistat on glucose tolerance and progression to type 2 diabetes in obese adults. Arch Intern Med 2000; 160:1321. Balsiger BM, Murr MM, Poggio JL, Sarr MG. Bariatric surgery. Surgery for weight control in patients with morbid obesity. Med Clin North Am 2000; 84:477. Gastrointestinal surgery for severe obesity: Proceedings of an NIH Consensus Development Conference, March 25–27, 1991. Am J Clin Nutr 1992; 55(Suppl 2):615S.

CHAPTER 127 STARVATION Principles and Practice of Endocrinology and Metabolism

CHAPTER 127 STARVATION RUTH S. MACDONALD AND ROBERT J. SMITH Metabolic Response to Starvation Hormonal Control and Adaptation During Starvation Insulin Glucagon Growth Hormone and Insulin-Like Growth Factor-I Thyroid Hormone Glucocorticoids Catecholamines Leptin Impact of Starvation on Endocrine and Nonendocrine Disease Chapter References

Protein-energy malnutrition is a major public health problem in many parts of the world. Severe protein-energy malnutrition is characterized by cessation of growth in children, body wasting, mental apathy, and loss of pigmentation of the hair and skin.1 When both dietary energy sources and protein are deficient, the syndrome called marasmus develops, characterized by generalized emaciation, an absence of subcutaneous fat, muscle wasting, and wrinkled and dry skin (Fig. 127-1, left). A severe deficiency of protein, with either low or adequate energy intake, causes the syndrome called kwashiorkor, characterized by emaciated limbs as evidence of a loss of lean body mass but a swollen abdomen secondary to edema and hepatomegaly2 (see Fig. 127-1, right). Features of marasmus and kwashiorkor often occur simultaneously or at different times in one person. Both conditions are associated with reduced resistance to disease and infection, and with diminished capacity for recovery from concurrent illnesses.1,3,4

FIGURE 127-1. Typical appearance of children with advanced protein-energy malnutrition. Left, Marasmus. Right, Kwashiorkor. (Photograph from Kivu region, Republic of the Congo, by R. J. Smith.)

Periods of partial or total starvation and resulting malnutrition often occur during the course of catabolic illnesses or after injury. During starvation in otherwise normal persons, there is a coordinated adaptive response within different tissues such that nutrient stores (glycogen, triglycerides, and protein) are used efficiently. These metabolic changes occur in response to alterations in nutrient availability and in the levels of several hormones. In treating patients with both malnutrition and an additional illness, it is important to understand these normal nutrient and hormonal control mechanisms and the possible alterations in the normal adaptive responses.

METABOLIC RESPONSE TO STARVATION The metabolic adaptation to starvation can be divided into four temporal phases: the postabsorptive period (5–6 hours after a meal), early starvation (1–7 days), intermediate starvation (1–3 weeks), and prolonged starvation (more than 3 weeks). During the first three phases, a series of metabolic changes occurs progressively, until a more stable, near steady state is reached during prolonged starvation.5 Metabolic fuel use and production by liver, muscle, adipose tissue, kidney, and brain during the four phases of starvation are summarized in Table 127-1. During the postabsorptive period and early in starvation, glucose is readily available from hepatic stores of glycogen and is consumed as a principal fuel by the brain, muscle, kidney, and other tissues. As hepatic glycogen stores become depleted during the first 24 hours of fasting, plasma levels of glucose decrease modestly, gluconeogenesis becomes a progressively more important source of glucose, and alternative substrates begin to replace glucose as a metabolic fuel.6 It has been estimated that £90% of available glucose derives from gluconeogenesis after a 40-hour fast.7 Free fatty acid release from adipose tissue triglyceride stores is augmented progressively during fasting, resulting in increased plasma fatty acid levels, consumption of fatty acids by skeletal muscle and kidney, and sparing of glucose. The synthesis of ketone bodies by the liver is activated by the increased availability of fatty acids and a rising glucagon/insulin ratio, elevating circulating levels of ketone bodies as starvation continues.8 Skeletal muscle, kidney, and brain begin to metabolize ketone bodies in proportion to their plasma concentrations, further reducing the requirement for glucose.

TABLE 127-1. Adaptations in Metabolic Fuel Use and Production during Progressive Starvation

The net breakdown of muscle protein during starvation reflects the requirements of the liver and kidney for substrates for gluconeogenesis and ammoniagenesis, respectively. Alanine and other amino acids released from muscle are converted into glucose in the liver,9 whereas muscle-derived glutamine10 is utilized primarily in renal ammoniagenesis or as a fuel and precursor of gluconeogenic amino acids in the gastrointestinal tract.11 During early starvation, the increased urinary levels of metabolically generated anions in the form of ketone bodies require an accompanying excretion of sodium. This probably explains the natriuresis and diuresis, correlated with rapid weight loss, that occur during the first days of a fast.12 As total body sodium becomes depleted and ketone body production increases, ammonium derived primarily from glutamine replaces sodium as the primary urinary cation. As starvation progresses and the levels of ketone bodies and free fatty acids continue to rise, diminished requirements for endogenous glucose production enable an adaptive conservation of body protein.13 Thus, during long-term fasting, amino-acid conversion to glucose is minimized, and the net breakdown of muscle protein is governed largely by the glutamine requirement for maintenance of acid-base balance through renal ammoniagenesis pathways.14 This series of metabolic adaptations during fasting assures adequate metabolic fuels for all tissues and the efficient utilization of nutrient stores. The net result is a

remarkable capacity for survival in the absence of food intake. The maximal duration of fasting compatible with life in initially normal-weight persons is demonstrated by the unfortunate example of political protesters in Northern Ireland, who died after 45 to 76 days of essentially total fasting.15 Obese patients, who may have as much as four times the normal caloric reserve, have been treated clinically with total fasting for more than 200 days without serious complications,16 although the apparent risk of sudden death from cardiac failure makes fasting of this duration inadvisable.17 Thus, the length of time a person can survive starvation differs with body composition. Generally, the loss of one-third to one-half of body nitrogen is incompatible with life in lean or obese patients, but the large fat stores in obesity allow longer-term sparing of body nitrogen. Patients with debilitating illnesses or malnourished, wasted persons may have a markedly decreased tolerance for nutrient deprivation.

HORMONAL CONTROL AND ADAPTATION DURING STARVATION The most important factor controlling the metabolic adaptation to starvation is insulin. Without a regulated decrease in the secretion and the circulating levels of insulin, endogenous glucose-generating pathways cannot be activated, and alternative fuel stores cannot be mobilized. Superimposed on these dominant actions of insulin are the effects of multiple other catabolic and anabolic hormones, including glucagon, growth hormone, growth factors, thyroid hormone, and glucocorticoids. Regulated change in the levels of each of these hormones is essential for effective metabolic adaptation during fasting. INSULIN The changes in circulating hormone levels during starvation are illustrated in Figure 127-2. Plasma insulin levels decrease to approximately half the normal postabsorptive levels during the first 3 days of food withdrawal and then stabilize at low but physiologically significant concentrations.18 This adaptation in insulin secretion develops gradually during fasting and, in turn, normalizes only after several days of refeeding. Thus, when glucose or other food is ingested after a period of fasting, there is a subnormal insulin secretory response.19 In some nondiabetic persons, glucose intolerance may be observed after periods of inadequate food ingestion; therefore, the results of glucose tolerance tests can be difficult to interpret unless patients have had adequate carbohydrate and total caloric intake for several days before testing.

FIGURE 127-2. Changes in hormone levels during fasting. A, Insulin. B, Glucagon. (Data in A and B from Marliss EB, Aoki TT, Unger RH, et al Glucagon levels and metabolic effects in fasting man. J Clin Invest 1970; 49:2256.) C, Growth hormone and insulin-like growth factor-I (IGF-I) (Data in C from Cahill GF Jr, Herrera MG, Morgan AP, et al. Hormone fuel interrelationships during fasting. J Clin Invest 1966; 45:1751; and Isley WL, Underwood LE, Clemmons DR. Dietary components that reg ulate serum somatomedin-C concentrations in humans. J Clin Invest 1983; 71:175.) D, Triiodothyronine (T3) and reverse triiodothyronine (rT3). (Data in D from Merimee TJ, Fineberg ES. Starvation-induced alterations of circulating thyroid hormone concentrations in man Metabolism 1976; 25:79; and Gardner DF, Kaplan MM, Stanley CA, Uti ger RD. Effect of triiodothyronine replacement on the metabolic and pituitary responses to starvation. N Engl J Med 1979; 300:579.)

GLUCAGON During starvation, the blood glucose concentration is maintained in part through the action of glucagon. Serum glucagon levels increase within the first 48 hours and continue to rise throughout the first several days of fasting in normal humans.18 In obese persons undergoing prolonged fasting, plasma glucagon concentrations rise initially but return toward prefast levels after 3 to 4 weeks.20 Probably, in early starvation, hyperglucagonemia is attributable to decreased clearance of the hormone rather than to increased secretion.20 During prolonged starvation, glucagon levels return toward normal because of a decrease in secretion, with continued decreased clearance of the hormone. Pancreatic stores of glucagon do not appear to be diminished after short-term fasting, as evidenced by an exaggerated glucagon secretory response to arginine infusion.21 With low concentrations of insulin in the fasting state, glucagon contributes to the maintenance of glucose homeostasis by stimulating hepatic gluconeogenesis and glycogenolysis.8,22 An increase in the plasma glucagon/insulin ratio during early starvation also promotes the generation of alternative fuels by increasing hepatic synthesis of ketone bodies and mobilizing free fatty acids from adipose tissue.23 In prolonged starvation, circulating glucagon returns to postabsorptive levels concurrent with the reduced demand for glucose (see Chap. 134). GROWTH HORMONE AND INSULIN-LIKE GROWTH FACTOR-I Serum growth hormone levels also become elevated early in starvation.24 The role of growth hormone as a metabolic regulator during fasting is unclear, however, because individual values are variable and do not correlate closely with metabolic adaptations. Because patients with growth hormone deficiency can become hypoglycemic during fasting,25 it seems reasonable to conclude that adequate levels of the hormone are important for the maintenance of blood glucose in the fasting state but that elevated concentrations have an otherwise minor role in the metabolic adaptation to fasting. Possibly, elevated growth hormone secretion during fasting is a consequence, in part, of the normal feedback relation between insulin-like growth factor-I (IGF-I) and growth hormone.26,26a IGF-I (designated somatomedin C in earlier literature) is a peptide growth factor that functions as an important mediator of the anabolic actions of growth hormone27 (see Chap. 12 and Chap. 173). Growth hormone stimulates IGF-I synthesis in the liver and some peripheral tissues, and there is evidence that IGF-I in turn exerts negative feedback inhibition on growth hormone secretion.28 Although growth hormone appears to be the principal regulator of IGF-I, other factors, such as thyroid hormone, insulin, estrogens, and nutritional status, also influence IGF-I levels.29 As a consequence, circulating levels of IGF-I decrease during fasting in spite of elevated growth hormone levels (see Fig. 127-2),30,31 and feedback inhibitory effects of IGF-I on growth hormone secretion do not occur. During fasting, growth hormone binding in the liver is reduced in parallel with decreased IGF-I mRNA synthesis, suggesting growth hormone resistance as a mechanism contributing to decreased circulating IGF-I.32 Decreased IGF-I also occurs during protein restriction, however, even though liver growth hormone binding is not altered.33 Thus, protein restriction appears to affect growth hormone regulation of circulating IGF-I through a postreceptor mechanism. In addition to these effects on the liver, fasting reduces IGF-I mRNA levels in peripheral tissues, including kidney, muscle, gut, and brain, while simultaneously increasing IGF-I–receptor mRNA.34 The combination of low IGF-I and high growth hormone levels may have adaptive value by diminishing the energy expenditure necessary for growth-related processes and yet enabling growth hormone to promote the mobilization of alternative fuels through its lipolytic actions.31 The availability of IGF-I to tissues is directly influenced by circulating IGF-binding proteins (IGFBPs). IGFBP-3 is thought to be the major carrier of IGFs in the circulation. Prolonged fasting or protein depletion is correlated with decreased concentrations of IGFBP-3.35 IGFBP-2 and IGFBP-1 may mediate cellular transport of the IGFs. Both IGFBP-2 and IGFBP-1 tend to increase during prolonged fasting, likely as a consequence of decreased insulin concentrations.35 In malnourished patients given nutritional support, the direction of changes in serum IGF-I levels correlates closely with the direction of changes in nitrogen balance. Refeeding adequate energy but low protein only partially restores circulating IGF-I concentrations. When total parenteral nutrition is provided to nutritionally deprived patients, serum IGF-I levels rise at a rate corresponding to the improvement in nitrogen balance.36 This does not appear to reflect simply increased synthesis of hepatic proteins secondary to improved nutrition, because the levels of other hepatic secretory proteins are not consistently raised during the time IGF-I increases. IGF-I may prove to be a sensitive indicator of the response to nutritional rehabilitation.37 THYROID HORMONE One of the energy-conserving adaptations during prolonged starvation is a decrease in the basal metabolic rate, which is mediated at least in part by altered levels of triiodothyronine (T3). In normal humans undergoing a short-term fast, serum T3 levels decrease, whereas levels of thyroxine (T4), thyroid-stimulating hormone, and the response of thyroid-stimulating hormone to thyroid-releasing hormone are unchanged.38,39 and 40 Experimentally, starvation causes an increased sensitivity of pituitary

thyrotrope cells to T3. The decrease in serum T3 coincides with an increase in less biologically potent reverse T3 (see Fig. 127-2), suggesting that peripheral tissues convert a greater proportion of T4 into reverse T3 during fasting (see Chap. 30 and Chap. 36). A similar hormonal pattern develops in adults with protein-energy malnutrition.42 If T3 levels are maintained by the administration of T3 during fasting, an increase in the catabolism of nutrient stores occurs, as demonstrated by elevated levels of glucose, free fatty acids, and ketone bodies,43 and by increased urinary urea excretion.39 Therefore, it is thought that diminished levels of T3 are an important determinant of the decreased oxygen consumption observed in prolonged starvation and may be protective in limiting muscle protein breakdown. Reduced thyroid receptor protein levels also occur in starvation,43 and thus, by reducing tissue responsiveness to the catabolic effects of T3, may also have a role in the adaptation to starvation. GLUCOCORTICOIDS Although early work provided evidence for an absence of changes in the pituitary-adrenal axis in obese patients undergoing prolonged starvation,45 more recent studies have demonstrated modest elevations in glucocorticoid levels during fasting.46 In catabolic states associated with injury or systemic illness (see later), more marked glucocorticoid elevations are believed to contribute to tissue breakdown. It is possible that the effects of glucocorticoids in starvation as compared to systemic disease states may depend on the magnitude of change in circulating levels. Patients with glucocorticoid deficiency frequently have fasting hypoglycemia, because adequate glucocorticoid levels are necessary to maintain the activity of pyruvate carboxylase, a rate-limiting enzyme for gluconeogenesis,47 and for the release of gluconeogenic amino acids from skeletal muscle.48 Thus, it appears that normal or modestly elevated levels are necessary for the adaptive response to starvation. CATECHOLAMINES Increased sympathetic nervous system activity may contribute to the elevation of plasma free fatty acids and glucagon concentrations, and to the maintenance of blood glucose levels during fasting, but several lines of evidence suggest that this is not an important factor.49 For example, normal elevations in free fatty acid levels occur during fasting in patients who have undergone adrenalectomy, adrenergic blockade, or even complete disruption of the sympathetic efferent pathways at the level of the cervical cord.50,51 and 52 Similarly, catecholamine deficiency has not been associated with the development of fasting hypoglycemia. Therefore, it appears that sympathetic activity neither initiates nor maintains the metabolic adaptations to starvation. LEPTIN The obesity gene product, leptin, is released from adipose tissue, and circulating concentrations increase with increased body fat content. Leptin has been proposed as an afferent signal between adipose tissue and the central nervous system that regulates body composition. In accordance with this role, leptin levels decrease with loss of body weight and reduction in adipose tissue mass.52a In short-term fasting, however, leptin levels decrease more rapidly and are more rapidly restored with refeeding than is body fat.53,54 In mice, repletion of leptin during fasting partially reverses changes in thyroid, adrenal, and gonadal hormones, suggesting a role for decreasing leptin as an activator of neuroendocrine adaptations to starvation.55,56 The regulation of leptin may involve both glucose and insulin concentrations, as no decrease in leptin occurs during fasting when glucose or insulin concentrations are maintained.54

IMPACT OF STARVATION ON ENDOCRINE AND NONENDOCRINE DISEASE The adaptive metabolic responses to starvation make it possible for humans to survive long periods of inadequate nutrient intake.57 Because the metabolic adaptation to fasting is largely the result of endocrinologic regulation, the normal response can be markedly altered by endocrine disorders. For example, lack of a single hormone, such as cortisol58 or growth hormone,59 can cause severe hypoglycemia within a few hours of food deprivation. Fortunately, these primary endocrinologic diseases are relatively uncommon and generally recognized early. Much more commonly, patients with nonendocrine disease are nutritionally depleted but unable to generate a normal endocrine response to partial or total fasting. Consequently, malnutrition can have a significant effect on morbidity and mortality. For example, after traumatic injuries, major operative procedures, burns, systemic infections, or other serious illnesses, a well-characterized catabolic stress response develops.60 An important adaptive function of this response appears to be the accelerated mobilization of body nutrient stores, which are used in tissue repair, by inflammatory cells, and for processes such as acute-phase protein synthesis. The metabolic alterations of stress depend in large part on changes in hormone levels, which, as illustrated in Figure 127-3, are different from the endocrine response to fasting. Levels of sympathetic nervous system activity and circulating catechola-mines rise early,61 followed somewhat later by more sustained elevations of glucocorticoids,62 glucagon,63 and growth hormone.64 Insulin levels may be decreased early but ultimately are normal or elevated.65 In addition, wounds, sepsis, or inflammation induce a variety of hormonal mediators, or cytokines, such as tumor necrosis factor and the interleukins, which exacerbate skeletal muscle breakdown and debilitation.66 Attenuation of the metabolic effects of these stress mediators through nutritional intervention is limited, although specific nutrient formulas in combination with trophic hormones potentially may provide beneficial effects.67

FIGURE 127-3. Effects of burn injury on basal hormone levels. Plasma hormone concentrations were determined after 8 hours of fasting in 15 subjects with 25% to 90% burns. (Redrawn from Wolfe RR, Durkot MJ, Allsop JR, Burke JF. Glucose metabolism in severely burned patients. Metabolism 1979; 28:1031.)

When decreased nutrient intake or total fasting coincides with severe injury or illness, the normal metabolic adaptation to fasting does not develop. Instead of decreased energy requirements and a lowered basal metabolic rate, energy requirements remain increased as a result of the demands for tissue repair processes, inflammatory cell formation and function, hyperdynamic circulation, and, sometimes, fever-associated shivering. Increased concentrations of stress hormones, in particular glucocorticoids, greater than the levels characteristic of fasting may lead to the net catabolism of tissue protein and may increase negative nitrogen balance.68 Insulin levels remain normal or elevated in spite of fasting, reflecting a state of insulin resistance that prevents insulin from counteracting the catabolic effects of the stress hormones. In spite of this insulin resistance, there may be enough antilipolytic action of insulin in some patients to diminish free fatty acid mobilization and ketone body synthesis and, thus, an impaired transition to lipid fuel consumption during fasting. Consequently, in some patients with multiple traumatic injuries, levels of plasma ketone bodies remain low during fasting.69 This is associated with increased urinary nitrogen excretion and, probably, less effective protein conservation. As in starvation, adaptation to a catabolic stress ultimately is characterized by a reduction of muscle protein catabolism and increased use of free fatty acids and ketone bodies.70 Depending on the demands for glucose consumption and overall energy use, and on the effectiveness of alternative fuel mobilization, however, the adaptation may be delayed for many weeks. Thus, even modest degrees of undernutrition or short periods of total fasting can lead to significant nutritional depletion, such that the provision of adequate nutritional support in the setting of catabolic diseases may be essential for optimal recovery or even survival.71 For this reason, parenteral feeding should be considered early in the course of critically ill patients if oral food intake is inadequate. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7.

Olsen RE. Protein-calorie malnutrition. New York: Academic Press, 1975. Winick M. Long term effects of kwashiorkor. J Pediatr Gastroenterol Nutr 1987; 6:833. Wolstenholme GEW, O'Connor M. Nutrition and infection. Boston: Little, Brown, 1967. Adhikari M, Gita-Ramjee, Berjak P. Aflatoxin, kwashiorkor, and morbidity. Nat Toxins 1994; 2:1. Cahill GF, Owen OE, Morgan AP. The consumption of fuels during prolonged starvation. Adv Enzyme Regul 1968; 6:143. Webber J, Macdonald IA. The cardiovascular, metabolic and hormonal changes accompanying acute starvation in men and women. Br J Nutr 1994; 71:437. Katz J, Tayek JA. Gluconeogenesis and the Cori cycle in 12-, 20-, and 40-h-fasted humans. Am J Physiol 1998; 275:E537.

8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

Gelfand RA, Sherwin RS. Glucagon and starvation. In: Lefebvre PJ, ed. Glucagon II. Handbook of experimental pharmacology, vol 66[II]. Berlin: Springer-Verlag, 1983:223. Felig P, Marliss E, Owen OE, Cahill GF Jr. Blood glucose and gluconeogenesis in fasting man. Arch Intern Med 1969; 123:293. Marliss EB, Aoki TT, Pozefsky T, et al. Muscle and splanchnic glutamine and glutamate metabolism in postabsorptive and starved man. J Clin Invest 1971; 50:814. Windmueller HG. Glutamine utilization by the small intestine. Adv Enzymol Relat Areas Mol Biol 1982; 53:202. Sigler MH. The mechanism of the natriuresis of fasting. J Clin Invest 1975; 55:377. How ketones spare protein in starvation. (Editorial). Nutr Rev 1989; 3:80. Aoki TT, Mueller WA, Brennan MF, Cahill GF Jr. Metabolic effects of glucose in brief and prolonged fasted man. Am J Clin Nutr 1975; 28:507. Kerndt PR, Naughton JL, Driscoll CE, Loxterkamp DA. Fasting: the history, pathophysiology and complications. West J Med 1982; 137:379. Thomson TJ, Runcie J, Miller V. Treatment of obesity by total fasting for up to 249 days. Lancet 1966; 2:992. Garnett ES, Barnard DL, Ford J, et al. Gross fragmentation of cardiac myo-fibrils after therapeutic starvation for obesity. Lancet 1969; 1:914. Marliss EB, Aoki TT, Unger RH, et al. Glucagon levels and metabolic effects in fasting man. J Clin Invest 1970; 49:2256. Fink G, Gutman RA, Cresto JC, et al. Glucose-induced insulin release patterns: effect of starvation. Diabetologia 1974; 10:421. Fisher M, Sherwin RS, Hendler R, Felig P. Kinetics of glucagon in man: effects of starvation. Proc Natl Acad Sci U S A 1976; 73:1735. Aguilar-Parada E, Eisentraut AM, Unger RH. Effects of starvation on plasma pancreatic glucagon in normal man. Diabetes 1969; 18:717. Ven de Werve G, Hue L, Hers HG. Hormonal and ionic control of the gly-cogenolytic cascade in rat liver. Biochem J 1977; 162:135. McGarry JD, Foster DW. Regulation of hepatic fatty acid oxidation and ketone body production. Annu Rev Biochem 1980; 493:395. Cahill GF Jr, Herrera MG, Morgan AP, et al. Hormone-fuel interrelationships during fasting. J Clin Invest 1966; 45:1751. Merimee TJ, Felig P, Marliss E, et al. Glucose and lipid homeostasis in the absence of human growth hormone. J Clin Invest 1971; 50:574. Abe H, Molitch ME, Van Wyk JJ, Underwood LE. Human growth hormone and somatomedin C suppress the spontaneous release of growth hormone in unanesthetized rats. Endocrinology 1983; 113:1319.

26a.Thissen JP, Underwood LE, Ketelslegers JM. Regulation of insulin-like growth factor-I in starvation and injury. Nutr Rev 1999; 57:167. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52.

Chochinov RH, Daughaday WH. Current concepts of somatomedin and other biologically related growth factors. Diabetes 1976; 25:994. Clemmons DR, Underwood LE. Somatomedin-C/insulin-like growth factor I in acromegaly. J Clin Endocrinol Metab 1986; 15:629. Phillips LS, Goldstein S, Gavin JR III. Nutrition and somatomedin XVI: somatomedins and somatomedin inhibitors in fasted and refed rats. Metabolism 1988; 37:209. Isley WL, Underwood LE, Clemmons DR. Dietary components that regulate serum somatomedin-C concentrations in humans. J Clin Invest 1983; 71:175. Merimee TJ, Zapf J, Froesch ER. Insulin-like growth factors in the fed and fasted states. J Clin Endocrinol Metab 1982; 55:999. Straus DS, Takemoto CD. Effect of fasting on insulin-like growth factor-I (IGF-I) and growth hormone receptor mRNA levels and IGF-I gene transcription in rat liver. Mol Endocrinol 1990; 4:91. Maiter DM, Maes M, Underwood LE, et al. Early changes in serum concentrations of somatomedin-C induced by dietary protein deprivation: contributions of growth hormone receptor and post-receptor defects. J Endocrinol 1988; 118:113. Lowe WJ, Adamo M, Werner H, et al. Regulation by fasting of insulin-like growth factor-I and its receptor: effects on gene expression and binding. J Clin Invest 1989; 84:619. Clemmons DR, Underwood LE. Nutritional regulation of IGF-I and IGF binding proteins. Annu Rev Nutr 1991; 11:393. Clemmons DR, Underwood LE, Dickerson RN, et al. Use of plasma somatomedin-C/insulin-like growth factor I measurements to monitor the response to nutritional repletion in malnourished patients. Am J Clin Nutr 1985; 41:191. Unterman TG, Vazquez RM, Slas AJ, et al. Nutrition and somatomedin XIII: usefulness of somatomedin-C in nutritional assessment. Am J Med 1985; 78:228. Merimee TJ, Fineberg ES. Starvation-induced alterations of circulating thyroid hormone concentrations in man. Metabolism 1976; 25:79. Gardner DF, Kaplan MM, Stanley CA, Utiger RD. Effect of triiodothyronine replacement on the metabolic and pituitary responses to starvation. N Engl J Med 1979; 300:579. Vagenakis AG. Thyroid hormone metabolism in prolonged experimental starvation in man. In: Vigersky RA, ed. Anorexia nervosa. New York: Raven Press, 1977:243. Hugues JN, Enjalbert A, Burger AG, et al. Sensitivity of thyrotropin (TSH) secretion to 3,5,3'-triiodothyronine and TSH-releasing hormone in rats during starvation. Endocrinology 1986; 119:253. Chopra IJ, Smith SR. Circulating thyroid hormones and thyrotropin in adult patients with protein-calorie malnutrition. J Clin Endocrinol Metab 1975; 40:221. Carter WJ, Shakir KM, Hodges S, et al. Effect of thyroid hormone on metabolic adaptation to fasting. Metabolism 1975; 24:1177. Tagami T, Nakamura H, Sasaki S, et al. Starvation-induced decrease in the maximal binding capacity for triiodothyronine of the thyroid hormone receptor is due to a decrease in the receptor protein. Metabolism 1996; 45:970. Sabeh G, Alley RA, Robbins TJ, et al. Adrenocortical indices during fasting in obesity. J Clin Endocrinol Metab 1969; 29:373. Bergendahl M, Vance ML, Iranmanesh A, et al. Fasting as a metabolic stress paradigm selectively amplifies corticol secretory burst mass and delays the time of maximal nyctohemeral cortisol concentrations in men. J Clin Endo-crinol Metab 1996; 81:692. Baxter JD, Forsham PH. Tissue effects of glucocorticoids. Am J Med 1972; 53:573. Bondy PK, Ingle DJ, Meeks RC. Influence of adrenal cortical hormones upon the level of plasma amino acids in eviscerate rats. Endocrinology 1954; 55:354. Jung RT, Shetty PS, James WPT. Nutritional effects on thyroid and catecholamine metabolism. Clin Sci 1980; 58:183. Levy AC, Ramey ER. Effect of autonomic blocking agents on depot fat mobilization in normal and adrenalectomized animals. Proc Soc Exp Biol Med 1958; 99:637. Misbin RI, Edgar PJ, Lockwood DH. Adrenergic regulation of insulin secretion during fasting in normal subjects. Diabetes 1970; 19:688. Brodows RG, Campbell RG, Al-Aziz AJ, Pi-Sunyer FX. Lack of central autonomic regulation of substrate during early fasting in man. Metabolism 1976; 25:803.

52a.Klein S, Horowitz JF, Landt M, et al. Leptin production during early starvation in lean and obese women. Am J Physiol Endoc Metab 2000; 278:E280. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

Weigle DS, Duell PB, Connor WE, et al. Effect of fasting, refeeding, and dietary fat restriction on plasma leptin levels. J Clin Endocrinol Metab 1997; 82:561. Boden G, Chen X, Moxxoli M, Ryan I. Effect of fasting on serum leptin in normal human subjects. J Clin Endocrinol Metab 1996; 81:3419. Ahima RS, Prabakaran D, Mantzoros C, et al. Role of leptin in the neuroendocrine response to fasting. Nature 1996; 382:250. Flier JS. What's in a name? In search of leptin's physiologic role. J Clin Endocrinol Metab 1998; 83:1407. Joslin EP. Treatment of diabetes mellitus. Philadelphia: Lea & Febiger, 1916:243. Bondy PK. Disorders of the adrenal cortex. In: Wilson JD, Foster DW, eds. Williams textbook of endocrinology, 7th ed. Philadelphia: WB Saunders, 1985:816. Goodman HG, Grumbach MM, Kaplan SJ. Growth and growth hormone II: a comparison of isolated growth-hormone deficiency and multiple pituitary-hormone deficiencies in 35 patients with idiopathic hypopituitary dwarfism. N Engl J Med 1968; 278:57. Wolfe RR, Durkot MJ, Allsop JR, Burke JF. Glucose metabolism in severely burned patients. Metabolism 1979; 28:1031. Jaattela A, Alho A, Avikainen V, et al. Plasma catecholamines in severely injured patients: a prospective study on 45 patients with multiple injuries. Br J Surg 1975; 62:177. Vaughan GM, Becker RA, Allen JP, et al. Cortisol and corticotrophin in burned patients. J Trauma 1982; 22:263. Wilmore DW, Lindsey CA, Moylan JA, et al. Hyperglucagonaemia after burns. Lancet 1974; 1:73. Ross H, Johnston IDA, Welborn TA, Wright AD. Effect of abdominal operation on glucose tolerance and serum levels of insulin, growth hormone, and hydrocortisone. Lancet 1966; 2:563. Black PR, Brooks DC, Bessey PQ, et al. Mechanisms of insulin resistance following injury. Ann Surg 1982; 196:420. Hardin TC. Cytokine mediators of malnutrition: clinical implications. Nutr Clin Pract 1993; 8:55. Wilmore DW. Catabolic illness. N Engl J Med 1991; 325:695. Tishler ME, Leng E, Al-Kanhal M. Metabolic response of muscle to trauma: altered control of protein turnover. In: Dietze G, Kleinberger W, eds. Clinical nutrition and metabolic research proceedings, 7th congress. Munich: ESPEN, 1985:40. Smith R, Fuller DJ, Wedge JH, et al. Initial effect of injury on ketone bodies and other blood metabolites. Lancet 1975; 1:1. Krause MV, Mahan LK. The metabolic stress response and methods for providing nutritional care to stressed patients. In: Food, nutrition, and diet therapy, 6th ed. Philadelphia: WB Saunders, 1979:694. Ingenbleek Y, Bernstein LH. The nutrionally dependent adaptive dichotomy (NDAD) and stress hypermetabolism. J Clin Ligand Assay 1999; 22:259.

CHAPTER 128 ANOREXIA NERVOSA AND OTHER EATING DISORDERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 128 ANOREXIA NERVOSA AND OTHER EATING DISORDERS MICHELLE P. WARREN AND REBECCA J. LOCKE Anorexia Nervosa Characteristics Clinical Syndrome Endocrinopathy Neuroendocrine and Psychological Characteristics and Interrelations Osteoporosis Bulimia Characteristics Clinical Manifestations Treatment of Anorexia Nervosa and Bulimia Chapter References

Starvation engenders various adaptive metabolic and endocrine changes that decrease caloric demands and permit survival. The starvation associated with the syndrome of anorexia nervosa combines medical, endocrine, and psychological manifestations. The abnormality is partly hypothalamic in origin and represents an adaptation to the starvation state. 1,2 Although not associated with starvation, bulimia (abnormal means of purging calories or weight) similarly involves endocrine and behavioral symptoms.

ANOREXIA NERVOSA CHARACTERISTICS Anorexia nervosa usually occurs during adolescence and in women younger than 25 years. Amenorrhea, weight loss, and behavioral changes constitute the classic triad, any one of which may precede the others. The incidence differs greatly among population groups, and high-risk populations are found.3 For example, 1 in 100 middle-class adolescent girls and 1 in 20 to 1 in 5 professional ballet dancers have anorexia.4,5 and 6 The Diagnostic and Statistical Manual of Mental Disorders, Revised Third Edition, estimates that the rate of anorexia in girls between the ages of 12 and 18 years ranges from 1 in 800 to 1 in 100.7 The high incidence of this disorder in dancers derives from the rigid standards for thinness as well as the many hours of exercise this profession entails.5 (Increased levels of activity and restricted eating can induce self-starvation in rats, providing an interesting animal model.8) On the other hand, the condition is rare among blacks, including black ballet dancers, despite their exposure to the rigid standards of competition and weight restriction.5,9 This difference may relate to different social or cultural influences or, perhaps, to more efficient metabolic mechanisms. The risk that a sister of a patient with anorexia nervosa will acquire the illness is 6%; this fact, along with studies of monozygous twins, suggests that inborn metabolic factors contribute to the syndrome.10 The incidence of anorexia nervosa is increased in Turner syndrome, diabetes mellitus, and Cushing disease.6,11 Family histories of depression, alcohol, and drug abuse or dependence may be risk factors for the disorder.12 The female/male ratio is 9:1.13 The condition has been reported in men who are training for competitive activity while restricting their weight.14 The age of onset shows a bimodal pattern, with a high incidence between 13 and 14 years, and again between 17 and 18 years. The dieting behavior is related to pubertal maturational development and may coincide with the rapid accumulation of fat that is normal at that time. Dieting in this age group also is related to negative feelings about body image independent of weight.15,16 CLINICAL SYNDROME Anorexia nervosa represents a prototype of “hypothalamic amenorrhea.” The reproductive and physiologic adjustments appear to be an adaptive phenomenon appropriate for the semistarved state. Generally, recovery from the amenorrhea and the psychiatric disturbance parallels the weight gain. The criteria for the diagnosis of anorexia nervosa appear in the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition.17 This new definition contains bulimic subgroups. The diagnosis of anorexia nervosa depends on finding a symptom complex that includes severe weight loss (usually to less than 80% of ideal body weight; Fig. 128-1); behavioral changes such as hyperactivity and preoccupation with food; and perceptual changes, in particular, a distorted view of the body accompanied by an unreasonable concern about being “too fat.” The amenorrhea may occur at any time, even preceding the syndrome, but often is related to the start of the food restriction, even if weight loss has been slight. If the weight is lost before menarche, patients may have primary amenorrhea.

FIGURE 128-1. Young woman with severe anorexia nervosa.

The hyperactivity may begin in the guise of an athletic pursuit. Excessive exercise is often an integral aspect of the disorder in its acute phases. Moreover, the combination of strenuous physical activity and immoderate dieting has been found to lead to comorbidity.18 Anorectic patients may demonstrate an intense interest in low-calorie foods, with a large intake of diet sodas and raw vegetables, and avoidance of fried foods or other products high in calories. Hypercarotenemia may give a yellow cast to the skin, especially on the palms and soles; the sclerae remain clear. The high serum carotene level is only partly attributable to an increased intake of raw vegetables; metabolism of carotene, a precursor of vitamin A, also is decreased.6 The possible complaints and physical findings of the anorectic patient include abdominal pain, intolerance to cold, vomiting, hypotension, hypothermia, dry skin, lanugo-type hair, bradycardia, a systolic murmur, pedal edema, petechiae, and acrocyanosis. Despite the frequent leukopenia, which may be marked, the risk of infection is not increased, and cell-mediated immunity is intact. Anemia, thrombocytopenia, and hypoplastic bone marrow may be present. Severe hypoalbuminemia is rare; nevertheless, pitting edema may occur, especially with refeeding. Dehydration may cause an elevated blood urea nitrogen level, which returns to normal after rehydration. Electrolyte abnormalities and hypophosphatemia may occur. Occasionally, hypoglycemia has been severe enough to cause coma.11 Hematologic abnormalities, including decreased total leukocyte, monocyte, neutrophil, and platelet counts, appear to be correlated with total body fat–mass depletion.19 Numerous medical problems occur in anorexia nervosa, including salivary gland enlargement, pericardial effusion, pancreatitis, pancreatic insufficiency, delayed gastric emptying, poor intestinal motility, liver dysfunction with increased hepatic enzyme levels, pneumomediastinum, kidney stones, coagulopathies, bilateral peroneal nerve palsies, and deficiencies of thiamine and zinc. As a heat-conserving mechanism, marked vasoconstriction of the extremities may be present. If the weight loss occurs at or before the growth spurt, permanent growth deficiency may result. Duodenal ulcers may occur, and disregarding complaints of abdominal pain is unwise.6 Electrocardiographic changes include low-voltage, inverted T waves and, sometimes, arrhythmias. Hypoglycemia has been associated with coma and with low or absent insulin and C peptide levels, suggesting that this problem results from malnutrition.20 Osteoporosis and fractures occur as a result of long-term poor nutritional intake and prolonged estrogen deficiency.6,21,22 and 23 Computed tomographic scanning or

magnetic resonance imaging of the brain occasionally demonstrates enlargement of the cortical sulci and subarachnoid spaces, as well as cerebral atrophy; surprisingly, in one patient, the atrophy was reversed with weight gain.6 Changes in the brain due to anorexia nervosa may be irreversible. Weight recovery does not seem to reverse unilateral temporal lobe hypoperfusion, gray matter volume deficits, or various localized brain dysfunction.24,25 and 26 ENDOCRINOPATHY HYPOTHALAMIC DYSFUNCTION Low serum levels of luteinizing hormone (LH) and follicle-stimulating hormone (FSH) are associated with a profound estrogen deficiency.3 Serum levels of triiodothyronine (T3) and thyroxine (T4) may be low. The serum cortisol concentration may be high, a finding that differentiates anorexia nervosa from pituitary insufficiency. Gonadotropin Abnormalities. In anorexia nervosa, the normal episodic, pulsatile variation in the secretion of LH is absent, and a pattern typical of early puberty may be seen27 (Fig. 128-2). These abnormalities normalize with weight gain. In addition, the pattern of gonadotropin secretion can be normalized by the pulsatile administration of luteinizing hormone– releasing hormone (LHRH)28; if this hormone is injected every 2 hours, menstrual bleeding and ovulation also can be induced. Interestingly, similar disturbances in LH secretion are noted in normal women who are exposed to a low-calorie diet.29

FIGURE 128-2. Plasma luteinizing hormone (LH) concentration measurements taken every 20 minutes for 24 hours during acute exacerbation of anorexia nervosa (upper panel) and after clinical remission with return of body weight to normal (lower panel). The latter represents a normal adult pattern. (REM, rapid eye movement.) (Boyar RN, Katz J, Finkelstein JW, et al. Anorexia nervosa: immaturity of the 24-hour luteinizing hormone secretory pattern. N Engl J Med 1974; 291:861.)

The amenorrhea probably is secondary to altered signals reaching the medial central hypothalamus from the arcuate nucleus or higher levels. (The arcuate nucleus probably is the center responsible for the episodic stimulation of LHRH.) In a subset of patients, pulsations of LH are restored by administration of naloxone, suggesting that increased opioid activity participates in the suppressed LHRH pulsations. Experiments with another opioid inhibitor, naltrexone, have shown variable effects on LH secretion, a finding which indicates that the suppression of LH is not consistently opioid linked.30,31 The pattern of response to injected LHRH is immature, resembling that seen in prepubertal children—the FSH response is much greater than that of LH. The normalization of LH/FSH ratios with repeated injections of LHRH suggests that the pituitary gonadotropes have become sluggish because of the lack of endogenous stimulation, and that the episodic stimulation is important in determining the relative amounts of LH and FSH secreted. Moreover, patients who recover partially from anorexia nervosa tend to have exaggerated responses to injected LHRH.32 These changes are seen in normal children during early puberty. Thus, all of these changes indicate that the hypothalamic signals of the central nervous system revert to a prepubertal or pubertal state. Hypometabolic Manifestations. Despite their marked cachexia, patients with anorexia nervosa have clinical and metabolic signs suggestive of hypothyroidism: constipation, cold intolerance, bradycardia, hypotension, dry skin, prolonged ankle reflexes, low basal metabolic rate, and carotenemia.3 In addition, the altered metabolism of certain sex steroids, such as testosterone, is analogous to that seen in hypothyroidism.3 Some of these changes suggest a compensatory hypometabolism. Studies of the circulatory system show that, during maximal exercise, the attainable oxygen uptake and heart rate are low in children with anorexia nervosa; the maximal aerobic power (VO2max) appears to be decreased disproportionately to the circulatory and body dimensions.3 An adaptation to caloric restriction, with metabolic rates reduced in proportion to the absolute reduction in body weight, has been documented in animals; this mechanism of energy conservation also may be operative in persons with anorexia. The hypometabolism that occurs in anorexia nervosa reverses with refeeding. Resting energy expenditure and thermic response to food are decreased.33 This hypometabolism appears to be an appropriate mechanism to conserve energy. The low serum T3 levels in anorexia nervosa may be explained by an alteration in T4 conversion (see Chap. 30 and Chap. 36). In anorexia nervosa, as in starvation, the peripheral deiodination of T4 is diverted from the formation of the active T 3 to the production of the inactive reverse T3. Fasting decreases the hepatic uptake of T4, with a proportionate decrease in T3 production. (This “low T3 syndrome” rarely may mask hyperthyroidism.) The low serum T 4 value in some patients with anorexia nervosa is somewhat more difficult to explain. Low T4 euthyroidism has been seen in seriously ill patients in whom there is a dysfunctional state of deficient T4 binding with a normal availability of peripheral tissue sites for free T4. Presumably, a similar mechanism may be operative in anorexia nervosa. The secretion of thyroid-stimulating hormone (TSH) is normal, but an augmented serum TSH response to thyrotropin-releasing hormone (TRH) stimulation is seen, and the peak is delayed from 30 to between 60 and 120 minutes.3 This change may reflect an altered setpoint for endogenous TRH regulation. One study shows a subnormal response of T4 and T3 to TRH, with a normal TSH response, suggesting chronic understimulation of the thyroid as a consequence of hypothalamic hypothyroidism.34 In another study, the level of TRH in cerebrospinal fluid was found to be low.35 Hypothalamic-Adrenal Interrelations. Levels of serum cortisol often are elevated in anorexia nervosa; 24-hour studies demonstrate normal episodic and circadian rhythms of serum cortisol, but considerably higher levels36 (Fig. 128-3). This change, which also can be seen in malnutrition, is secondary to prolongation of the half-life of cortisol because of reduced metabolic clearance. The urinary levels of corticosteroids, including 17-hydroxysteroids and 17-ketosteroids, usually are low. The production rate of cortisol may be elevated; suppression with dexamethasone is inadequate, and the cortisol concentrations may exceed the binding capacity of cortisol-binding globulin.2 In addition, affinity of serum cortisol-binding globulin for cortisol is decreased. Thus, unbound serum cortisol may be increased, so that it becomes available to the tissues. Higher levels of cortisol might be expected to suppress corticotropin secretion. However, the circadian rhythm that is maintained at the higher serum cortisol levels suggests that a new setpoint has been determined by the hypothalamic–pituitary–adrenal axis. Work on the hypothalamic–pituitary–adrenal pathways suggests activation of this axis. Cortisol level is elevated and responses to corticotropin-releasing hormone (CRH) are abnormal. CRH is increased in the cerebrospinal fluid of patients with anorexia.37 CRH is known to suppress LH pulses in both humans and animals, and may augment dopaminergic and opiodergic inhibition of gonadotropin-releasing hormone.

FIGURE 128-3. Hourly mean serum cortisol level derived from average of samples obtained 20 minutes before, on, and after the hour in 10 patients with anorexia nervosa, compared with 6 normal controls matched for age and sex. Circadian rhythm of cortisol remains intact in anorexia nervosa but the level is higher. (SD, standard deviation.) (From Boyar RM, Hellman LD, Roffwang H, et al. Cortisol secretion and metabolism in anorexia nervosa. N Engl J Med 1977; 296:190.)

Miscellaneous hypothalamic abnormalities in anorexia nervosa include a deficiency in the handling of a water load (probably as a result of a mild diabetes insipidus, characterized by an erratic serum vasopressin response to osmotic challenge); abnormal thermoregulatory responses with exposure to temperature extremes; and a lack of shivering. Leptin Levels. An intriguing aspect of the disorder involves leptin, the hormone secreted by the adipocyte that seems to be a critical link between metabolic and reproductive pathways. Leptin appears to modulate food intake by affecting appetite, energy requirements, and eating behavior.38 Low leptin levels, which have been reported in hypothalamic amenorrhea,39,40 have been linked to low bone mass.40,41 Ob/ob rodents lacking an active form of leptin tend to be not only overweight but also amenorrheic and infertile.42,43 A critical leptin level exists below which menstruation stops.44 Leptin levels are reduced in both cycling and amenorrheic athletes, and a diurnal pattern of 24-hour leptin levels is strikingly absent in amenorrheic athletes.45 Anorexia nervosa patients consistently exhibit hypoleptinemia.46,48 and 49 A threshold effect for leptin occurs at low body weights. Leptin levels correlate linearly with body mass index in normal persons and recovering anorectic patients49a but uncouple in a nonlinear fashion in untreated anorectics.46 A strong correlation is found between leptin levels and absolute body fat mass; however, fasting behavior, regardless of body fat, may also be correlated with low leptin levels.50,51 A rapid, disproportional lowering of leptin levels often accompanies starvation.50,52 Anorectic patients, thus, may have three factors leading to hypoleptinemia: low body fat, excessive exercise, and fasting behavior itself. However, leptin levels normalize and can even peak above normal levels with incomplete weight gain.47,48 As leptin acts to reduce food intake, this premature restoration of leptin levels with only minor weight gain can be extremely detrimental to anorectic patients by actually perpetuating the disorder. OTHER PITUITARY HORMONE ABNORMALITIES Serum growth hormone levels are elevated in starvation or any other restriction of food beyond the normal 12- to 15-hour overnight fast. Generally, basal serum growth hormone levels are higher than normal in anorexia nervosa but respond normally to provocative stimuli. These high levels are associated with a decreased level of somatomedin C (i.e., insulin-like growth factor-I), as is also found in starvation. Occasionally, low serum growth hormone levels are seen, with blunted responses to insulin-induced hypoglycemia.6 Insulin-like growth factor is decreased, and both these abnormalities resolve with nutritional therapy. Nutritional deprivation may alter the growth hormone–insulin-like growth factor axis by down-regulation of growth hormone receptor or postreceptor.53 A comparison of normal and anorectic subjects found that the elevation of growth hormone secretion in anorexia nervosa may be due to an increased frequency of secretory pulses superimposed on heightened tonic growth hormone secretion. Thus, the elevation may not simply result from a malnutrition-induced impairment of insulin-like growth factor-I production. Instead, it may reflect a complex hypothalamic dysregulation of growth hormone release.54 Sleep in anorexia nervosa has been the subject of ongoing investigation.55 Growth hormone levels associated with delta-wave sleep are normal.6 A paradoxic growth hormone secretory response to the infusion of TRH has been observed in underweight persons in both anorexia nervosa and starvation. The basal serum prolactin levels are normal, and TRH-stimulated prolactin levels also are normal, although the time of the peak prolactin level is delayed. With recovery (weight gain), all of these endocrine changes normalize. However, despite the normalization of serum gonadotropin secretory patterns, amenorrhea persists in 30% of patients.6 ESTROGEN ABNORMALITIES Sonographic imaging of the ovaries of patients with anorexia nervosa demonstrates cystic involvement resembling that seen at adolescence.56,57 The low serum estradiol levels are partly attributable to the lack of ovarian stimulation. However, estrogen metabolism also is altered: the metabolism of estradiol, which normally proceeds with 16a-hydroxylation, is decreased in favor of 2-hydroxylation and the resulting formation of catechol estrogen (2-hydroxyestrone).2 This latter compound has features of an antiestrogen because it has no intrinsic bioactivity. Thus, the extraordinarily low serum estrogen levels sometimes seen in anorexia nervosa are compounded by an endogenously produced antiestrogen. Furthermore, the lack of adipose tissue may deny to patients extraovarian sources of estrogen: normally, fat converts androstenedione to estrone. NEUROENDOCRINE AND PSYCHOLOGICAL CHARACTERISTICS AND INTERRELATIONS The altered behavior patterns in anorexia nervosa are distinctive. Sometimes, periods of gorging alternate with food avoidance and starvation. Altered food intake combined with activity changes also may accompany hypothalamic tumors and other hypothalamic syndromes. Such changes also have been documented in rats with lesions of the ventromedial nucleus of the hypothalamus, a finding which led investigators to conclude that the ventromedial nucleus inhibits food intake and promotes activity—similar to the pattern seen in anorexia nervosa.3 Some of the changes of anorexia nervosa, including the lowered metabolic rate, a decrease in attainable oxygen uptake and VO2max, an increase in serum cortisol (which stimulates gluconeogenesis and decreases peripheral glucose utilization), and the diminished serum gonadotropin levels (with a consequent loss of fertility), are appropriate adaptations to starvation.3 Because of these neuroendocrine changes, the speculation had been that abnormalities of neurotransmission participate in the pathogenesis of the condition. In particular, excess dopamine and norepinephrine have known effects on behavior and appetite. In addition, b-endorphin affects feeding behavior; this hormone is thought to modulate the secretion of LHRH from the hypothalamus, and administration of its antagonist, naloxone, can restore LH pulsations in some persons with anorexia nervosa. Patients with anorexia nervosa and those with secondary amenorrhea due to a hypothalamic cause commonly share a need for achievement and approval. On fear-of-failure scales, patients with anorexia score the highest. Another early marker for anorexia nervosa is perceptual distortion. These patients consider themselves to be too fat despite their low weights. They consistently overestimate body size; this overestimation tends to disappear with weight gain. The perceptual distortion may reinforce dieting behavior, perpetuating the condition. Anorexic behavior scales have been said to aid in differentiating persons with anorexia nervosa from those with secondary amenorrhea of other causes. Often, the scale consists of a psychological profile that is either self-administered or administered by a physician, nurse, or clinical psychologist. One study indicated that this scale is useful and accurate in distinguishing healthy persons from those with anorectic behavior.58 Thus, it may prove useful in the early diagnosis of anorexia nervosa, particularly in patients with secondary amenorrhea only and little, if any, weight loss. OSTEOPOROSIS A leading complication of amenorrhea seen with anorexia and weight loss is osteopenia. This is present in spinal, radial, and femoral sites, and is associated with fractures.59,60,61,62,63,64 and 65 After 12 years, 67% of chronic anorectic patients suffered from medical comorbidity, most often involving osteoporosis and renal disease.66 Longitudinal studies on bone density show little or no reversal with resolution of the amenorrhea.60,63 These observations are important because hypoestrogenism in young adulthood may predispose to premature osteoporosis in later life. Increases in bone mass may occur in young persons before the return of normal menses, but bone mass remains below that of normal control subjects, possibly because of prolonged hypoestrogenism in adolescence, and, thus, may have permanent effects on peak bone mass.64 Loss of bone mass also may occur. The effects of estrogen replacement on these changes need to be studied to determine whether the trend toward decreased bone mass can be reversed by therapy. Unfortunately, estrogen replacement does not restore bone mass in patients with anorexia nervosa.67,68 In general, 25-hydroxyvitamin D, 1,25 dihydroxyvitamin D, and osteocalcin levels are normal, although osteocalcin levels may be depressed because of lower bone turnover.62 When reversal of osteoporosis occurs, it is seen more often with recovery from anorexia and is associated more tightly with weight gain than with return of menses. Some studies suggest, however, that bone mass does not appear to recover even with weight gain,68a calcium supplementation, and exercise. One study found that 14 out of 18 recovered anorexia nervosa patients exhibited osteopenia.69 The increased cortisol levels seen in anorexia have been suggested as a mechanism for the osteopenia.70

BULIMIA CHARACTERISTICS Bulimia usually is a condition of young women, often related to previous anorectic behavior. These persons gorge themselves and use unusual means to lose weight, such as vomiting, enemas, and abuse of laxatives or diuretics. Gorging episodes may alternate with periods of severe food restriction. Most commonly, this syndrome

occurs in high school and college students. Among males, bulimia is more common than is anorexia nervosa. According to the Diagnostic and Statistical Manual of Mental Disorder, Revised Third Edition, 4.5% of girls and women younger than 20 years of age have bulimia, whereas other sources estimate that at least 20% of this population exhibits bulimic behaviors.71 This syndrome may occur in 3% to 15% of university students.72,73 The weight may fluctuate, but usually not to dangerously low levels. The patient often has a history of other impulsive behavior, such as alcohol or drug use. Depression is common. Stealing and shoplifting, as well as unrestrained sexual promiscuity, may be part of the syndrome—unlike in anorexia nervosa, in which patients generally are sexually inactive. Depression and obsessive-compulsive behavior often coexist with eating disorders, particularly when bulimia is also present. Other comorbid problems include drug abuse and alcoholism.74 Patients with bulimia tend to be slightly older than those with anorexia, usually between 17 and 25 years. A separate condition, known as bulimia nervosa, has been described, in which bulimic behavior has evolved from the prior, more restrictive anorexia nervosa–type pattern. CLINICAL MANIFESTATIONS Persons with bulimia have a wide variety of superimposed medical problems, including vomiting-related tooth decay, parotid enlargement, stomach rupture, pancreatitis, metabolic alkalosis, and carpalpedal spasm. Occasionally, persons with bulimia have menstrual irregularities despite normal weight.6 Some patients are anovulatory, yet have adequate estrogen secretion. Evidence exists for hormonal defects in CRH and cholecystokinin regulation.75,76 Bulimic behavior often is secretive; many patients do not admit to these patterns even when questioned directly. The condition often is chronic, and increased anxiety, irritability, depression, and poor social functioning are common. The relapse rate for bulimia nervosa can be as high as 30%, although risk of relapse declines after approximately 4 years.77 Eventually, protein and calorie malnutrition may supervene and contribute to the development of the reproductive disorder. Several neurologic problems are associated with bulimia, including Huntington chorea and seizure disorders. Bulimia also may follow encephalitis and can be seen in association with the hypersomnia of Kleine-Levin syndrome (periodic attacks of hypersomnia and bulimia, often secondary to encephalitis, head injury, or hypothalamic tumor) and with parkinsonism; the latter patients improve in their eating patterns with treatment.3

TREATMENT OF ANOREXIA NERVOSA AND BULIMIA The treatment of anorexia nervosa and bulimia is controversial. All treatment modalities are directed toward the reestablishment of normal weight and eating habits. Treatment has included combinations of psychotherapy, including family therapy, psychoanalysis, and drug therapy. Tricyclic antidepressants, cyproheptadine, L -dopa, and metoclopramide have been used with variable success. Antidepressants may be the most successful, particularly for bulimia, in which drugs such as fluoxetine have a 30% success rate.78 Behavior modification has been attempted, again with variable efficacy. The early recognition of anorectic behavior is essential so that patients may be treated before the full-blown syndrome sets in. Because amenorrhea occurs early in the disease process, it often is the first symptom that causes patients to seek help. Thus, physicians should be particularly attentive to a history of dieting and weight loss in their young patients with amenorrhea. Patients who are 75% of ideal body weight or below need immediate and aggressive intervention. Dietary therapy is important because the response to psychotherapy is improved with nutritional rehabilitation. Usually, this is best accomplished in a hospital setting by a team consisting of a psychiatrist or a psychologist, an internist or a pediatrician, and, if possible, a nutritionist with special interest and expertise in anorexia nervosa and related eating disorders. Ninety percent of standard body weight may be a reasonable target weight for treatment, as it seems to be an average weight at which menses resume. Research shows that 86% of patients who reach this target resume menstruating within 6 months. Interestingly, the resumption does not depend on body fat, but rather on restoration of hypothalamic–pituitary–ovarian function.79 Considerable caution should be taken when treating severely malnourished patients. Rapid refeeding of patients who are 98% of patients with new-onset type 1A diabetes and prediabetes express at least one antibody, and >80% express two or more. Specificity and sensitivity are much higher with these biochemical assays than with cytoplasmic islet cell antibody testing. In contrast to cytoplasmic islet cell antibody testing, with its inherent problems of reproducibility, biochemical determination of autoantibodies is remarkably stable in the prediabetic phase. International workshops to standardize insulin and GAD radioassays are under way. Such assays should rapidly replace standard cytoplasmic islet cell antibody testing in both the diagnosis and prediction of type 1A diabetes. FIRST-PHASE INSULIN SECRETION AS AN INDEX OF EARLY TYPE 1 DIABETES MELLITUS Approximately 3% of nondiabetic relatives of patients with type 1A diabetes have positive results on screening assays for islet cell autoantibodies. When such antibodies are detected, intravenous glucose-tolerance testing can be used to assess first-phase insulin release as a measure of subclinical B-cell dysfunction (Fig. 136-4). The loss of first-phase insulin secretion, as well as its rate of fall, aids in predicting the time of onset of overt diabetes.47,48 At the time of initial detection of islet cell antibodies, one of four patients has first-phase insulin secretion below the first percentile of normal persons. Almost all persons who develop type 1A diabetes lose first-phase insulin secretion before they develop overt diabetes. For patients with initially normal insulin release, intravenous glucose-tolerance testing is performed again in 3 to 6 months and, depending on its stability, at subsequent 3- to 12-month intervals. Immunologically and endocrinologically, persons with abnormal results may be alerted to the risk of type 1A diabetes and advised concerning routine home monitoring for glucosuria or capillary blood glucose determination.

FIGURE 136-4. Loss of first-phase insulin secretion in a prediabetic twin with islet cell antibody. The y-axis gives insulin concentrations at the times indicated after the intravenous injection of glucose. Initial phase of insulin release (1 and 3 minutes) is progressively lost. (DM, diabetes mellitus; F, fasting.) (Adapted from Srikanta S, Ganda OP, Eisenbarth GS, Soeldner JS. Islet cell antibodies and beta cell function in monozygotic triplets and twins initially discordant for type I diabetes. N Engl J Med 1983; 308:322.)

To aid in predicting the time of onset of type 1A diabetes among antibody-positive relatives of persons with type 1A diabetes, the following mathematic formula has been developed: years to diabetes = –0.12 + 1.35 ×loge (insulin secretion) –0.59 loge (insulin autoantibody concentration). This simple formula appears to account for ~50% of the variance in the time of diabetes onset.48 Immunologic assays also are being used to aid in the classification of patients with diabetes. Insulin dependence is a physiologic state that can evolve slowly, even in type 1A diabetes, from a stage in which hyperglycemia is controlled with diet or oral medication to a stage in which death occurs in the absence of insulin therapy (Fig. 136-5). Antiislet antibodies are found in as many as 10% of patients with classic type 2 diabetes at the time of diagnosis, and over the ensuing 5 years, many of these patients become insulin dependent.35 Epidemiologic data from Japan, Pittsburgh, the Netherlands, and Poland indicate that 1 in 200 children die of ketoacidosis at the time of diagnosis of their diabetes. Such deaths probably could be prevented by early detection and treatment of diabetes.

FIGURE 136-5. Stages in the development of type 1 diabetes, beginning with genetic predisposition and ending with insulin-dependent diabetes with essentially complete B-cell destruction. (Adapted from Eisen-barth GS. Type I diabetes mellitus: a chronic autoimmune disease. N Engl J Med 1986; 314:1360.)

IMPLICATIONS FOR PREVENTIVE THERAPY The newer immunologic knowledge concerning type 1A diabetes and the success of a wide variety of immunotherapies in preventing the diabetes of BB rats and NOD mice have led to trials of immunotherapy in patients with recent-onset type 1A diabetes, and to a few trials in persons at high risk for the development of type 1A diabetes. These trials indicate that limiting B-cell destruction is possible. Toxic side effects of the most powerful (and most effective) drugs are a serious problem, however. For example, cyclosporine A is nephrotoxic at the dosages that appear to be required to induce remissions of type 1A diabetes,49,50 and azathioprine may be associated with epithelial malignancies. Prednisone given after the onset of type 1A diabetes is ineffective, and a series of other therapies, such as plasmapheresis and treatment with antithymocyte globulin or monoclonal antibody T12, produce no long-term benefit. A series of new agents that are not significantly immunosuppressive yet are able to limit B-cell destruction in animal models of type 1A diabetes are being studied. All immunologic attempts to limit B-cell destruction are considered investigational and should be used only under the oversight of a human investigation committee.51 Trials investigating the prevention of type 1A diabetes and the amelioration of further B-cell loss after diabetes onset are concentrating on nonimmunosuppressive therapies. Such trials include administration of the vitamin nicotinamide,52 which modestly delays diabetes onset in NOD mice; vaccination with bacille Calmette-Guérin53; administration of oral insulin54; and therapy with parenteral insulin.55 Nicotinamide may have an effect by limiting free radical damage to B cells. In the NOD mouse model, a single injection of bacille Calmette-Guérin (BCG) prevents diabetes but not insulitis. Such therapy may limit B-cell destruction by altering cytokines produced by infiltrating T cells. A completed German trial of nicotinamide found no benefit, but a larger European trial is continuing. In randomized trials, BCG had no beneficial effect. Oral administration of insulin delays or prevents type 1 diabetes in NOD mice. Its effect is likely due to the generation of T cells (by peptides of insulin present in the intestinal mucosa) that suppress inflammation. The most dramatic prevention of diabetes in both the NOD mouse and the BB rat has been obtained with parenteral administration of insulin. Such therapy prevents not only diabetes, but also infiltration of islets by T cells and destruction of B cells. A small pilot trial53 of intravenous insulin and low-dose subcutaneous insulin for the prevention of diabetes in high-risk relatives of patients with diabetes suggests that such therapy may delay the onset of type 1 diabetes, and a large U.S. prevention trial (DPT-1 [Diabetes Prevention Trial-1]) is under way. AUTOIMMUNE POLYGLANDULAR FAILURE Approximately 20% of patients with type 1A diabetes develop other organ-specific autoimmune diseases (see Chap. 197), such as celiac disease, Graves disease, hypothyroidism, Addison disease, and pernicious anemia.56,57 Some patients develop multiple disorders as a part of two inherited polyendocrine autoimmune syndromes (type I and type II). The type I syndrome usually has its onset in infancy, with hypoparathyroidism, mucocutaneous candidiasis, and, somewhat later, Addison disease and other organ-specific disorders. Fifteen percent of these children develop type 1A diabetes. This disorder is inherited in an autosomal recessive manner with no HLA association due to mutations of a gene (AIRE) located on chromosome 21. This mutated gene codes for a DNA-binding protein that is expressed in the thymus. The type II polyendocrine autoimmune syndrome (Addison disease, type 1 diabetes [50% of patients], Graves disease, hypothyroidism, myasthenia gravis, and other organ-specific diseases) is strongly HLA-associated and has an onset from late childhood to middle age. In these families, a high prevalence of undiagnosed organ-specific auto-immune disease is seen, and, at a minimum, thyroid function tests should be performed in the first-degree relatives of these patients. Biochemical evaluation for adrenal insufficiency and pernicious anemia should be performed if any suggestive symptom or sign is present (e.g., decreasing insulin requirements can herald the development of Addison disease in a patient with type 1 diabetes before electrolyte abnormalities or hyperpigmentation develop). Excellent autoantibody tests can facilitate the detection of Addison disease (21-hydroxylase autoantibodies) or celiac disease (transglutaminase autoantibodies) in patients with type 1A diabetes. As many as 1 in 200 patients with type 1A diabetes develop Addison disease, and 1 in 20 develop celiac disease. CHAPTER REFERENCES 1. Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care 1997; 20:1183. 2. World Health Organization Expert Committee on Diabetes Mellitus. Second report. Geneva: World Health Organization, 1980:646. Technical report series 1980. 3. Marble A, Ferguson BD. Diagnosis and classification of diabetes mellitus and the non-diabetic melliturias. In: Marble A, Krall LP, Bradley RF, et al., eds. Joslin's diabetes mellitus, 12th ed. Philadelphia: Lea & Febiger, 1985:332. 4. Polonsky KS, Stuns J, Bell GI. Non-insulin-dependent diabetes mellitus: a genetically programmed failure of the beta cell to compensate for insulin resistance. N Engl J Med 1996; 334:777. 5. Nelson RL. Oral glucose tolerance test: indications and limitations. Mayo Clin Proc 1988; 63:263. 6. Ganda OP, Srikanta S, Brink SJ, et al. Differential sensitivity to beta cell secretagogues in “early” type I diabetes. Diabetes 1984; 33:516. 7. Bergman RN, Finegood DJ, Ader M. Assessment of insulin sensitivity in vivo. Endocr Rev 1985; 6:45. 8. Fajans SS, Conn JW. An approach to the prediction of diabetes mellitus by modification of the glucose tolerance test with cortisone. Diabetes 1954; 3:296. 9. Koenig RJ, Cerami A. Hemoglobin A1c and diabetes mellitus. Annu Rev Med 1980; 31:29. 10. Starkman HS, Soeldner JS, Gleason RE. Oral glucose tolerance—relationship with hemoglobin A 1c. Diabetes Res Clin Pract 1987; 6:343. 11. Garg SK, Potts RO, Ackerman NR, et al. Correlation of fingerstick blood glucose measurements with Gluco Watch biographer glucose results in young subjects with type 1 diabetes. Diabetes Care 1999; 22:1708. 12. Agner E, Thorsteinssen B, Eriksen M. Impaired glucose tolerance and diabetes mellitus in elderly subjects. Diabetes Care 1982; 5:600. 13. Gepts W. The pathology of the pancreas in human diabetes. In: Adreani D, Federlin KF, DiMario U, Heding LG, eds. Immunology in diabetes. London: Kimpton Publishers, 1984:21. 14. Gepts W. Islet cell morphology in type I and type II diabetes. In: Irvine WJ, ed. Immunology in diabetes. Edinburgh: Teviot Scientific Publications, 1982:255. 15. Eisenbarth GS. Type I diabetes mellitus: a chronic autoimmune disease. N Engl J Med 1986; 314:1360. 16. Makino S, Harada M, Kishimoto Y, Hayashi Y. Absence of insulitis and overt diabetes in athymic nude mice with NOD genetic background. Jikken Dobitsu Exp Anim 1986; 35:495. 17. Inoue H, Tanizawa Y, Wasson J, et al. A gene encoding a transmembrane protein is mutated in patients with diabetes mellitus and optic atrophy (Wolfram syndrome). Nat Genet 1998; 20:143. 18. Mordes JP, Greiner DL, Rossini AA. Animal models of autoimmune diabetes mellitus. In LeRoith D, Taylor SI, Olefsky JM, eds. Diabetes mellitus. Philadelphia: Lippincott–Raven Publishers, 1996:349. 19. Feingold KR, Lee TH, Chug MY, et al. Muscle capillary basement membrane width in patients with Vacor-induced diabetes mellitus. J Clin Invest 1986; 78:102. 20. Jackson RA, Buse JB, Rifai R, et al. Two genes required for diabetes in BB rats. J Exp Med 1984; 159:1629. 21. Nepom GT, Kwok WW. Perspectives in diabetes: molecular basis for HLA-DQ associations with IDDM. Diabetes 1998; 47:1177. 22. Todd JA. From genome to aetiology in a multifactorial disease, type 1 diabetes. Bioessays 1999; 21:164. 23. Bellgrau D, Eisenbarth GS. Immunobiology of autoimmunity. In: Eisen-barth GS, ed. Molecular mechanisms of endocrine and organ specific autoimmunity. Austin: RG Landes Company, 1991:1. 24. Redondo MJ, Kawasaki E, Mulgrew CL, et al. DR and DQ associated protection from type 1 diabetes: comparison of DRB1*1401 and DQA1*0102-DQB1*0602. J Clin Endocrinol Metab 2000; In press. 25. Nepom GT. Immunogenetics and IDDM. Diabetes Rev 1993; 1:93. 26. Kawasaki E, Noble J, Erlich H, et al. Transmission of DQ haplotypes to patients with type 1 diabetes. Diabetes 1998; 47:1971. 27. Erlich HA, Griffith RL, Bugawan TL, et al. Implication of specific DQB1 alleles in genetic susceptibility and resistance by identification of IDDM siblings with novel HLA-DQB1 allele and unusual DR2 and DR1 haplo-types. Diabetes 1991; 40:478. 28. Greenbaum CJ, Cuthbertson D, Eisenbarth GS, et al. Islet cell antibody positive relatives with HLA-DQA1*0102, DQB1*0602: Identification by the Diabetes Prevention Trial-1. J Clin Endocrinol Metab 2000; 85:1255. 29. Bennett ST, Lucassen AM, Gough SCL, et al. Susceptibility to human type I diabetes at IDDM2 is determined by tandem repeat variation at the insulin gene minisatellite locus. Nat Genet 1995; 9:284. 30. Barnett AH, Eff C, Leslie RDG, Pyke DA. Diabetes in identical twins: a study of 200 pairs. Diabetologia 1981; 20:87. 31. Blom L, Dahlquist G, Nystrüm L, et al. The Swedish childhood diabetes study—social and perinatal determinants for diabetes in childhood. Diabetologia 1989; 32:7. 32. Menser MA, Forrest JM, Brensby RD. Rubella infection and diabetes mellitus. Lancet 1981; 1:57. 33. Yoon JW, London WT, Curfman BL, et al. Coxsackie virus B4 produces transient diabetes in non-human primates. Diabetes 1986; 35:712. 34. Bingley PJ, Gale EAM. Current status and future prospect for prediction of IDDM. In: Palmer JP, ed. Prediction, prevention, and genetic counseling in IDDM. Chichester, England: John Wiley, 1996:227. 35. Turner R, Stratton I, Horton V, et al. UKPDS 25: autoantibodies to isletcell cytoplasm and glutamic acid decarboxylase for prediction of insulin requirement in type 2 diabetes. UK Prospective Diabetes Study Group. Lancet 1997; 30:1288. 36. Gorsuch AN, Spencer KM, Lister J, et al. Evidence for a long prediabetic period in type I (insulin-dependent) diabetes mellitus. Lancet 1981; 2:1363. 37. Palmer JP, Asplin CM, Clemons P, et al. Insulin antibodies in insulin-dependent diabetics before insulin treatment. Science 1983; 222:1337. 38. Vardi P, Tuttleman M, Grinbergs M, et al. Consistency of antiislet autoimmunity in “pretype I diabetics” and genetically susceptible subjects: evidence from an ultrasensitive competitive insulin autoantibody (CIAA) radioimmunoassay. Diabetes 1986; 35(Suppl 1):86A. 39. Gale EAM. Islet cell autoantibodies: a family story. Eur J Endocrinol 1996; 135:643. 40. Kuglin B, Bertrams J, Linke C, et al. Prevalence of cytoplasmic islet cell antibodies and insulin autoantibodies is increased in subjects with genetically defined high risk for insulin-dependent diabetes mellitus. Klin Wochenschr 1989; 67:66. 41. Radetti G, Paganini C, Gentili L, et al. Frequency of Hashimoto's thyroiditis in children with type 1 diabetes mellitus. Acta Diabetol 1995; 32:121. 42. Baekkeskov S, Aanstoot H, Christgau S, et al. Identification of the 64K autoantigen in insulin dependent diabetes as the GABA-synthesizing enzyme glutamic acid decarboxylase. Nature 1990; 347:151. 43. Castano L, Russo E, Zhou L, et al. Identification and cloning of a granule autoantigen (carboxypeptidase H) associated with type I diabetes. J Clin Endocrinol Metab 1991; 73:1197. 44. Pietropaolo M, Castano L, Babu S, et al. Islet cell autoantigen 69 kDa (ICA69): molecular cloning and characterization of a novel diabetes associated autoantigen. J Clin Invest 1993; 92:359. 45. Verge CF, Stenger D, Bonifacio E, et al. Combined use of autoantibodies (IA-2ab, Gadab, IAA, ICA) in type 1 diabetes: combinatorial islet autoantibody workshop. Diabetes 1998; 47:1857. 46. Nayak RC, Omar MAK, Rabizadeh A, et al. “Cytoplasmic” islet cell antibodies: evidence that the target antigen is asialoglycoconjugate. Diabetes 1985; 34:617. 47. Srikanta S, Ganda OP, Soeldner JS, Eisenbarth GS. First-degree relatives of patients with type I diabetes: islet cell antibodies and abnormal insulin secretion. N Engl J Med 1985; 313:461.

48. 49. 50. 51. 52. 53. 54. 55. 56. 57.

Eisenbarth GS, Gianani R, Yu L, et al. Dual parameter model for prediction of type 1 diabetes mellitus. Proc Assoc Am Physicians 1998; 110:126. Stiller CR, Dupré J, Gent M, et al. Effect of cyclosporine immunosuppression in insulin-dependent mellitus of recent onset. Science 1984; 223:1362. Feutren G, Asson G, Karsenty G, et al. Cyclosporine increases the rate and length of remissions in insulin-dependent diabetes of recent onset: results of a multi-center trial. Lancet 1986; 2:119. Gottlieb PA, Eisenbarth GS. Diagnosis and treatment of pre-insulin dependent diabetes (IDDM). Annu Rev Med 1998; 49:391. Lampeter EF, Klinghammer A, Scherbaum WA, et al. The Deutsche Nicotin-amide Intervention Study: an attempt to prevent type 1 diabetes. DENIS Group. Diabetes 1998; 47:980. Sadelain MWJ, Qin H-Y, Lauzon J, Singh B. Prevention of type I diabetes in NOD mice by adjuvant immunotherapy. Diabetes 1990; 39:583. Zhang JZ, Davidson L, Eisenbarth GS, Weiner HL. Suppression of diabetes in nonobese diabetic mice by oral administration of porcine insulin. Proc Natl Acad Sci U S A 1991; 88:10252. Keller RJ, Eisenbarth GS, Jackson RA. Insulin prophylaxis in individuals at high risk of type I diabetes. Lancet 1993; 341:927. Nuefeld M, Maclaren N, Blizzard RM. Two types of autoimmune Addison's disease associated with different polyglandular autoimmune (PGA) syndromes. Medicine (Baltimore) 1981; 60:355. Verge C, Eisenbarth GS. Autoimmune polyendocrine syndromes. In: Wilson JD, Foster DW, eds. Williams textbook of endocrinology, 9th ed. Philadelphia: WB Saunders, 1998:1651.

CHAPTER 137 ETIOLOGY AND PATHOGENESIS OF TYPE 2 DIABETES MELLITUS AND RELATED DISORDERS Principles and Practice of Endocrinology and Metabolism

CHAPTER 137 ETIOLOGY AND PATHOGENESIS OF TYPE 2 DIABETES MELLITUS AND RELATED DISORDERS C. RONALD KAHN Non–Insulin-Dependent (Type 2) Diabetes Mellitus Pathogenesis Insulin Synthesis and Secretion Insulin Resistance Events in the Development of Type 2 Diabetes Genetics of Type 2 Diabetes Related Conditions Impaired Glucose Tolerance Gestational Diabetes Mellitus Maturity-Onset Diabetes of Youth Nondiabetic Mellituria Renal Glucosuria Nonglucose Mellituria Hereditary Fructose Intolerance Galactosemia Essential Fructosuria Essential Pentosuria Chapter References

NON–INSULIN-DEPENDENT (TYPE 2) DIABETES MELLITUS Type 2 diabetes mellitus is by far the most prevalent endocrine disease. It is estimated to affect >15 million people in the United States, approximately one-third of whom are undiagnosed.1,2 The prevalence increases with age, and >9% of those older than 65 years have the disease. The prevalence is higher in Mexican Americans, blacks, and Native Americans, reaching as high as 50% among adult Pima Indians. Development of type 2 diabetes is strongly influenced by genetic factors and environmental factors, including obesity, decreased physical activity, and a low level of physical fitness.3 PATHOGENESIS Although type 2 or non–insulin-dependent diabetes mellitus is the more common form of the disease, the exact nature of its pathogenesis remains controversial, and in contrast to type 1 diabetes, in which immunologic markers confirm the pathogenesis of the disease, no specific diagnostic tests are available for type 2 diabetes. Type 2 diabetes has strong genetic influences and occurs in identical twins with almost total concordance4,5; however, defining the exact genes involved has posed a great challenge. In type 2 diabetes, B-cell mass is relatively well preserved, but insulin secretion in response to specific secretagogues such as glucose is reduced, and clear evidence exists for resistance to insulin action in the peripheral tissues.6,7 and 9 The precise genetic defects have been identified in some of the minor forms of type 2 diabetes, including those in rare patients with genetic syndromes of insulin resistance (see Chap. 146) and in patients with maturity-onset diabetes of youth (MODY). For the common form of type 2 diabetes, controversy continues over whether the decreased insulin secretion or insulin resistance is the principal factor in the pathogenesis of the disease, and which occurs first in the longitudinal development of the syndrome. Uncertainty also exists about the extent of the heterogeneity and the identity of the primary lesion. INSULIN SYNTHESIS AND SECRETION Even before the introduction of radioimmunoassays, morphologic studies suggested that the pancreas of a patient with type 2 diabetes has at least 50% (and sometimes up to 100%) of the normal B-cell mass, whereas that of a patient with type 1 diabetes of more than a few years' duration has virtually no B cells.6 This finding is consistent with the data on extractable insulin as measured in bioassay and radioimmunoassay. Immunoassay of plasma insulin and of C peptide has confirmed the presence of functioning B cells in patients with this disease, but the extent of function varies considerably depending on the type of stimulus used, the body weight of the patient, and the stage of disease.8,9 and 10 In long-standing disease, islets do produce amyloid deposits, which are partly comprised of a second hormone (termed islet amyloid polypeptide or amylin), which is secreted by B cells.10 Although some studies have implicated this peptide in the abnormal B-cell function of type 2 diabetes, this hypothesis remains unproven, and it seems more likely that the amyloid deposits are a marker of chronic insulin hypersecretion. In most individuals with type 2 diabetes, basal insulin levels are normal or elevated, and the degree of elevation correlates with the degree of obesity.11 Elevated insulin levels also occur in thin individuals with type 2 diabetes. Although a few patients have been identified with a defect in the conversion of proinsulin to insulin or mutant insulin molecules, in most patients with type 2 diabetes, the proportion of proinsulin to insulin is normal or only slightly increased, and the insulin and proinsulin have normal receptor binding and bioactivity. Although basal insulin levels usually are elevated in patients with type 2 diabetes, the insulin secretory responses to oral glucose differ considerably, depending on the extent of glucose intolerance. In patients with normal fasting glucose levels and 2-hour postprandial levels of 200 mg/dL.7,8

FIGURE 137-1. Relation between insulin secretion and glucose level 2 hours after a glucose load. Data were obtained for Pima Indians with various degrees of glucose intolerance. Similar results have been observed in whites. (Data from Savage PJ, Dippe SE, Bennett PH, et al. Hyperinsulinemia and hypoinsulinemia: insulin responses to oral carbohydrate over a wide spectrum of glucose tolerance. Diabetes 1975; 24:262.)

In contrast to the relative preservation of insulin response to meals and to oral glucose, a loss of acute-phase (first-phase) insulin release in response to intravenous glucose occurs in virtually all patients with significant fasting hyperglycemia.9 Acute insulin release in response to b-adrenergic stimuli, amino acids, and other insulin secretagogues in these same individuals often is normal, suggesting a specific defect in glucorecognition rather than a general defect in B-cell function.12,13 This pattern of response resembles that seen in patients in the early, preclinical phase of type 1 diabetes. In studies of perfused pancreases from animal models of type 2 diabetes, although first-phase secretion may be lost, glucose maintains its ability to potentiate arginine-induced insulin secretion.14 The preservation of response to an oral

glucose load probably is a reflection of the importance in this response of the potentiation of the glucose effect by gastrointestinal hormones. In humans with type 2 diabetes, however, the glucose potentiation is also blunted.12 The lost first-phase response to glucose can be restored, at least partially, by a-adrenergic blockade, opiate-receptor blockade, inhibitors of prostaglandin synthesis, reduction of plasma glucose by dietary restriction, use of oral hypoglycemic agents, or even administration of insulin.15 The fact that this lesion is functional and at least partially reversible makes it potentially amenable to therapeutic manipulation. A fragment of glucagon-like peptide-1 (GLP-1) may potentiate glucose-induced insulin secretion in persons with type 2 diabetes; this offers a possible new avenue for therapy that is currently being explored in clinical trials.16 INSULIN RESISTANCE Virtually all patients with type 2 diabetes have some degree of insulin resistance.17 Conditions associated with the development of insulin resistance, especially obesity and advancing age, greatly increase the risk of type 2 diabetes. Insulin resistance correlates with certain patterns of obesity and is greater in individuals with central obesity than in those with more generalized obesity.18 At any given body weight, a high waist-to-hip ratio correlates with insulin resistance and increased risk of type 2 diabetes. Insulin resistance and hyperinsulinemia are also associated with hypertension and hypertriglyceridemia, deceased high-density lipoprotein cholesterol, and increased risk of atherosclerosis and cardiovascular disease.19,20 The association of insulin resistance with these features in the absence of clinical diabetes has been referred to as the metabolic syndrome or syndrome X (see Chap. 145).19 In cases of type 2 diabetes and syndrome X, insulin resistance is suggested by the elevated insulin levels and the fact that the patient develops glucose intolerance with circulating insulin levels well above those seen in the type 1 diabetic. The simplest test of insulin sensitivity is the measurement of the fall in plasma glucose in response to a given dose of exogenous insulin (i.e., an insulin-tolerance test). In normal individuals, glucose usually falls by >50% in response to a dose of 0.05 to 0.1 U per kilogram of body weight. In type 2 diabetics, this response may be markedly blunted, and as much as 0.3 U/kg may be required to produce a 50% fall in glucose. Because the variability of endogenous insulin secretion and the counterregulatory hormones released during hypoglycemia may modify the response to the insulin test, more sophisticated measures of insulin resistance have been developed in which these factors are minimized. These tests include the measurement of steady-state glucose during simultaneous insulin and glucose infusion in which pancreatic insulin is suppressed with propranolol and epinephrine or somatostatin or the use of the euglycemic insulin clamp technique.21 In the latter, dose-response curves for the effect of insulin on glucose disposal in normal humans can be constructed (Fig. 137-2). When this test is used, the nature of the insulin resistance can be dissected into changes in median effective dose (ED50) on dose-response curves (i.e., changes in insulin sensitivity) and changes in maximal insulin effect (i.e., changes in insulin responsiveness).21,22 In type 2 diabetes, decreased sensitivity in insulin action to decreased splanchnic glucose output (i.e., an effect primarily on the liver) and decreased sensitivity and decreased responsiveness of insulin action to increased glucose utilization (i.e., effects primarily on muscle and fat tissue) are seen.21,23

FIGURE 137-2. Glucose disposal during a euglycemic clamp in normal patients and in those with type 2 diabetes. Predicted level of insulin resistance, assuming the only defect to be that of insulin binding, was compared with the observed data. The observed data indicate more severe insulin resistance and thus suggest the presence of a postbinding defect as well. (Redrawn from data of Scarlett JA, Gray RS, Griffin J, et al. Insulin treatment reverses the insulin resistance of type II diabetes mellitus. Diabetes Care 1982; 5:353.)

The resistance to insulin that occurs in the patient with type 2 diabetes mellitus could result from defects at several levels of insulin action (see Chap. 135). To produce a signal at the target cell, insulin must bind to its receptor, generate a transmembrane signal by activation of the insulin-receptor kinase, and initiate a complex network of intracellular signals that ultimately culminate in the activation and inhibition of the different cellular processes responsible for the physiologic effects of insulin. Studies of tissues taken from animal models of type 2 diabetes, as well as biopsies of tissues from humans with the disease, have revealed multiple alterations, including defects at the insulin receptor (e.g., binding and kinase activation) and at several of the postreceptor steps involved in insulin action. Decreased insulin-receptor binding has been described in obese and thin individuals with type 2 diabetes23,24 (Fig. 137-3). This decrease in binding is attributable to a decrease in receptor number, with no changes in receptor affinity, and is thought to be secondary to down-regulation of the receptor by the elevated basal endogenous insulin level.16,24 Similar decreases in insulin binding are observed in patients with impaired glucose tolerance and in some obese individuals with normal glucose tolerance. These findings indicate that the decrease in insulin receptors alone probably does not entirely account for the insulin resistance. Because “spare” receptors for insulin action are present in many tissues, a decrease in receptors would be expected to produce only a shift in the dose-response curve, with no changes in maximal response (i.e., decreased sensitivity).22

FIGURE 137-3. Defect in insulin-receptor binding in diabetes and obesity. The negative relation between receptor concentration and plasma insulin concentration is illustrated. Data are plotted as insulin binding versus plasma insulin concentration. Shaded areas indicate the normal range. Data on the left are from thin and obese patients with type 2 diabetes and data on the right are from obese individuals with and without diabetes. (Data for diabetic patients from Kahn CR. Insulin action, diabetogenes, and the cause of type II diabetes [Banting lecture]. Diabetes 1994; 43[8]:1066; and from Ferrannini E, Mari A. How to measure insulin sensitivity. J Hypertens 1998; 16[7]:895. Data for obese patients from Caro JF, Sinha MK, Raju SM, et al. Insulin receptor kinase in human skeletal muscle from obese subjects with and without non–insulin-dependent diabetes. J Clin Invest 1987; 79:1330.)

As noted previously, euglycemic clamp studies have indicated that both decreased sensitivity and decreased responsiveness to insulin are present, indicating postbinding defects in insulin action.23 These include defects in activation of the insulin-receptor kinase and phosphorylation of intracellular substrates.23,25,26,27 and 28 Sequence variations have been identified in several of the proteins involved in insulin signaling, and in the case of insulin-receptor substrate-1 (IRS-1), these occur with increased frequency in patients with type 2 diabetes.29 Most studies have indicated that a defect in glucose transport is also present due to a defect in glucose transporter translocation30,31 (Fig. 137-4). Finally, a defect also is present in glycogen synthesis, which forms the major component of nonoxidative glucose metabolism.6,31 Which of these defects is primary and whether other defects are present in the intracellular steps of insulin action remain unknown; however, some of these alterations can be found in normoglycemic offspring of parents with type 2 diabetes.31

FIGURE 137-4. Defect in insulin-receptor kinase activity in type 2 diabetes, in which the activity of this enzyme in liver is decreased by ~50%, even when expressed per receptor. (NIDDM, non–insulin-dependent diabetes mellitus.) (Redrawn from Caro JF, Ittoop O, Pories WJ, et al. Studies in the mechanism of insulin resistance in the liver from humans with non–insulin-dependent diabetes. J Clin Invest 1986; 78:249.)

Studies in both animal models and cell culture have demonstrated that a number of defects in insulin signaling may be secondary to different metabolic abnormalities present in the type 2 diabetic patient. For example, prolonged exposure to high insulin levels produces postbinding desensitization and receptor down-regulation, suggesting a role for increased basal insulin levels in these defects.32 The increased levels of tumor necrosis factor-a (TNF-a) and free fatty acids (FFAs), which are also found in obesity, produce insulin resistance by inhibiting the insulin-receptor kinase and by altering postreceptor metabolism, respectively.33,34 and 35 Hyperglycemia, both directly and through the production of intermediates like glucosamine, can also contribute to increasing levels of insulin resistance.36 Thus, like the defects in insulin secretion, the defects in insulin action are largely reversible when the diabetes is treated and the metabolic abnormalities are corrected. This is true whether the treatment involves diet, oral hypoglycemic agents, or intensive insulin treatment.36,37 and 38 A schematic diagram illustrating the various factors involved in the pathogenesis of type 2 diabetes is shown in Figure 137-5.

FIGURE 137-5. Schema for pathogenesis of type 2 diabetes, illustrating insulin resistance to glucose uptake at muscle and fat and failure of insulin to suppress hepatic glucose output, coupled with a defect in glucose sensing at the B cell.

EVENTS IN THE DEVELOPMENT OF TYPE 2 DIABETES Like type 1 diabetes, type 2 diabetes is preceded by a long pre-diabetic phase in which glucose-tolerance tests are normal, but insulin resistance is present17,39 and 40 (Fig. 137-6). Many patients also pass through a stage of impaired glucose tolerance in which basal and stimulated insulin levels are increased, findings which further suggest that insulin resistance precedes the functional insulin deficiency. At this stage, decreases in insulin-receptor binding and insulin action in muscle and fat can be detected. Fasting hyperglycemia suggests unsuppressed gluconeogenesis and indicates further resistance to insulin action at the liver. These patients also have a significant functional defect in insulin secretion, especially in glucose recognition.12,13

FIGURE 137-6. Model of the progressive pathogenesis of type 2 diabetes mellitus.

GENETICS OF TYPE 2 DIABETES Although type 2 diabetes has a very strong genetic influence, the genes leading to development of this disease are poorly understood. In most patients, type 2 diabetes probably is polygenic in nature, and the polygenes involved may be different among different families or population groups. However, a small number of patients show a monogenic form of type 2 diabetes. These include patients with maturity-onset diabetes of youth (MODY; described below), a few families in whom genetic defects in the insulin molecule lead to generation of a mutant insulin or a failure to process proinsulin and hyperproinsulinemia,41,42 patients with defects in mitochondrial DNA, and patients with genetic defects in the insulin receptor. In each of these cases, the resulting syndrome usually has clinical features that are distinct from typical type 2 diabetes. Genetic defects in the insulin receptor have been described in at least 100 individuals, virtually all with different mutations in the gene. Most of these patients have syndromes of severe insulin resistance (e.g., leprechaunism, Rabson-Mendenhall syndrome, or the type A syndrome of insulin resistance and acanthosis nigricans).43,44 These patients are discussed in Chapter 146. A variant of type 2 diabetes associated with a point mutation in the gene encoding the transfer RNA for leucine has been described. This gene is present in mitochondrial DNA rather than in nuclear DNA. Because mitochondrial DNA is inherited almost exclusively from the mother, this form of diabetes is characterized by a maternal inheritance pattern.45,46 This condition of diabetes is also associated with nerve deafness and some somatic defects, and may represent as much as 0.8% of type 2 diabetes in some populations. RELATED CONDITIONS

IMPAIRED GLUCOSE TOLERANCE Impaired glucose tolerance (IGT) is present in 7% to 11% of the population, depending on the diagnostic criteria used and the group studied. Recommendations by the American Diabetes Association Expert Committee on the Diagnosis and Classification of Diabetes Mellitus suggested that this diagnosis be based simply on a fasting glucose level of 110 to 125 mg/dL.47 Many of the characteristics of this population, other than the level of hyperglycemia, are similar to those of patients with type 2 diabetes, suggesting that IGT may be an intermediate step in the development of overt type 2 diabetes.47,48 IGT may also be a component of syndrome X. The presence of IGT significantly increases the risk of subsequent development of type 2 diabetes, although in many individuals, IGT is transient or does not progress. Researchers at the National Institutes of Health are about to undertake a major study to determine if the progression of IGT to type 2 diabetes can be prevented by changes in lifestyle or pharmacologic intervention. GESTATIONAL DIABETES MELLITUS Gestational diabetes mellitus is defined as the development of diabetes during pregnancy in a woman with no previous history of disease (see Chap. 156).49,50,51 and 52 Gestational diabetes may occur in 2% to 5% of all pregnancies. Although some of these cases represent the coincidence of detection of type 1 or type 2 diabetes with pregnancy, in most patients with true gestational diabetes, the hyperglycemia disappears after delivery. These women have an increased risk of developing type 2 diabetes, which is usually estimated at 2% to 3% per year of follow-up.49,50 Pathophysiologically, women with gestational diabetes have insulin resistance, as measured in a euglycemic clamp, and decreased early insulin response to intravenous glucose.51 These changes are similar to those observed in type 2 diabetes and may reflect an unmasking of the diabetic state by the hormonal milieu of pregnancy, particularly the increased levels of placental lactogen. MATURITY-ONSET DIABETES OF YOUTH From a genetic perspective, the best-characterized subset of patients with type 2 diabetes are those with MODY.53,53a,54,55,56,57,58 and 59 As implied by the name, this form is characterized by an early age of onset and a strong family history, with affected members in at least three generations, suggesting an autosomal dominant mode of inheritance. Although the typical patient in these kindreds is first found to be diabetic in the teens or 20s, some may be diagnosed as early as 5 years of age. As in other forms of type 2 diabetes, the level of hyperglycemia increases gradually, allowing treatment with diet or sulfonylureas, although in the latter stages of disease some patients require insulin for control of blood glucose. These patients develop the typical chronic complications of diabetes mellitus, including retinopathy, nephropathy, and neuropathy. Pathophysiologically, the disorder is heterogeneous, but in many families, the major defect appears to be relatively low insulin secretion. Genetic studies have revealed at least five subtypes of MODY. These have been designated MODY1 through MODY5, based on the order in which the genetic loci were identified. All five forms of MODY thus far identified are due to genetic defects impairing B cell function. The most common form of MODY in the United States and in most European countries is MODY3, which is present in 25% to 50% of families that meet the clinical criteria for MODY. This form is due to genetic defects in the nuclear transcription factor hepatic nuclear factor-1a (HNF-1a), a gene that plays a major role in control of insulin synthesis and B-cell function.53,54 From 10% to 40% of the kindreds have MODY2, in which the molecular abnormalities are genetic defects in the enzyme glucokinase.55,56 Glucokinase is the major form of hexokinase, the enzyme responsible for phosphorylation of glucose to glucose-6-phosphate in the liver and in islets of Langerhans. In the latter site, the enzyme is one of the rate-limiting steps in glucose sensing by the B cell. In comparison with other forms of type 2 diabetes, MODY2 is usually characterized by relatively mild degrees of hyperglycemia and glucose intolerance. MODY1, MODY4, and MODY5 are also due to defects in the transcription factors HNF-4a, PDX-1, and HNF-1b, respectively. Each of these affects only a few families. In all forms of MODY, the gene involved may have a wide variety of mutations, including point mutations, nonsense mutations leading to premature termination, and splicing defects. This makes screening for these defects more difficult in individual patients. Almost all patients are heterozygous for the mutation, with one normal allele and one mutant allele. The mechanism by which the mutant enzyme becomes dominant over the normal enzyme is poorly understood. The genetic defect in 15% to 30% of families has not yet been determined.57,58 and 59

NONDIABETIC MELLITURIA The presence of some form of sugar in the urine is not necessarily diagnostic of diabetes. At one time, the forms of nondiabetic mellituria were confused with diabetes, but this occurs rarely now that the diagnosis of diabetes is based primarily on blood glucose measurement. Moreover, urinary glucose is determined with more specific glucose oxidase detection systems (e.g., Tes-Tape, Clinistix, Diastix, Chemstrip uG). Nonetheless, sugar in the urine may be due to diseases other than diabetes. RENAL GLUCOSURIA The most commonly recognized nondiabetic mellituria is renal glucosuria.60 This disorder accounted for as many as 1 of 500 patients referred for evaluation of diabetes when the diagnosis depended primarily on urine glucose measurements. The diagnosis of renal glycosuria requires the detection of glucosuria with simultaneous normal plasma glucose on timed urine samples taken during an oral glucose-tolerance test. Renal glucosuria usually is not associated with other urinary tubular absorptive defects (e.g., aminoaciduria of the Fanconi syndrome should be excluded) and appears to be benign. Some patients with renal glucosuria have mistakenly been treated for decades with insulin. With increased reliance on measurements of serum and capillary glucose and on hemoglobin A1C, such mistaken diagnoses should be extremely rare. Normally, glucose is actively reabsorbed at the proximal renal tubule after filtration at a concentration equal to that in plasma. In normal individuals, the reabsorptive capacity of the proximal tubule exceeds the filtered load of glucose, and little glucosuria occurs with a plasma glucose concentration of 75% of the pancreas is removed or destroyed. Endocrine secretion from all islet-cell types is reduced, resulting not only in insulin deficiency under basal and/or stimulated conditions, but also in diminished pancreatic glucagon, somatostatin, and pancreatic polypeptide (PP) secretion. Despite preserved secretion of glucagon-like substances of duodenal origin in patients who have not undergone pancreatoduodenectomy, reduced levels of pancreatic glucagon account for the relative resistance of these patients to ketoacidosis, notwithstanding their insulin deficiency. In addition, average daily insulin requirements are lower in pancreatectomized diabetic patients than in matched patients with type 1 diabetes.92,93 Hypoglycemic events, however, occur with increased frequency in pancreatectomized patients,94,95 and their response to spontaneous or induced hypoglycemia is delayed compared to that of patients with both type 1 and type 2 diabetes.95,96 PATHOPHYSIOLOGY A marked decrease in insulin secretion is the major factor underlying postpancreatectomy diabetes. Plasma glucagon may be low, normal, or even high.97 This is partly due to cross-reactivity of the antibodies used in the immunoassay with glucagon-like substances of gastrointestinal origin, the levels of which may be elevated after pancreatectomy. The absence of a glucagon response to arginine stimulation and the very low levels of 3500-Da glucagon indicate deficient A-cell function.92 The low levels of pancreatic glucagon are responsible for the relative resistance to ketosis characteristic of this group of patients. This, along with increased cellular insulin binding and enhanced hepatic and extrahepatic sensitivity to exogenous insulin,94 is also responsible for the high frequency of hypoglycemic episodes. Pancreatectomized subjects also display subnormal epinephrine responses to insulin-induced hypoglycemia, which are significantly more severe than those seen in patients with type 1 diabetes matched for autonomic neuropathy or for the level of C peptide.96 Because free insulin levels are not significantly different in pancreatectomized patients from those in other insulin-requiring diabetic patients, the absence of spontaneous glucose recovery seen in these patients is probably secondary to the combined defects in glucagon and epinephrine response. PANCREATITIS CLINICAL FEATURES Glucose intolerance may be part of the clinical picture in both acute and chronic pancreatitis. Hyperglycemia is often transient in acute pancreatitis, and may be severe. Abnormalities may persist in 3% to 5% of patients.98 Chronic pancreatitis may be associated with diabetes in up to 60% of patients, with the incidence varying with the geographic location. This reflects the fact that alcoholic pancreatitis, the most common cause of chronic pancreatitis in the United States, is relatively infrequently associated with diabetes, whereas diabetes occurs frequently with fibrocalcific pancreatitis, especially in the tropical regions of the world. This has led to the use of the terms “tropical diabetes”99 and “J-type diabetes” (“J” is for Jamaica). These forms of secondary diabetes often include degrees of both pancreatic exocrine and endocrine dysfunction. Diabetes secondary to pancreatitis differs somewhat from diabetes secondary to pancreatectomy, with a reduced frequency of hypoglycemic events, a higher incidence of development of ketosis, and, generally, a greater requirement for insulin. PATHOPHYSIOLOGY Patients with pancreatitis exhibit variable decreases in the plasma insulin level. This decrease correlates with increased pancreatic enzyme release in acute pancreatitis and decreased pancreatic exocrine function in chronic cases. Carbohydrate intolerance usually has been attributed to the diminished insulin secretion98; however, patients with chronic pancreatitis often have normal basal plasma insulin levels but decreased insulin responses to glucose, arginine, and glucagon challenge.98,100 Studies in the canine pancreatic duct ligation model of chronic pancreatitis101 suggest that abnormal islet responsiveness with resultant circulating insulin deficiency may be associated with pancreatic acinar fibrosis, whereas the islets of Langerhans remain histologically and ultrastructurally intact. Insulin deficiency alone may not be the only factor responsible for the development of secondary diabetes in the setting of pancreatitis. Hepatic resistance to insulin, accompanied by loss of sensitivity to insulin-induced hepatic glucose suppression, is a prominent feature of canine and rodent chronic pancreatitis.102 Deficiency of PP has been implicated as a factor in this resistance, and infusion of bovine PP ameliorates glucose intolerance and restores suppression of hepatic glucose output by insulin in PP-deficient animals103,104 and patients with chronic pancreatitis.105 In addition, the basal pancreatic glucagon level is significantly higher in patients than in controls,106 and it is increased further after oral glucose and arginine infusion.100 Evidence exists that “pancreatic” glucagon may also be secreted by the duodenal mucosa107 and that this may be the source of the elevated glucagon levels found in patients with pancreatitis and B-cell secretory deficiency. In contrast, because the duodenum is often removed along with the pancreas during pancreatectomy, duodenal glucagon may not be available in pancreatectomized patients.

HEMOCHROMATOSIS CLINICAL FEATURES Hemochromatosis includes various clinical conditions in which cirrhosis is accompanied by a markedly increased hepatic iron content, as well as by variable amounts of iron deposition in other tissues (see Chap. 131). Hemochromatosis may be a primary (idiopathic) familial disorder or secondary to iron overload. The latter form of hemochromatosis occurs in some patients with chronic hemolytic anemia or thalassemia, in patients receiving multiple transfusions, or in persons ingesting large amounts of iron, such as the Bantus, who drink excessive amounts of Kaffir beer. Clinical diabetes or impaired glucose tolerance occurs in 75% to 90% of patients with primary hemochromatosis and up to 65% of patients with hemochromatosis secondary to hemolytic anemia, multiple transfusions, or iron ingestion, and usually precedes other manifestations of the disease.108 and 110 The diabetes of both primary and secondary hemochromatosis often requires insulin administration, and ketosis can occur. PATHOPHYSIOLOGY Three factors contribute to the abnormal glucose tolerance in hemochromatosis: cirrhosis, pancreatic iron deposition, and the coexistence of primary diabetes mellitus. Although the presence of cirrhosis increases the likelihood of abnormal glucose metabolism, hepatic iron content, serum ferritin levels, or the extent of liver damage correlate poorly with the presence of impaired glucose homeostasis. A positive family history of diabetes may, perhaps, be the best predictor of glucose intolerance, at least among patients with primary, hereditary hemochromatosis.111 Hepatic insulin resistance clearly plays an important role in patients with both varieties of hemochromatosis.112,113 Defective first-phase insulin secretion is also observed, however, even in the absence of significant degrees of islet iron deposition. Glucoseand insulin-tolerance testing and euglycemic clamp studies109,110,113 suggest that the defect in insulin secretion is more important than is insulin resistance in this syndrome. This diminished insulin secretion, however, may be preceded by insulin resistance and increased insulin secretion,110 a pattern reminiscent of the natural history of type 2 diabetes mellitus. Overall, however, the relative contributions of genetic factors versus iron overload to the pathophysiology of diabetes secondary to hemochromatosis are not yet clear.

DRUG-INDUCED DIABETES A wide variety of pharmacologic agents can induce glucose intolerance (Table 139-3). Individual drugs may affect glucose homeostasis by interfering with insulin secretion, insulin action, or both. Whether its effect is primarily on insulin secretion or insulin action, a drug itself may mediate the effect directly or indirectly through balance of insulin counterregulatory hormones or cations (e.g., K+) critical to the mechanisms that control insulin release. Glucohomeostatic effects of supraphysiologic levels of GH, glucocorticoids, and catecholamines best illustrate the direct and indirect consequences of pharmacologically altered insulin action. As the links between altered insulin sensitivity and altered insulin secretion tighten, assigning a drug an effect based solely on insulin action becomes more and more difficult. Clinically, these agents may uncover previously silent insulin-secretory defects or insulin resistance and consequently induce glucose intolerance in a previously undiagnosed patient or worsen the diabetic state when administered to patients with antecedent diabetes mellitus.

TABLE 139-3. Diabetogenic Pharmacologic Agents

IMPAIRED INSULIN SECRETION (DIRECT EFFECT)—PROTOTYPE: PHENYTOIN SODIUM The diabetogenic effect of phenytoin sodium in humans and animals has been known for many years.114 Early studies in the isolated perfused rat pancreas showed that high concentrations of this drug completely inhibited both first and second phases of insulin release.115 Phenytoin sodium has been shown to reversibly inhibit calcium inflow into the B cell through voltage-dependent calcium channels.116 When given orally to normal subjects in doses sufficient to achieve therapeutic blood levels, the drug produces a significant decrease in both early and late insulin responses to intravenous glucose, accompanied by a significant rise in plasma glucose level.117 Another pharmacologic agent in this category that deserves special mention is pentamidine isethionate, a drug used in the treatment of trypanosomiasis and leishmaniasis and now extensively used to treat Pneumocystis carinii pneumonia. Given the wide use of this drug in the management of patients with acquired immunodeficiency syndrome, its effects on glucose homeostasis are being noted with increasing frequency.118 The chemical structure of pentamidine isethionate is somewhat similar to that of the diabetogenic rodenticide pyriminil (Vacor).119 When pentamidine isethionate is incubated in vitro with islet cells, an initial passive release of insulin occurs, followed by a significant decrease in B-cell response to glucose and theophylline. The proposed mechanism of pentamidine toxicity, therefore, is an early cytolytic release of insulin followed by B-cell exhaustion and insulin deficiency. IMPAIRED INSULIN SECRETION (INDIRECT EFFECT)—PROTOTYPE: THIAZIDE DIURETICS The incidence of glucose intolerance in patients taking thiazide diuretics varies between 10% and 40%, depending on the duration of administration of the drug.120 Thiazides were thought to adversely affect insulin secretion and promote glucose intolerance indirectly through production of hypokalemia,121 in a manner similar to that proposed for primary hyperaldosteronism. Although prevention or correction of hypokalemia does ameliorate thiazide-induced glucose intolerance, direct effects of thiazides themselves on pancreatic B-cell secretion have been described. Unlike the structurally related compound diazoxide, thiazides do not hyperpolarize B cells by opening the adenosine triphosphate (ATP)–sensitive potassium channels closed by the sulfonylureas.122 Instead, they, like diphenylhydantoin, may affect stimulus-secretion coupling in the B cell by inhibiting calcium uptake.123 Similarly, the loop diuretics, thought to share with the thiazides an indirect effect on B-cell secretion mediated through hypokalemia, also directly affect insulin secretion by inhibiting chloride pump function in the B-cell membrane.124 IMPAIRED INSULIN ACTION—PROTOTYPE: GLUCOCORTICOIDS The mechanism of glucocorticoid-induced glucose intolerance has been presented earlier. Like the catecholamines, glucocorticoids increase glucose production as well as decrease insulin secretion and insulin-stimulated glucose utilization (see Table 139-2). Although glucocorticoid administration was used formerly as a diagnostic maneuver in an attempt to precipitate abnormal glucose tolerance in individuals with a family history of diabetes, this test has proved unreliable because all individuals are prone to the diabetogenic actions of these drugs. IMPAIRED INSULIN SECRETION AND IMPAIRED INSULIN ACTION—PROTOTYPE: b-ADRENERGIC BLOCKERS The effects of b-adrenergic blocking agents on intermediary metabolism can be deduced from a knowledge of the adrenergic control of insulin secretion, glucose production, and glucose utilization. Generally, nonselective b-blockade decreases pancreatic insulin secretion. Glycogenolysis and gluconeogenesis also may be impaired, but the clinical effect of these may be offset by the reduction in insulin secretion. Consequently, a worsening of glucose tolerance in patients receiving b-blockers is not unusual.125 The occurrence of hypoglycemia in these patients is somewhat less common.126 The a-adrenergic contribution to hepatic glycogenolysis may be the reason for the low frequency of this finding. Relatively selective b-blockade (e.g., with drugs such as atenolol) would be expected to have a smaller effect on insulin secretion and on recovery from insulin-induced hypoglycemia, because the stimulatory pancreatic a-adrenoreceptor is a2 in subtype and the hepatic

b-adrenoreceptor is b2 in subtype (see Chap. 85). The opiates also fall into this category of pharmacologic agents. The hyperglycemic effect of morphine has been well described,127 and heroin addicts have elevated basal plasma insulin levels but markedly reduced plasma insulin responses to intravenous glucose in comparison with age-, gender-, and weight-matched controls.128 Beta-endorphin can be found in human pancreatic tissue.129 This observation raises the question of a possible role for endogenous opioids in the intrapancreatic control of insulin secretion.

GENETIC SYNDROMES ASSOCIATED WITH IMPAIRED GLUCOSE TOLERANCE The list of genetic syndromes that include glucose intolerance as part of their profile is extensive and growing (see Table 139-1). Among members of this list, relatively pure defects in insulin secretion are represented by diseases such as cystic fibrosis, whereas leprechaunism may be regarded as a prototypical syndrome of insulin resistance. However, multiple pathophysiologic mechanisms that affect both glucose production and glucose utilization combine, in most cases, to produce the full-blown syndromes and the glucose intolerance that characterizes them. Advances in the molecular genetics and molecular pathophysiology of these syndromes likely will shed light not only on the dysregulated glucose handling in these syndromes but also on mechanisms of altered glucose homeostasis common to secondary as well as to primary forms of diabetes.

COMPLICATIONS OF SECONDARY DIABETES Over the past 20 years, the question of whether the vascular complications of diabetes are a direct consequence of the metabolic abnormalities present in the diabetic patient or are secondary to another genetic lesion has remained controversial. The results of the Diabetes Control and Complications Trial provide compelling evidence that strict control of plasma glucose levels in patients with type 1 diabetes can retard the progression of microangiopathic changes in the eye, kidney, and nerves.130 Whether such therapy can actually prevent these complications is still unknown. This question has been difficult to answer for at least two reasons: first, there is a lack of information about a possible underlying genetic defect in diabetes. Second, it is not known whether hyperglycemia itself or some other factor in the diabetic milieu (insulin, IGF-I, other growth factors) initiates the sequence of events leading to microangiopathy. Even less is known about the relationship of these factors, if any, to macrovascular changes. Patients with glucose intolerance secondary to one of the diseases of hormone overproduction have a notable lack of microangiopathic complications. This is not surprising because frank diabetes is relatively uncommon in this group (affecting 200 U per day.15,84 The plasma of patients with immune insulin resistance can bind >50 U of insulin per liter, which means that the amount introduced by subcutaneous injections is a small proportion of the total amount of bound insulin. Sometimes these patients were extremely difficult to treat because they had poor control even when given thousands of units of insulin per day. Such patients have no meaningful peak of whatever insulin they are given and sometimes are best treated with U500 regular insulin administered twice a day. A species switch to human insulin sometimes leads to a reduction in the dosage requirement, and even lispro insulin has helped when antibodies are strongly directed against human insulin.85 Another strategy has been to give corticosteroids to reduce antibody production. (Some success also has been obtained with sulfated insulin, which is not as well bound by antibodies because epitopes are hidden by the added sulfate groups.84 However, sulfated insulin is no longer routinely made by any companies.)

LOCALIZED REACTIONS TO INSULIN Localized reactions to insulin used to be seen commonly with the initiation of insulin treatment, but with the introduction of purified insulin, they are now rarely significant. Reactions used to be noted in as many as 50% of patients but now are found in 24 hours. Another approach to prolonging the action of insulin is to couple a fatty acid chain to Lys (B29), a modification that prolongs insulin's action because the fatty acid binds to albumin.88 Recombinant human insulin-like growth factor-I (IGF-I) combined with insulin has been given by injection on an experimental basis to patients with type 1 diabetes and has resulted in some improvement in control.89 The longterm influence of IGF-I on the micro- and macrovascular complications of diabetes remains to be determined. The administration of insulin by the nasal route has been studied by several groups, but problems with nasal irritation and variable absorption have made it problematic.90 The use of inhaled insulin may be more promising.91 The absorption of this form of insulin is very rapid and could be very useful when given before meals. Variable-rate insulin pumps implanted in the abdominal wall with delivery of insulin into the peritoneal space have received considerable attention in the past few years as potential treatment options for patients with both type 1 and type 2 diabetes.92 Some patients have been treated successfully with these devices for >5 years. Intraperitoneal insulin delivery is attractive because of the theoretic advantages of portal insulin delivery, including the possibility of reducing systemic hyperinsulinemia and better regulating intermediary metabolism. The devices are “open-loop” systems that rely on self-monitoring of the glucose level. A “closed-loop” system, in which a monitor capable of continuous glucose sensing is linked to an intraperitoneal pump, would be a major advance. A large and complex closed-loop artificial pancreas has been available for hospitalized patients for some time, but it is impractical for home use and rarely is used now, even in hospitals. Progress in development of glucose sensors has been slow, but some advances have been made with a noninvasive approach in which absorbance spectra are measured using light spectros-copy in the near infrared range.93 Another approach has been to use a glucose oxidase–sensing system on the tip of a needle, which could provide a constant readout of subcutaneous glucose levels for a period of several days. Whole pancreas transplantation is now being offered by many medical centers but is largely restricted to patients who also are receiving kidney transplants with immunosuppression (see Chap. 144). Islet transplantation with human islet allografts is being studied as an experimental procedure in a small number of centers; results are improving but still are not equal to those obtained with whole pancreas transplants (see Chap. 144). CHAPTER REFERENCES 1. Rosenzweig JL. Principles of insulin therapy. In: Kahn CR, Weir GC, eds. Joslin's diabetes mellitus, 13th ed. Philadelphia: Lea & Febiger, 1994:460.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59.

Schernthaner G. Immunogenicity and allergenic potential of animal and human biosynthetic human insulin. Diabetes Care 1993; 16(Suppl 3):133. Chance RE, Frank BH. Research, development, production, and safety of biosynthetic human insulin. Diabetes Care 1993; 16(Suppl 3):133. Heinemann L, Richter B. Clinical pharmacology of human insulin. Diabetes Care 1993; 16(Suppl 3):90. Berger W, Keller U, Honegger B, Jaeggi E. Warning symptoms of hypoglycemia during treatment with human and porcine insulin in diabetes mellitus. Lancet 1989; 1:1041. Cryer PE. Hypoglycemia unawareness in IDDM. Diabetes Care 1993; 16(Suppl 3):40. Galloway JA, Kooper SA, Spradlin CT, et al. Biosynthetic human proinsulin: review of chemistry, in vitro and in vivo receptor binding, animal and human pharmacology studies, and clinical experience. Diabetes Care 1992; 14:666. Holleman F, Hoekstra JBL. Insulin lispro. N Engl J Med 1997; 337:176. Lalli C, Ciofetta M, Del Sindaco P, et al. Long-term intensive treatment of type 1 diabetes with the short-acting insulin analog lispro in variable combination with NPH insulin at mealtime. Diabetes Care 1999; 22:468. Rassam AG, Zeise TM, Burge MR, et al. Optimal administration of lispro insulin in hyperglycemic type 1 diabetes. Diabetes Care 1999; 22:133. Joseph SE, Korzon-Burakowska A, Woodworth JR, et al. The action profile of lispro is not blunted by mixing in the syringe with NPH insulin. Diabetes Care 1999; 21:2098. Mohn A, Ross KM, Matyka KA, et al. Lispro of regular insulin for multiple injection therapy in adolescence: differences in free insulin and glucose levels overnight. Diabetes Care 1999; 22:27. Brunell RL, Llewelyn J, Anderson Jr JH, et al. Metaanalysis of the effect of insulin lispro on severe hypoglycemia in patients with type 1 diabetes. Diabetes Care 1998; 21:1726. Melki V, Renard E, Lassmann-Vague V, et al. Improvement of HbA1C and blood glucose stability in IDDM patients treated with lispro insulin analog in external pumps. Diabetes Care 1999; 21:977. Galloway JA, de Shazo RD. Insulin chemistry, pharmacology, dosage algorithms and the complications of insulin treatment. In: Rifkin H, Porte D Jr, eds. Ellenberg and Rifkin's diabetes mellitus theory and practice, 4th ed. New York: Elsevier, 1990:497. Ahmed ABE, Home PD. Optimal provision of daytime NPH insulin in patients using the insulin analog lispro. Diabetes Care 1998; 21:1707. Anderson JH, Massey EH. Flocculated humulin N insulin. (Letter). N Engl J Med 1987; 316:1027. Zinman B, Ross S, Campos RV, et al. Effectiveness of human Ultralente versus NPH insulin in providing basal insulin replacement for an insulin lispro multiple daily injection regimen. Diabetes Care 1999; 22:603. Sherwin RS, Kramer KJ, Tobin JD, et al. A model of the kinetics of insulin. J Clin Invest 1974; 53:1481. Stimmler L. Disappearance of immunoreactive insulin in normal and adult onset subjects. Diabetes 1967; 16:652. Navelesi R, Pilo A, Ferrannini E. Kinetic analysis of plasma disappearance in nonketotic diabetic patients and in normal subjects. J Clin Invest 1978; 61:197. Polonsky KS, Given BD, Hirsch L, et al. Quantitative study of insulin secretion and clearance in normal and obese subjects. J Clin Invest 1988; 81:435. Ferrannini E, Wahren J, Faber OK, et al. Splanchnic and renal metabolism of insulin in human subjects: a dose response study. Am J Physiol 1983; 244:E517. Binder C, Lauritzen T, Faber O, Prammer S. Insulin pharmacokinetics. Diabetes Care 1984; 7:188. Galloway JA, Spradlin CT, Howey DC, Dupre J. Intrasubject differences in pharmacokinetic and pharmacodynamic responses: the immutable problem of present day treatment? In: Serrano-Rio M, Lefebvre PJ, eds. Diabetes 1985, New York: Elsevier, 1986:877. Kurtz AB, Nabarro JDN. Circulating insulin-binding antibodies. Diabetologia 1980; 19:329. Hildebrant P, Birch K, Sestoft O, Nielson ST. Orthostatic changes in subcutaneous blood flow and insulin absorption. Diabetes Res 1985; 2:187. Heine RJ, Bilo HJG, Fonk T, et al. Absorption kinetics and action profiles of mixtures of short- and intermediate-acting insulins. Diabetologia 1984; 27:558. Colagiuri S, Villalobos S. Assessing effect of mixing insulins by glucose-clamp technique in subjects with diabetes mellitus. Diabetes Care 1986; 9:579. Galloway JA, Spradlon T, Jackson RL, et al. Mixtures of intermediate acting insulin: an update. In: Skyler JS, ed. Insulin update: 1982. Amsterdam: Excerpta Medica, 1982:111. Koivisto VA, Tuominen JA, Ebeling P. Lispro Mix25 insulin as premeal therapy in type 2 diabetic patients. Diabetes Care 1999; 22:459. Simone EA, Wegmann DR, Eisenbarth GS. Immunologic “vaccination” for the prevention of autoimmune diabetes (type 1A). Diabetes Care 1999; 22(Suppl 2):B7. Berger M, Jorgens V, Muhihauser I. Rationale for the use of insulin therapy alone as the pharmacological treatment of type 2 diabetes. Diabetes Care 1999; 22(Suppl 3):C71. The Diabetes Control and Complications Trial Research Group. The effects of intensive treatment of diabetes on the development and progression of longterm complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329:977. Intensive blood-glucose control with sulfonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). United Kingdom Prospective Diabetes Study Group. Lancet 1998; 352:837. Effect of intensive blood-glucose control with metformin on complications in overweight patients with type 2 diabetes (UKPDS 34). United Kingdom Prospective Diabetes Study Group. Lancet 1998; 352:854. Nathan DM. Long-term complications of diabetes mellitus. N Engl J Med 1993; 328:1676. Boyne MS, Saudek CD. Effect of insulin therapy on macrovascular risk factors in type 2 diabetes. Diabetes Care 1999; 22(Suppl 3):C45. American Diabetes Association. Standards of medical care for patients with diabetes mellitus. Diabetes Care 1999; 22(Suppl 1):S32. Bolli GB. How to ameliorate the problem of hypoglycemia in intensive as well as nonintensive treatment of type 1 diabetes. Diabetes Care 1999; 22(Suppl 2):B43. Yki-Jarvinen H, Kauppila M, Kujansuu E, et al. Comparison of insulin regimens in patients with NIDDM. N Engl J Med 1992; 327:1426. Yki-Jarvinen H, Ryysy L, Nikkila K, et al. Comparison of bedtime insulin regimens in patients with type 2 diabetes mellitus: a randomized trial. Ann Intern Med 1999; 130:389. Buse JB. Overview of current therapeutic options in type 2 diabetes. Diabetes Care 1999; 22:C65. Bode BW, Steed RD, Davidson PC. Reduction in severe hypoglycemia with long-term continuous subcutaneous insulin infusion in type 1 diabetes. Diabetes Care 1996; 19:324. Gillespie SA, Kulkarni KD, Daly AE. Using carbohydrate counting in diabetes clinical practice. J Am Diet Assoc 1998; 98:897. Farkas-Hirsch R, Hirsch IB. Continuous subcutaneous insulin infusion: a review of the past and its implementation for the future. Diabetes Spectrum 1994; 7:80. Zinman B, Tildesley H, Chaisson JL, et al. Insulin lispro in CSII: results of a double-blind crossover study. Diabetes 1997; 46:440. Gill GV, Walford S, Alberti KGMM. Brittle diabetes. Present concepts. Diabetologia 1985; 28:579. Paulsen EP, Courtney GW, Duckworth WC. Insulin resistance caused by massive degradation of subcutaneous insulin. Diabetes 1979; 28:640. Home PD, Massi-Benedetti M, Gill GV, et al. Impaired subcutaneous absorption of insulin in “brittle” diabetes. Acta Endocrinol (Copenh) 1982; 101:414. Wolfsdorf JI, Anderson BJ, Pasquarello C. Treatment of the child with diabetes. In: Kahn CR, Weir GC, eds. Joslin's diabetes mellitus, 13th ed. Philadelphia: Lea & Febiger, 1994:530. Tattersall RB, Lowe J. Diabetes and adolescence. Diabetologia 1981; 20:517. Kitzmiller JL. Sweet success with diabetes: the development of insulin therapy and glycemic control for pregnancy. Diabetes Care 1993; 16(Suppl 3):107. Hadden DR. How to improve prognosis in type 1 diabetic pregnancy: old problems, new concepts. Diabetes Care 1999; 22(Suppl 2):B104. Agner T, Damni P, Binder C. Remission in IDDM: prospective study of basal C-peptide and insulin dose in 265 consecutive patients. Diabetes Care 1987; 10:164. Perlman K, Erlich RM, Filler RM, Albisser AM. Sustained normoglycemia in newly diagnosed type I diabetic subjects: short term effects and one year followup. Diabetes 1984; 33:995. Rabkin R, Simon NM, Steiner S, Colwell JA. Effect of renal disease on renal uptake and excretion of insulin in man. N Engl J Med 1970; 282:182. Rabkin R, Ryan MP, Duckworth WC. The renal metabolism of insulin. Diabetologia 1984; 27:35. Amair P, Khanna R, Leibel B, et al. Continuous ambulatory dialysis in diabetics with endstage renal disease. N Engl J Med 1982; 306:625.

59a. Fritsche A, Stumvoll M, Hüring HU, Gerich JE. Reversal of hypoglycemia unawareness in a long-term type 1 diabetic patient by improvement of b-adrenergic sensitivity after prevention of hypoglycemia. J Clin Endocrinol Metab 2000; 85:523. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85.

Korytkowski MT, Mokan M, Veneman TF, et al. Reduced b-adrenergic sensitivity in patients with type 1 diabetes and hypoglycemia unawareness. Diabetes Care 1998; 21:1939. Meyer C, Grosmann R, Mitrakou A, et al. Effects of autonomic neuropathy on counterregulation and awareness of hypoglycemia in type 1 diabetic patients. Diabetes Care 1998; 21:1960. Somogyi M. Insulin as a cause of extreme hyperglycemia and instability. Bull St Louis Med Soc 1938; 32:498. Gerich JE, Campbell PJ. Overview of counterregulation and its abnormalities in diabetes mellitus and other conditions. Diabetes Metab Rev 1988; 4:93. Clutter WE, Rizza RA, Gerich JE, Cryer PE. Regulation of glucose metabolism by sympathochromaffin catecholamines. Diabetes Metab Rev 1988; 4:1. Lerman IG, Wolfsdorf JI. Relationship of nocturnal hypoglycemia to daytime glycemia in IDDM. Diabetes Care 1988; 11:636. Hirsch IB, Smith LJ, Havlin CE, et al. Failure of nocturnal hypoglycemia to cause daytime hyperglycemia in patients with IDDM. Diabetes Care 1990; 13:133. Gerich JE. Glucose counterregulation and its impact on diabetes mellitus. Diabetes 1988; 37:1608. Bolli G, Fanelli CG, Perrielbo G, De Feo P. Nocturnal blood glucose control in type I diabetes mellitus. Diabetes Care 1993; 16(Suppl 3):71. Campbell P, Bolli G, Cryer P, Gerich J. Pathogenesis of the dawn phenomenon in patients with insulin-dependent diabetes mellitus. N Engl J Med 1985; 312:1473. Evans DJ, Pritchard-Jones K, Trotman-Dickenson B. Insulin oedema. Postgrad Med J 1986; 62:665. O'Hare JA, Ferriss JB, Twomey B, O'Sullivan DJ. Poor metabolic control, hypertension and microangiopathy independently increase the transcapillary escape rate of albumin in diabetes. Diabetologia 1983; 25:260. Wheatly T, Edwards OM. Insulin edema and its clinical significance: metabolic studies in three cases. Diabet Med 1985; 2:400. Reeves W, Allen B, Tattersell R. Insulin-induced lipoatrophy: evidence for an immune pathogenesis. Br Med J 1980; 280:1500. Young RJ, Hannan W, Frier B, et al. Diabetic lipohypertrophy delays insulin absorption. Diabetes Care 1984; 7:479. Rowe JW, Young JB, Minaker KM, et al. Effect of insulin and glucose infusions on sympathetic nervous system activity in normal man. Diabetes 1981; 30:219. Page MM, Watkins PJ. Provocation of postural hypotension by insulin in diabetic autonomic neuropathy. Diabetes 1976; 25:90. Chance R, Root M, Galloway J. The immunogenicity of insulin preparations. Acta Endocrinol (Copenh) 1976; 83(Suppl 205):185. Kahn C, Mann D, Rosenthal A, et al. The immune response to insulin in man: interaction of HLA alloantigens and the development of the immune response. Diabetes 1982; 31:716. Kahn C, Rosenthal A. Immunologic reactions to insulin: insulin allergy, insulin resistance and the autoimmune insulin syndrome. Diabetes Care 1979; 2:283. Stewart W, McSweeney S, Kellett M, et al. Increased risk of severe protamine reactions in NPH-insulin-dependent diabetics undergoing cardiac catheterization. Circulation 1984; 70:788. Feinglos M, Jegasothy B. “Insulin” allergy due to zinc. Lancet 1979; 1:122. Waldhüusl W, Bratusch-Marrain P, Kruse V, et al. Effect of insulin antibodies on insulin pharmacokinetics and glucose utilization in insulin-dependent diabetic patients. Diabetes 1985; 34:166. Van Haeften T, Bolli G, Dimitriadis G, et al. Effect of insulin antibodies and their kinetic characteristics on plasma free insulin dynamics in patients with diabetes mellitus. Metabolism 1986; 35:1649. Davidson J, DeBra D. Immunologic insulin resistance. Diabetes 1978; 27:307. Uahtela JT, Antonen J, Knip M, et al. Severe insulin-mediated human insulin resistance: successful treatment with the insulin analog lispro. Diabetes Care 1997; 20:71.

85a. Gale EA. A randomized, controlled trial comparing insulin lispro with human soluble insulin in patients with Type 1 diabetes on intensified insulin therapy. The UK trial group. Diabet Med 2000; 17:209. 86. 87. 88. 89. 90. 91. 92. 93.

Home PD, Lindholm A, Hylleberg B, et al. Improved glycemic control with insulin aspart. Diabetes Care 1999; 21:1904. Gillies PS, Figgitt DP, Lamb HM. Insulin glargine. Drugs 2000; 59:253. Markussen J, Havelund S, Kurtzhals P, et al. Soluble fatty acid acylated insulins bind to albumin and show protracted action in pigs. Diabetologia 1996; 39:281. Thrailkill KM, Quattrin T, Baker U, et al. Cotherapy with recombinant human insulinlike growth factor I and insulin improves glycemic control in type 1 diabetes. Diabetes Care 1999; 22:585. Salzman R, Manson JE, Griffing GT, et al. Intranasal aerosolized insulin: mixed meal studies and long-term use in type I diabetes. N Engl J Med 1985; 312:1078. Gelfand B, Sherwyn L, Schwartz L, et al. Pharmacological reproducibility of inhaled human insulin premeal dosing in patients with type 2 diabetes. Diabetes 1998; 47(Suppl 1):A99. Saudek CD. Future developments in insulin delivery systems. Diabetes Care 1993; 16(Suppl 3):122. Gough DA, Armour JC, Baker DA. Advances and prospects in glucose assay technology. Diabetologia 1997; 40(Suppl 2):102.

CHAPTER 144 PANCREAS AND ISLET TRANSPLANTATION Principles and Practice of Endocrinology and Metabolism

CHAPTER 144 PANCREAS AND ISLET TRANSPLANTATION GORDON C. WEIR Pancreas Transplantation Islet Transplantation Human Islet Allografts Allografts in the Absence of Autoimmunity Autografts The Two Major Barriers to Successful Islet Transplantation What Sources of Insulin-Producing Tissue are Available? Xenotransplantation Efforts to Expand B Cells and to Create Insulin-Producing Cells with Genetic Engineering Immune Attack on Transplanted Islets Therapeutic Approaches of Rejection and Autoimmunity Immunobarrier Technology Gene-Transfer Technology to Protect Islets Predicting the Future Chapter References

Although insulin was discovered many years ago, the complications of diabetes still produce devastating consequences. The link between high blood-glucose levels and the complications of retinopathy, nephropathy, and neuropathy is now established beyond doubt.1 Although we know more than ever about insulin therapy, largely thanks to self blood-glucose monitoring, only a small proportion of people with type 1 diabetes obtain good enough glycemic control to avoid complications (see Chap. 136, Chap. 138 and Chap. 143). And even for those, the cost in terms of lifestyle change and monetary expense is substantial. Type 2 diabetes, being greater than ten times more common than type 1 diabetes, causes even more problems in terms of health outcomes, economic cost, and personal tragedy; yet this, too, has proved difficult to control (see Chap. 137 and Chap. 139). Although solutions to the problem of diabetes could come from various unknown sources, an obvious path is some form of B-cell replacement therapy. This is an accepted goal for type 1 diabetes, but even type 2 diabetes could be greatly helped by such a therapy. Major contributing factors to the development of type 2 diabetes are the Western way of life, with its obesity and sedentary lifestyle, and strong genetic factors that work together to cause insulin resistance (see Chap. 141, Chap. 145 and Chap. 146). Diabetes only develops when the pancreatic B cells (beta cells) can no longer compensate. Relative B-cell failure is a critical factor that could be remedied by B-cell replacement therapy. There are two potential routes for B-cell replacement, one being some form of B-cell transplantation in the form of a pancreas transplant or as insulin-producing cells; the other is a mechanical B cell, a path that continues to be elusive.

PANCREAS TRANSPLANTATION The first pancreas transplants were performed in the 1960s, but the procedure was not widely applied until the mid-1980s.2,3,4 and 5 Since 1978, >9000 pancreas transplants have been performed, and currently >1000 are done yearly, ~70% of these in the United States. The most common transplants are simultaneous kidney/pancreas (SKP) transplants, given to type 1 diabetic patients with advanced nephropathy who need a kidney transplant. Much less common are pancreas transplants done after a kidney allograft (pancreas after kidney; PAK), which usually require more immunosuppression, and even fewer receive a pancreas transplant alone (PTA), although a single center has reported 225 cases of PTA.6 In making decisions, patients with failing kidneys are often urged to accept a kidney from a living related or unrelated donor because of the benefits obtained from receiving a kidney alone, rather than waiting for an SKP from a cadaver donor. Some patients are judged to have so much trouble with their diabetes (with problems such as hypoglycemia unawareness) that a PTA is recommended. Some centers have experience with living related donors who provide the distal portion of their pancreases, but there continues to be uncertainty about the overall benefit of this approach, because the donors can experience dangerous surgical complications and often have glucose intolerance after the procedure, with some developing diabetes.7 The best results are obtained when a kidney and pancreas are transplanted simultaneously (SKP), with experienced centers finding that ~85% of the pancreases maintain perfect control 1 year after the transplant, and ~50% of the transplants are still working well after 5 years. Fewer good results are reported for PAK and PTA, but with more aggressive immunosuppression better results are being obtained. Although the first experimental transplants used enteric drainage, for a period of ~10 years the favored approach was to place the pancreas with the duodenum attached in the right lower pelvis, with pancreatic juices draining through the duodenum into the anastomosed bladder. Because many patients have had problems with acidosis, dehydration, and a variety of bladder problems, many centers have now switched back to enteric drainage.8 The most commonly used immuno-suppression in the past has been triple therapy with cyclosporine, azathioprine, and prednisone, but now many centers are switching to tacrolimus (FK-506) and mycophenolate mofetil, and continuing with prednisone.9 Antibodies against thymocytes and the interleukin-2 (IL-2) receptor are also used. In general, higher doses of immunosuppression are used for pancreas transplants than for kidney transplants alone, which is worrisome because of the increased incidence of infection and malignancy. The immunosuppression is required, not only for allograft rejection but also for autoimmunity. An interesting set of pancreas/kidney transplants was performed between identical twins, who were not given immunosuppressive medication. As expected there was no rejection of the exocrine pancreas or liver, but diabetes recurred with immune destruction of the islet demonstrated by biopsy, no doubt mediated by the autoimmunity that caused the diabetes in the first place.10 The extra surgery of a pancreas transplant is accompanied by considerable morbidity. Patients often have long hospitalizations and readmissions for such problems as intraabdominal infection and vascular thrombosis. There is even a report showing a modest increase in mortality, which is not surprising considering the major surgery being performed in patients with advanced diabetic complications.11 Successful pancreas transplants render recipients normoglycemic and insulin-independent, such that they have normal glucose levels around the clock, normal glycohemoglobin levels, and no dietary restrictions.11a This occurs even though patients may be taking medications such as cyclosporine or tacrolimus that can inhibit insulin secretion,12 and glucocorticoids with their separate diabetogenic effects. Questions have been raised about whether there is more reactive hypoglycemia than in normal subjects, but such episodes are uncommon and rarely significant. Debate continues about how much benefit is provided from pancreas transplants. Most of the recipients already have established complications, so it is not surprising that improvement of these has been modest at best. Various studies have found that there may be some stabilization of retinopathy and improvement in nerve conduction velocity, but these small changes seem to have little clinical impact. Protection of the transplanted kidney from the characteristic histologic lesions of diabetic nephropathy has been demonstrated. It was reported that when the kidneys of patients with PTA were biopsied 5 years after the pancreas transplant, no benefit was seen, but after a biopsy at 10 years there was impressive reversion of histology toward normal.13 A provocative study suggests that patients with autonomic neuropathy before a pancreas transplant have better survival 7 years later than those with failed grafts14; although encouraging, this finding needs to be confirmed. The most obvious benefit of pancreas transplants is that patients believe that their quality of life is improved, particularly with freedom from insulin injections, hypoglycemic episodes, and food restrictions. However, quality of life has proved to be a very difficult parameter to evaluate.2,15 Indeed, it has been difficult to show that patients' lives are truly improved using such standard parameters as whether they are more active or have better performance at work. The most striking finding is that patients are very happy to be free of their diabetes, but it is difficult to know what value to place on this psychological benefit. With pancreas transplantation providing benefits that are counterbalanced by substantial risk and cost, active debate continues about overall value and about how to handle the financial burdens, which are typically more than $100,000 per patient.15 Although there has been a call for controlled clinical trials to answer some of these questions, it will be difficult to organize such large trials; thus, research will probably be restricted to small studies. In the meantime, patients will continue to want pancreas transplants in spite of the risks. The psychological desire for such transplants will continue to be strong, and some patients will benefit from being relieved of the devastating impact of hypoglycemia unawareness. Important questions are being asked about the potential benefits of pancreas transplants alone for patients with early proteinuria, which still has a poor prognosis. Pancreas transplants usually provide truly normal glucose levels, which is considerably better than the control achieved by the Diabetes Control and Complications Trial (DCCT), which had strong protective effects on diabetic complications.1 Until islet transplantation becomes successful and accessible, the potential benefits of pancreas transplantation will continue to be explored.

ISLET TRANSPLANTATION There was great excitement in the early 1970s when the first successful islet transplants were performed in rodents, because many expected that successful routine transplants for patients were only a few years away. Sadly, more than 25 years later, only a very small number of patients have received any benefit from B-cell

replacement therapy. Because of the high expectations, the issue has been emotionally charged, but, in spite of many missteps, substantial progress has been made. Moreover, there are different potential approaches that are emerging, some of which should lead to progress and eventual success. There are many political as well as strategic issues about how to approach the problem, which have been discussed elsewhere.16 HUMAN ISLET ALLOGRAFTS After much work with small and large animal models, it finally became possible in the late 1980s to start human islet allografts in immunosuppressed patients with kidney transplants.17,18,19,20,21 and 22 Some of these early transplants used islets obtained from as many as five donor pancreases, with most being cryopreserved and some being fresh. These islets are injected into the portal vein, whereupon they wedge in the hepatic sinusoids and engraft, receiving their vascular supply from host vessels growing into the islets. Islet allografts have few complications because islets can be delivered to the liver via the portal vein, either with a relatively simple transhepatic angiographic procedure or a more direct approach to the portal vein with dissection along the umbilical vein. Portal hypertension has only rarely been a problem. In spite of the excitement of a few early successes with some recipients who were insulin-free for more than 2 years, the initial results were disappointing. Data from the International Islet Transplantation Registry show that by the end of 1995, of 270 islet allografts, only 27 (10%) of the recipients remained insulin-free for more than a week.23 Some recipients even lost circulating C-peptide levels within a few days, a phenomenon called primary nonfunction, which may be secondary to rapid autoimmune destruction. The reasons for this apparent lack of success are not yet understood, but closer examination of the data indicates that many of the recipients who remained hyperglycemic have benefited from partial graft function, which has led to lower insulin requirements, smoother control, excellent glycohemoglobin levels, and often many fewer problems with hypoglycemia.24 Nonetheless, the lack of success is striking when compared to pancreas transplantation, with which there is often 85% graft function with insulin independence at 1 year. This suggests that the islets contained in one pancreas are sufficient, and that both transplant rejection and autoimmunity can be controlled by conventional immunosuppression. Somehow, the islets contained in their normal home in the pancreas must be less vulnerable to immune injury. It is also possible that the ability of the pancreas to generate new islets from ducts may be helpful. Over the past few years, a more rigorous approach to islet allografts seems to have led to better results in some centers.25 Some of this improvement is likely due to meticulous attention to detail. For example, the pancreas should have cold-ischemia-time of 6000 islet equivalents per kilogram should be given, and every effort should be made to maintain euglycemia with insulin therapy in the early posttransplant period.25a There is some evidence that the use of antithymocyte globulin is beneficial. Because of these encouraging results and the development of some attractive new approaches to immunosuppression, an increasing number of human allografts are expected to be performed in the next few years. ALLOGRAFTS IN THE ABSENCE OF AUTOIMMUNITY Cluster operations for abdominal cancer have been performed in which liver, pancreas, and other organs are removed; then a liver from a cadaver donor is transplanted with islets isolated from the pancreas of the same donor injected into the portal vein.23 These transplants more often produced normoglycemia than allograft islets given to patients with type 1 diabetes, with patients becoming insulin-independent 60% of the time, and one patient being off insulin for 5 years. Although transplantation of the liver may have had some beneficial modulating influence on the immune system, the absence of autoimmunity is suspected to be the major reason for the success. AUTOGRAFTS People with painful pancreatitis sometimes undergo pancreatectomy with resultant insulin-dependent diabetes. There has been some experience in isolating islets from these pancreases and injecting them back into the portal vein. These patients do remarkably well, with ~75% being insulin independent at 1 year.23 Sometimes, even 255 Å, pattern A) or smaller LDL particles (diameter of ³255 Å, pattern B). Individuals with pattern B have higher plasma TG and lower HDL cholesterol concentrations and are at increased risk of CHD.27 Because similar changes in plasma TG and HDL cholesterol concentrations are associated with resistance to insulin-mediated glucose uptake and/or hyperinsulinemia,28 pattern B is also probably associated with insulin resistance. Evidence has now been published demonstrating that healthy volunteers with small, dense LDL particles (pattern B) are relatively insulin resistant, glucose intolerant, hyperinsulinemic, hypertensive, and hypertriglyceridemic, and have a lower HDL cholesterol concentration.29 Thus, this change in LDL composition should be added to the cluster of abnormalities constituting syndrome X. HEMODYNAMICS SYMPATHETIC NERVOUS SYSTEM ACTIVITY Not only is the resting heart rate higher in patients with high blood pressure (HBP), but it also is a predictor of hypertension.30 Moreover, patients with HBP demonstrate resistance to insulin-mediated glucose disposal and/or compensatory hyperinsulinemia compared with normotensive individuals,5 and these changes can also be seen in normotensive first-degree relatives of patients with hypertension.31 A review article has summarized evidence that insulin resistance and compensatory hyperinsulinemia can predispose individuals to develop hypertension via stimulation of the sympathetic nervous system (SNS).32 Thus, the association between hypertension and increased heart rate could be secondary to enhanced SNS activity in insulin-resistant subjects. Evidence in support of this view has come from a study of normotensive, nondiabetic individuals in whom both insulin resistance and the plasma insulin response to glucose were significantly correlated with heart rate.33 Cross-sectional data do not provide proof of causal relationships, but the suggestion that SNS activity is increased in insulin-resistant individuals seems reasonable. This finding may explain why both resistance to insulin-mediated glucose disposal (or hyperinsulinemia) and an increase in heart rate have been shown to predict development of hypertension. SODIUM RETENTION The acute infusion of insulin increases renal sodium retention in both normal individuals and patients with HBP.12,13 Importantly, the ability of insulin to enhance renal sodium absorption was discerned in patients whose muscles are resistant to insulin-mediated glucose disposal.12,13 This represents another instance in which renal sensitivity to an action of insulin is maintained despite a loss of muscle insulin sensitivity, similar to the situation described with regard to renal uric acid clearance. Indeed, these two effects of insulin on the kidney seem to be related, in that the increase in sodium retention is associated with the decrease in uric acid clearance. If insulin enhances renal sodium retention in insulin-resistant subjects, one might predict that such individuals would also be salt sensitive. Although this issue has not

been definitively settled, evidence has been found in both normotensive and hypertensive individuals that the blood pressure is sensitive to salt in insulin-resistant individuals.34 Irrespective of the role played by insulin resistance and/or compensatory hyperinsulinemia in the pathogenesis of salt-sensitive hypertension, however, enhanced renal tubular sodium reabsorption appears to be part of syndrome X. HYPERTENSION Insulin-resistant and hyperinsulinemic individuals are at increased risk of developing HBP34a (see Chap. 82). The considerable evidence in support of this view5,31,32 can be summarized as follows: (a) cross-sectional studies have repeatedly shown that patients with HBP are insulin resistant and hyper-insulinemic compared with well-matched groups of normotensive individuals; (b) patients with secondary forms of hypertension are not insulin resistant, and effective drug treatment of patients with HBP does not restore insulin sensitivity to normal in insulin-resistant patients with hypertension; (c) normotensive first-degree relatives of patients with HBP are insulin resistant and hyperinsulinemic compared with normotensive subjects without a family history of hypertension; and (d) hyperinsulinemia predicts the development of hypertension in prospective studies. On the other hand, probably no more than half of patients with HBP are insulin resistant and hyperinsulinemic,35 and insulin-resistant and hyperinsulinemic individuals with normal blood pressure certainly exist. Thus, any effort to define a causal relationship between insulin resistance, hyperinsulinemia, and hypertension must take into account the fact that the defects in insulin metabolism can only be relevant to the development of hypertension in the subset of patients with HBP. Furthermore, even if insulin resistance and hyperinsulinemia are related to the pathogenesis of hypertension, they most likely play a permissive role, increasing the risk of an individual's developing hypertension. A good deal remains to be learned about the relationship between insulin resistance, hyperinsulinemia, and hypertension, but no doubt remains that an association between these variables exists. Therefore, the inclusion of hypertension in Figure 145-1 is certainly warranted. HEMOSTASIS Concentrations of plasminogen activator inhibitor-1 (PAI-1) are higher in patients with hypertriglyceridemia, hypertension, and CHD.36,37 and 38 Given the association between PAI-1, CHD, and the other features of syndrome X, PAI-1 concentrations possibly are related to insulin resistance and/or compensatory hyperinsulinemia.39,40 Perhaps the most compelling evidence comes from the European Concerted Action on Thrombosis and Disabilities Angina Pectoris Study,38 in which PAI-1 concentrations were found to be significantly associated with hyperinsulinemia, hypertriglyceridemia, and hypertension in 1500 patients with angina pectoris.38 The evidence that elevated fibrinogen levels are also part of syndrome X is not as strong. Although insulin resistance and fibrinogen levels have been shown to be correlated, the argument has been made that the relationship in this case is not an independent one but is the manifestation of an acute-phase reaction in patients with CHD.

CONCLUSION The ability of insulin to stimulate glucose uptake varies widely from person to person. In an effort to maintain the ambient plasma glucose concentration between ~80 and 140 mg/dL, the pancreatic B cell attempts to secrete whatever amount of insulin is required to accomplish this goal (see Chap. 135). The more resistant a normal individual is to insulin-mediated glucose disposal, the greater must be the degree of compensatory hyperinsulinemia. If the B cell cannot sustain this effort, type 2 diabetes ensues. Although hyperinsulinemia may prevent the development of type 2 diabetes, the ability of insulin-resistant individuals to maintain normal or near-normal glucose tolerance by secreting large amounts of insulin is hardly benign. Specifically, the combination of insulin resistance and compensatory hyperinsulinemia appears to lead to the cluster of abnormalities that make up the current version of syndrome X, as shown in Figure 145-1. In this formulation, resistance to insulin-mediated glucose disposal and compensatory hyperinsulinemia are viewed as the central defects. Arguments may continue regarding the relative importance of nature versus nurture in the genesis of insulin resistance, but general agreement exists that insulin resistance leads to an effort on the part of the B cell to secrete more insulin to prevent decompensation of glucose homeostasis. Normal, or near-normal, glucose tolerance can be maintained if insulin-resistant individuals are able to maintain a state of chronic hyperinsulinemia. Unfortunately, the consequences of this “victory” put an individual at greatly increased risk to develop all of the abnormalities shown in Figure 145-1. More to the point, all of the various facets of syndrome X are involved to a substantial degree in the cause and clinical course of CHD. On the basis of the previous considerations, one might suggest that resistance to insulin-mediated glucose disposal, and the manner in which the organism responds to this defect, play major roles in the pathogenesis and clinical course of what are often referred to as diseases of Western civilization. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

Ginsberg H, Kimmerling G, Olefsky JM, Reaven GM. Demonstration of insulin resistance in untreated adult onset diabetic subjects with fasting hyperglycemia. J Clin Invest 1975; 55:454. Reaven GM. Insulin resistance in noninsulin-dependent diabetes mellitus: does it exist and can it be measured? Am J Med 1983; 74(Suppl 1A):3. Reaven GM. Role of insulin resistance in human disease. Diabetes 1988; 37:1595. Bogardus C, Lillioja S, Mott DM, et al. Relationship between degree of obesity and in vivo insulin action in man. Am J Physiol 1985; 248(3 Pt 1):E286. Lillioja SD, Mott DM, Spraul M, et al. Insulin resistance and insulin secretory dysfunction as precursors of non-insulin-dependent diabetes mellitus. N Engl J Med 1993; 329:1988. Reaven GM, Brand RJ, Chen Y-DI, et al. Insulin resistance and insulin secretion are determinants of oral glucose tolerance in normal individuals. Diabetes 1993; 42:1324. Reaven GM, Miller RG. An attempt to define the nature of chemical diabetes using a multidimensional analysis. Diabetologia 1979; 16:17. Gertler MM, Garn SM, Levine SA. Serum uric acid in relation to age and physique in health and coronary heart disease. Am J Med 1951; 34:1421. Wyngaarden JB, Kelley WN. Gout. In: Metabolic basis of inherited disease, 5th ed. New York: McGraw-Hill, 1983:1043. Facchini F, Chen Y-DI, Hollenbeck CB, Reaven GM. Relationship between resistance to insulin-mediated glucose uptake, urinary uric acid clearance, and plasma uric acid concentration. JAMA 1991; 266:3008. Zavaroni I, Vazza S, Fantuzzi M, et al. Changes in insulin and lipid metabolism in males with asymptomatic hyperuricemia. J Intern Med 1993; 234:24. Quinones GA, Natali A, Baldi S, et al. Effect of insulin on uric acid excretion in humans. Am J Physiol 1995; 268:E1. Muscelli E, Natali A, Bianchi S, et al. Effect of insulin on renal sodium and uric acid handling in essential hypertension. Am J Hypertens 1996; 9:746. Reaven GM, Lerner RL, Stern MP, Farquhar JW. Role of insulin in endogenous hypertriglyceridemia. J Clin Invest 1967; 46:1756. Olefsky JM, Farquhar JW, Reaven GM. Reappraisal of the role of insulin in hypertriglyceridemia. Am J Med 1974; 57:551. Tobey TA, Greenfield M, Kraemer F, Reaven GM. Relationship between insulin resistance, insulin secretion, very low density lipoprotein kinetics, and plasma triglyceride levels in normotriglyceridemic man. Metabolism 1981; 30:165. Wilson DE, Chan I-F, Buchi KN, Horton SC. Postchallenge plasma lipoprotein retinoids: chylomicron remnants in endogenous hypertriglyceridemia. Metabolism 1985; 34:551. Chen Y-DI, Swami S, Skowronski R, et al. Differences in postprandial lipemia between patients with normal glucose tolerance and noninsulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1993; 76:172. Jeppesen J, Hollenbeck CB, Zhou M-Y, et al. Relation between insulin resistance, hyperinsulinemia, postheparin plasma lipoprotein lipase activity, and postprandial lipemia. Arterioscler Thromb Vasc Biol 1995; 15:320. Swenson TL. The role of the cholesteryl ester transfer protein in lipoprotein metabolism. Diabetes Metab Rev 1991; 7:139. Fidge N, Nestel P, Toshitsugu I, et al. Turnover of apoproteins A-I and A-II of high density lipoprotein and the relationship to other lipoproteins in normal and hyperlipidemic individuals. Metabolism 1980; 29:643. Chen Y-DI, Sheu WH-H, Swislocki ALM, Reaven GM. High density lipo-protein turnover in patients with hypertension. Hypertension 1991; 17:386. Golay A, Zech L, Shi M-Z, et al. High density lipoprotein (HDL) metabolism in noninsulin-dependent diabetes mellitus: measurement of HDL turnover using tritiated HDL. J Clin Endocrinol Metab 1987; 65:512. Brinton EA, Eisenberg S, Breslow JL. Human HDL cholesterol levels are determined by apoA-I fractional catabolic rate, which correlates inversely with estimates of HDL particle size. Effects of gender, hepatic and lipoprotein lipases, triglyceride and insulin levels, and body fat distribution. Arterioscler Thromb 1994; 14:707. Eckel RH. Lipoprotein lipase: a multifunctional enzyme relevant to common metabolic diseases. N Engl J Med 1989; 320:1060. Maheux P, Azhar S, Kern PA, et al. Relationship between insulin-mediated glucose disposal and regulation of plasma and adipose tissue lipoprotein lipase. Diabetologia 1997; 40:850. Austin MA, Breslow JL, Hennekens CH, et al. Low-density lipoprotein subclass patterns and risk of myocardial infarction. JAMA 1988; 260:1917. Reaven GM, Chen Y-DI, Jeppesen J, et al. Insulin resistance and hyperin-sulinemia in individuals with small, dense, low density lipoprotein particles. J Clin Invest 1993; 92:141. Laws A, Reaven GM. Evidence for an independent relationship between insulin resistance and fasting plasma HDL-cholesterol, triglyceride and insulin concentrations. J Intern Med 1992; 231:25. Selby JV, Friedman GD, Quesenberry CP. Precursors of essential hypertension: pulmonary function, heart rate, uric acid, serum cholesterol and other serum chemistries. Am J Epidemiol 1990; 131:1017. Reaven GM. Hypertension in diabetes. London: Martin Dunitz, 1999; in press. Reaven GM, Lithell H, Landsberg L. Hypertension and associated metabolic abnormalities—the role of insulin resistance and the sympathoadrenal system. N Engl J Med 1996; 334:374. Facchini FS, Stoohs RA, Reaven GM. Enhanced sympathetic nervous system activity—the linchpin between insulin resistance, hyperinsulinemia, and heart rate. Am J Hypertens 1996; 9:1013. Zavaroni I, Coruzzi P, Bonini L, et al. Association between salt sensitivity and insulin concentrations in patients with hypertension. Am J Hypertens 1995; 8:855.

34a.Weston PJ. Insulin resistance and hypertension: is impaired arterial baroreceptor sensitivity the missing link? Clin Science 2000; 98:125 35. Zavaroni I, Mazza S, Dall'Aglio E, et al. Prevalence of hyperinsulinaemia in patients with high blood pressure. J Intern Med 1992; 231:235. 36. Landin K, Tengvory L, Smith U. Elevated fibrinogen and plasminogen activator (PAI-1) in hypertension are related to metabolic risk factors for cardiovascular disease. J Intern Med 1990; 227:273. 37. Juhan-Vague I, Alessel MC, Vague P. Increased plasma plasminogen activator inhibitor 1 levels. A possible link between insulin resistance and athero-thrombosis. Diabetologia 1991; 34:457. 38. Juhan-Vague I, Thompson SG, Jespersen J, on behalf of the ECAT Angina Pectoris Study Group. Involvement of the hemostatic system in the insulin resistance syndrome. Arterioscler Thromb 1993; 13:1865. 39. Valle M, Gascon F, Martas R, et al. Infantile obesity: a situation of athero-thrombotic risk? Metabolism 2000; 49:672. 40. Meige JB, Mittleman MA, Nathan DM, et al. Hyperinsulinemia, hyperglycemia, and impaired hemostasis: the Framingham Offspring Study. JAMA 2000; 283:221.

CHAPTER 146 SYNDROMES OF EXTREME INSULIN RESISTANCE Principles and Practice of Endocrinology and Metabolism

CHAPTER 146 SYNDROMES OF EXTREME INSULIN RESISTANCE JEFFREY S. FLIER AND CHRISTOS S. MANTZOROS Insulin Receptors and Insulin Action Insulin Receptors Insulin Resistance: General Considerations Definitions and in Vivo Assessment of Insulin Resistance Biochemical Basis for Clinical Heterogeneity in Insulin-Resistant States Pathogenetic Mechanisms Responsible for Severe Insulin Resistance Nature and Pathogenetic Basis for Clinical Features Commonly Associated with Severe Insulin Resistance Acanthosis Nigricans Ovarian Dysfunction Specific Syndromes of Extreme Insulin Resistance Syndrome of Insulin Resistance Caused by Autoantibodies to the Insulin Receptor (Type B Insulin Resistance) Type A Syndrome of Insulin Resistance and its Variants Rabson-Mendenhall Syndrome Leprechaunism Lipodystrophic States Pseudoacromegaly Other Complex Syndromes of Insulin Resistance Subcutaneous Insulin Degradation Syndrome Treatment Chapter References

This chapter considers a group of syndromes that have in common severe tissue resistance to the actions of insulin. Insulin resistance that occurs in association with a variety of common states (e.g., obesity, noninsulin-dependent diabetes mellitus [type 2], polycystic ovary syndrome [PCOS], hypertension, and syndrome X1,2 and 3) is not addressed. The insulin resistance that is due to antiinsulin antibodies, as occurs in patients receiving insulin therapy for diabetes, is considered in Chapter 143. In this review of syndromes characterized by severe insulin resistance, the authors discuss the tools and criteria for diagnosis, current knowledge of pathophysiologic mechanisms, clinical phenotypes3,4 (Table 146-1), and current therapeutic approaches.

TABLE 146-1. Clinical Syndromes of Extreme Insulin Resistance

INSULIN RECEPTORS AND INSULIN ACTION Although insulin is best known for its ability to promote glucose metabolism, this hormone exerts a wide variety of effects at the cellular level. In addition to stimulating glucose and amino-acid transport, insulin also can activate or inactivate cytoplasmic and membrane enzymes, alter the rate of synthesis and degradation of various proteins and specific mRNAs, and influence the processes of cell growth and differentiation.1,2 and 3 These multiple effects vary widely from tissue to tissue and in dose-response and time course. Some effects, such as the stimulation of glucose transport activity, occur within seconds at very low insulin concentrations. At the other extreme, actions to promote cell growth in certain cells require hours and generally involve higher concentrations of the hormone. Further complicating the study of insulin action is the relationship between insulin and the so-called insulin-like growth factors (IGFs; see Chap. 12 and Chap. 173). These peptides (IGF-I and IGF-II) have major structural homologies with insulin, but they have little or no immunologic cross-reactivity with the hormone. Both IGF-I and IGF-II have distinct receptors to which insulin also can bind, but with reduced affinity.5 The IGF-I receptor can mediate many of the same acute metabolic events that are regulated by the insulin receptor. Generally, however, the IGFs, acting through the IGF-I receptor, have more potent effects on cell growth than does insulin, acting through the insulin receptor (IR). Some of the actions of insulin that are seen at very high concentrations appear to be exerted through binding to and activation of IGF receptors (rather than IRs). INSULIN RECEPTORS The IR is a glycoprotein composed of two distinct subunits, referred to as a and b, with molecular masses of 135,000 and 95,000 daltons, respectively.6 These two subunits are held together by disulfide bonds and arise from a common precursor proreceptor molecule that is encoded by a single gene. The number of receptors expressed per cell varies considerably, from several hundred per mature erythrocyte to several hundred thousand per adipocyte. Insulin binds to the extracellular subunits of its heterotetrameric receptor and activates the intracellular tyrosine kinase of the transmembrane b subunits6 (Fig. 146-1). This results in IR autophosphorylation and subsequent tyrosine phosphorylation of critical intracellular signaling intermediates. These include IR substrates (IRS-1, IRS-2), Shc, and Gab1.6,7 These molecules bind to and activate other downstream molecules, including the adapter proteins Grb2 and Nck, the tyrosine phosphatase Syp, and the phosphoinositide 3-kinase, which amplify and diversify the initial signal generated by insulin binding to its receptor.6 Several signaling pathways are subsequently activated, including the ras (Grb2-mSOS-Ras)-mitogen–activated protein kinase pathway, the pp70 kinase, the PKB/Akt (protein kinase B), and possibly other unidentified pathways. Activation of these molecules results in the well-documented insulin effects, such as stimulation of cellular glucose and amino-acid uptake, glycogen synthesis, lipogenesis, and mitogenesis.6 The complexity of this downstream cascade, which is far greater than was previously anticipated, is reviewed in more detail in Chapter 135.

FIGURE 146-1. Schematic representation of the insulin receptor and insulin signal transduction. The binding of insulin to the a subunit activates autophosphorylation of tyrosine residues on the b subunit, leading to activation of the intrinsic tyrosine kinase of the receptor. The major known substrate of the insulin receptor tyrosine kinase is IRS-1. Tyrosine phosphorylation of specific tyrosines on IRS-1 causes interaction with molecules such as syp, Nck, phosphatidyl inositol-3 (PI-3) kinase, and Grb through specific domains on these molecules (see text; see also Chap. 135).

RECEPTOR REGULATION IRs are not static components of the cellular machinery; rather, they have a half-life measured in hours. A major factor regulating the concentration of IRs is insulin itself. Thus, when cells are cultured in a medium containing insulin, they exhibit a time- and temperature-dependent decrease in the concentration of IRs, a phenomenon termed down-regulation.8 The mechanism for this may be complex, but typically appears to involve an insulin-induced acceleration of receptor degradation, and the number of IRs on cells has correlated inversely with the concentration of insulin to which the cells are tonically exposed in vivo. Many other modulators of receptor concentration or affinity have been described through in vivo or in vitro studies. These include various physiologic states (e.g., age, diurnal variation, diet, exercise, menstrual cycle, and pregnancy) and drugs (e.g., oral hypoglycemic agents, such as sulfonylureas and the biguanides, and corticosteroids). Other modulators are dietary maneuvers such as fasting or high-carbohydrate feeding, exercise, and the level of specific molecules that can influence receptor expression, such as hormones (cortisol, growth hormone [GH]), nucleotides, ketones, and autoantibodies against the receptor.9 In many diseases, one or more of these receptor modulators may be responsible for insulin receptor alterations and partially responsible for the clinical resistance to insulin (Table 146-2).

TABLE 146-2. Diseases and Clinical States with Insulin Resistance in Which Insulin Receptor Expression or Function May Be Altered

INSULIN RESISTANCE: GENERAL CONSIDERATIONS DEFINITIONS AND IN VIVO ASSESSMENT OF INSULIN RESISTANCE Insulin resistance has been broadly defined as “a state (of a cell, tissue, or organism) in which a greater than normal amount of insulin is required to elicit a quantitatively normal response.”3 Usually, this brings to mind the image of an insulin-treated diabetic patient who remains hyperglycemic despite large doses of exogenous insulin. Although such a patient certainly is insulin resistant, this clinical situation represents only one of many clinical presentations of the insulin-resistant disorders. However, insulin resistance can be selective (i.e., involving only certain aspects of insulin action).1,3 Patients with insulin resistance thus span a broad spectrum of glucose homeostasis: At one end, these patients may be grossly diabetic despite large doses of insulin; at the other end, they may be normoglycemic despite severe insulin resistance, which is overcome by the compensatory hypersecretion of endogenous insulin. Clinical assessment of insulin resistance relies on several tests, which, in order of increasing complexity, include (a) the determination of insulin levels in either the fasted state or after oral glucose-tolerance testing (OGTT), the results of which must be interpreted in the context of plasma glucose levels10; (b) the calculation of the homeostasis (HOMA) index; (c) the assessment of sequential plasma glucose levels after the intravenous administration of insulin (insulin tolerance test)10; (d) the estimation of an index of insulin sensitivity (Si), by applying the minimal model technique to data obtained from the frequently sampled intravenous glucose-tolerance test (FSIVGTT)11; and (e) the measurement of in vivo insulin-mediated glucose disposal by the euglycemic hyperinsulinemic clamp procedure.12 BIOCHEMICAL BASIS FOR CLINICAL HETEROGENEITY IN INSULIN-RESISTANT STATES The diverse cellular actions of insulin result from at least two factors: the generation of multiple distinct effects on postreceptor signaling pathways within target cells and the ability of insulin to bind to and act through the IGF-I receptor as well as through the classic IR. As one consequence of these complex signaling mechanisms, resistance to one action of insulin (e.g., its glucose-lowering effect) need not necessarily be associated with equally severe resistance to other important actions of insulin (e.g., antilipolysis, amino-acid uptake, or growth stimulation). It is likely that heterogeneity in the degree to which insulin action on various cellular pathways is impaired plays a central role in determining the clinical specificity of these heterogeneous disorders. PATHOGENETIC MECHANISMS RESPONSIBLE FOR SEVERE INSULIN RESISTANCE Several different classifications for the pathogenesis of insulin resistance have been proposed. One useful classification divides these disorders into those in which there is “primary” target cell resistance to insulin and those in which insulin resistance is caused by factors apart from the target cell. PRIMARY TARGET CELL DEFECTS Primary target cell defects can be due to defects in the IR itself, or to defects in signaling components apart from the IR (Table 146-3). Disorders at the level of the IR gene have been defined over the past few years.13,14 These may present as a marked reduction in the number of (functionally normal) IRs, due to multiple biochemical defects, including mutations in the receptor gene or its promoter, causing reduced ability to synthesize receptor mRNA, or to mutations in the receptor gene, causing impaired ability of the mature receptor protein to insert into or remain within the membrane. Alternatively, qualitative abnormalities of receptor function have been seen, also resulting from multiple gene defects. These have included receptor structural mutations causing reduced affinity of hormone binding, altered function of the receptor as a hormone-activated kinase, and impaired interaction of an activated receptor with other signaling components.

TABLE 146-3. Mechanisms for Extreme Target Cell Resistance to Insulin

AUTOANTIBODIES Severe insulin resistance also can be caused by several mechanisms apart from primary defects in cellular responsiveness. One well-described mechanism involves the spontaneous development of circulating autoantibodies that recognize determinants in the IR molecule. Anti-IR antibodies occur presumably as a result of either loss of immune tolerance or generation of an immune response to an exogenous antigen and autoanti-body formation through molecular mimicry.3 These antibodies

can lead to insulin resistance by sterically interfering with insulin binding,3 although some anti-IR antibodies appear to lead to IR activation, explaining the fasting hypoglycemia that may occur in these patients.1,3,4 and 5,15,16 ACCELERATED DEGRADATION The biochemical basis for insulin degradation in vivo is imperfectly understood. It is known that the in vivo clearance and subsequent degradation of circulating insulin are, to a large degree, a process that is mediated through the IR.17,18 Several proteolytic enzymes that may be responsible for hormonal degradation subsequent to receptor binding have been identified.17 Whether subcutaneously administered insulin is excessively degraded by extracellular enzymes in some patients is uncertain, and the enzymes responsible for such an activity have not been identified. OTHER POTENTIAL MECHANISMS Theoretically, a mutation in the insulin molecule might produce a molecule that would have characteristics of a competitive antagonist. Mutant insulins have been described,19 but these have been weak agonists that have reduced receptor affinity, without the properties of a competitive antagonist (i.e., they have full bioactivity for any amount of receptor occupancy). However, because these species bind to receptors with low affinity and are poorly cleared from the circulation, these mutant insulins circulate at high concentrations, and an insulin-resistant state can be mistakenly diagnosed. Unlike patients with true insulin-resistant states, patients with such mutant insulins are normally sensitive to exogenous insulin. The second potential mechanism would be a, heretofore undescribed, paracrine or circulating factor (hormonal, metabolic) that might induce a state of insulin resistance. The observation that tumor necrosis factor may be overproduced by adipocytes in obesity and may be capable of inducing insulin resistance is an example of such a mechanism.20 Current evidence indicates that the presence of a dominant negative mutation in human PPAR is associated with severe insulin resistance, diabetes mellitus, and hypertension, indicating that this nuclear receptor (which is responsible for adipocyte differentiation) is important in the control of insulin sensitivity, glucose homeostasis, and blood pressure in humans.21

NATURE AND PATHOGENETIC BASIS FOR CLINICAL FEATURES COMMONLY ASSOCIATED WITH SEVERE INSULIN RESISTANCE Many patients with extreme target tissue resistance to insulin do not have overt diabetes. However, nearly all such patients do manifest one or more of a group of characteristic clinical features that suggest the existence of severe insulin resistance. These features include the skin lesion acanthosis nigricans, ovarian hyperandrogenism, accelerated or impaired linear growth, lipoatrophy/lipohypertrophy, and a variety of others (see Table 146-1). All patients with syndromes of severe insulin resistance share a number of laboratory findings. Among these, hyperinsulinemia, resulting from increased insulin secretion to compensate for the peripheral insulin resistance is by far the most consistent finding.1,3,4 Additionally, impaired glucose tolerance or frank diabetes mellitus commonly, but not universally, occurs at a later stage.1,3,4 These manifestations depend on the ability of the pancreas to compensate for the peripheral insulin resistance by increasing insulin secretion.1,3,4 The molecular basis for the association of these features with severe tissue resistance to insulin is only partially understood, but the clinical importance of these associations is clear. ACANTHOSIS NIGRICANS THE LESION Acanthosis nigricans is a skin lesion characterized by brown, velvety, hyperkeratotic plaques, most often found in the axillae, the back of the neck, and other flexural areas22 (Fig. 146-2A). The condition ranges in severity from minimal cases with mild discoloration of limited areas to extreme cases in which the entire surface of the skin may be heavily involved. The pathologic changes are found primarily in the epidermis. There is a complex folding (papillomatosis) of an overgrown epidermis that, although only slightly thickened, has an increased number of cells per unit surface area (Fig. 146-2B). Other changes noted are hyperkeratosis and an increase in the number of melanocytes, the latter contributing to the darkened appearance (see Chap. 153 and Chap. 218).

FIGURE 146-2. A, Clinical appearance of acanthosis nigricans on the back of the neck. B, Photomicro graph of acanthotic skin, revealing papillomatosis, increased keratin, and thickening of epidermis.

CLINICAL ASSOCIATIONS The clinical associations of acanthosis nigricans fall into two main groups: malignant neoplasms and insulin-resistant states. No feature of the skin lesion itself (i.e., the site or histologic appearance) differentiates between these two groups. Acanthosis nigricans associated with insulin resistance appears to be more common than the form associated with internal malignancy. The lesion is found in all clinical conditions that are characterized by markedly reduced insulin action at the cellular level. These include genetic defects in insulin action, antireceptor antibody-induced insulin resistance, and the more common, and pathogenetically less well-defined, insulin-resistant states, such as those associated with obesity.1,2,3 and 4 The acanthosis that occasionally is present in various endocrinopathies (e.g., Cushing syndrome, acromegaly) also may reflect the insulin resistance that is commonly present in these disorders.22,23 CELLULAR MECHANISMS The precise mechanism responsible for the association between tissue insulin resistance and acanthosis nigricans is unknown. Perhaps the skin lesions are caused by high levels of circulating insulin acting through receptors for IGF in the skin.22 Given the fact that various growth factors are produced by tumors,24 this hypothesis also could account for malignancy-associated acanthosis, if tumor oncogene products were able to activate IGF receptors in the skin. CLINICAL IMPLICATIONS Because of an increased awareness of the association between acanthosis and insulin resistance, this skin condition is being recognized more often by internists, endocrinologists, and gynecologists; it is not as uncommon as previously thought. For example, it has been detected in as many as 10% of women being evaluated for polycystic ovary disease, none of whom had overt diabetes.25 Because multiple molecular defects, ranging from autoimmune receptor antibodies to genetic abnormalities of receptor molecules and obesity, may be responsible for this lesion, the referral of patients with acanthosis nigricans for metabolic evaluation should be considered. The possibility of a malignant tumor always should be raised, especially when the condition develops rapidly in an adult, although experience suggests that it is much less often a sign of a malignant neoplasm (see Chap. 219) than of an insulin-resistant state. The extent of a metabolic evaluation should depend on the clinical context, including the presence or absence of other clinical features of insulin resistance. The measurement of plasma glucose and insulin levels constitutes the minimum evaluation to determine the presence of insulin resistance in patients with acanthosis nigricans; if the glucose and insulin levels are normal, an insulin-resistant state can be ruled out. Hyperinsulinemia, with or without hyperglycemia, is consistent with insulin resistance, and this finding should prompt further studies, such as attempts to detect circulating anti-IR antibodies or studies of IR expression and function using receptors expressed on circulating blood cells or cultured skin fibroblasts. OVARIAN DYSFUNCTION Evidence from various sources suggests that insulin and IGFs are important regulators of ovarian function. Clinically, there is an association between hyperinsulinemic

states of tissue insulin resistance and ovarian hyperandrogenism. Ovarian hyper-androgenism has been seen in a wide range of insulin-resistant states, including genetic disorders of tissue resistance to insulin,26 autoimmune IR deficiency,26 a subset of patients with obesity and polycystic ovary disease in whom insulin resistance may have both genetic and nutritional components,25,26 and 27 and a much larger group of patients with typical polycystic ovary disease in whom a correlation between ambient insulin levels and the degree of hyperandrogenism has been defined.1,2,3 and 4,28 The breadth of this clinical association suggests that insulin may play a surprisingly pervasive role in the pathogenesis of these ovarian disorders.29,30 In further support of this notion, in vitro studies demonstrate that insulin and IGF both have specific receptors in human ovarian cells31,32 and exert multiple effects on ovarian growth and steroidogenesis33,34 and 35 (see Chap. 94, Chap. 96 and Chap. 101). MECHANISM FOR THE ASSOCIATION There are two major hypotheses for the association of ovarian hyperandrogenism and insulin resistance in these patient groups. The first views insulin as exerting its actions through IGF receptors in the ovary, a result of the extremely high circulating levels of insulin. This explanation requires that IGF receptor pathways are normal, or at least less impaired than IR pathways. Alternatively, as has been observed for other metabolic pathways (e.g., persistent antilipolysis, and the absence of ketosis despite marked hyperglycemia in patients with both genetic and immune-mediated insulin resistance), the action of insulin to promote events in the ovary through IRs may be maintained despite the loss of action on pathways related to glucose homeostasis. Additionally, excessive IR serine phosphorylation has been implicated as a potential mechanism for insulin resistance in a subset of PCOS patients.2 PATHOLOGIC FINDINGS Pathologic findings in the ovary are nonspecific, and range from typical findings of polycystic ovary disease to extreme cases of ovarian hyperthecosis. The latter is more prevalent in those patients with the greatest degrees of androgen overproduction. Testosterone levels in some of these patients are sufficiently high (>200 ng/mL) to strongly suggest the existence of an androgen-producing tumor. In such situations, the documentation of insulin resistance should serve to diminish this possibility markedly.

SPECIFIC SYNDROMES OF EXTREME INSULIN RESISTANCE CLINICAL PHENOTYPES Unique features associated with each syndrome, which have been recognized as a result of extensive studies, have led to the classification of patients with severe insulin resistance into several distinct phenotypes. SYNDROME OF INSULIN RESISTANCE CAUSED BY AUTOANTIBODIES TO THE INSULIN RECEPTOR (TYPE B INSULIN RESISTANCE) The existence of IR autoantibodies was first documented during the evaluation of three patients who exhibited extremely insulin-resistant diabetes and acanthosis nigricans.26 At least 50 additional patients have been described since. As is often found in autoimmune disease, the condition is more common in women. Cases have been recorded in various ethnic groups, including Japanese and Mexican, but most cases have been among blacks. The mean age of first diagnosis is between 30 and 40 years, but the diagnosis has been made as early as 12 years and as late as 78 years. Acanthosis nigricans is a characteristic feature, but the skin lesion occasionally has been absent. The most common clinical presentation is symptomatic diabetes, marked by symptoms of polyuria, polydipsia, and weight loss. Although plasma glucose values have varied, levels >300 mg/dL have been common. However, ketoacidosis is generally absent or mild. Resistance to exogenous insulin therapy is the hallmark of the disease, and this typically is noted to be present at the time of initial insulin use (unlike the insulin resistance caused by autoantibodies to insulin). The extent of the insulin resistance is indicated by the marked endogenous hyperinsulinemia, and by the fact that individual patients have failed to respond to insulin in dosages as high as 100,000 U per day. In these circumstances, patients may not derive any benefit from continued insulin therapy. A few patients have only mild glucose intolerance and, in a small subgroup, preexisting insulin-resistant diabetes may be followed by a phase of severe hypoglycemia.36 In other patients, in whom diabetes or glucose intolerance never developed, autoantibodies to the IR have caused hypoglycemia.37 CLINICAL COURSE AND THERAPY Over several years of follow-up, patients with this syndrome have had various outcomes. The spontaneous remission of insulin resistance with disappearance of receptor antibodies has been documented in a few patients.36 In another group, diabetes and severe insulin resistance have persisted for several years, and insulin therapy apparently has been of little or no benefit. Patients with marked hyperglycemia and refractory severe insulin resistance have been treated with various experimental regimens. These have included glucocorticoids,36 antimetabolites, and plasma exchange.38 Given the fluctuating course in the absence of therapy, and the few patients studied, it has been difficult to obtain a strong indication of the degree to which these therapies have been effective. At least 3 of 30 reported patients have died while experiencing spontaneous hypoglycemia after a prior diabetic phase, and although the pathogenetic basis for this transition has not been elucidated, there appears to be a significant risk for the development of hypoglycemia among those who have this condition. AUTOIMMUNE FEATURES Patients with autoantibodies to the IR characteristically also have symptoms or laboratory test results indicative of more widespread autoimmune disease (e.g., leukopenia [>80%], antinuclear antibodies [>80%], elevated sedimentation rate [>80%], elevated serum immunoglobulin G [IgG >80%], proteinuria [50%], alopecia [36%], nephritis [30%], hypocomplementemia [29%], arthritis [20%], and vitiligo [14%]).39 Prominent among these are alopecia, vitiligo, arthralgias and arthritis, Raynaud phenomenon, enlarged salivary glands, elevated sedimentation rate, leukopenia, hypergammaglobulinemia, and a positive antinuclear antibody test result. Approximately one-third of patients meet the established criteria for systemic lupus erythematosus, Sjögren syndrome, or some other distinct autoimmune entity. Among those patients with systemic lupus, lupus nephritis has been seen. Premenopausal women with IR autoantibodies may have ovarian hyperandrogenism of a type similar to that observed in patients with other syndromes of extreme tissue resistance to insulin. CHARACTERISTICS OF ANTIBODIES TO THE INSULIN RECEPTOR The extent of insulin binding to receptors on circulating monocytes of these individuals is markedly reduced and of low affinity, and sera from affected patients can reproduce these findings when exposed to normal IRs in vitro.40 The inhibitory capacity of these sera is due to antibodies, predominantly IgG,41 that bind to the IR molecule and sterically hinder insulin binding. The antibodies also can precipitate IRs from solution. Titers vary over a wide range and have been extremely high in some individuals, typically those with the greatest insulin resistance. Antibodies from these patients inhibit insulin binding to IRs from a wide variety of target tissues and a broad range of animal species, suggesting interaction with a highly conserved region of the receptor molecule. Indeed, a conserved epitope in the IR a subunit has been identified as the site of antibody binding in most cases.42 The antibodies have been polyclonal, and individual populations of antibodies in some sera may show some degree of specificity for receptors on a given target tissue. Essentially all sera from affected patients have proved capable of inhibiting insulin binding to the IR, although sensitive assays based on the ability to precipitate receptors from solution have not been designed. The existence of precipitating antibodies that do not inhibit binding is reported to occur,43 as seen in myasthenia gravis, in which antibodies to acetylcholine receptors of this type are the predominant antibody species. Antibodies from some patients also can inhibit IGF-I binding to its closely related receptor, although the functional significance of these antibodies is unknown. The ability of antibodies to bind to the IR and inhibit insulin binding provides an explanation for the reduced insulin binding and the observed insulin resistance. However, studies of the bioactivity of these antibodies in vitro are more complex. Sudden exposure of cells to antireceptor immunoglobulins elicits a wide range of insulin-like effects.44 This finding raises a potential paradox between in vitro and in vivo observations. A partial resolution of this paradox has come from in vitro studies in which the insulin-mimetic effects are seen to be transient, followed by insulin resistance. The latter is due to postreceptor desensitization of some step that is subsequent to insulin binding, as well as to an enhanced rate of receptor degradation.45 A persistent insulin-like action of these antibodies may account for the hypoglycemia that occurs during the course of the illness in some of these patients, but so far, in vitro examination of the antibodies from hyperglycemic or hypoglycemic patients has not provided an explanation for these differences. The passive transfer of antibodies obtained from one patient to rabbits has caused an insulin-resistant phenotype with postprandial hyperglycemia.46 However, animals so treated also have a tendency to fasting hypoglycemia, and this probably is a reflection of persistent insulin-like properties of the antibodies. TYPE A SYNDROME OF INSULIN RESISTANCE AND ITS VARIANTS CLINICAL FEATURES

The initial description of the type A syndrome of insulin resistance involved three peripubertal, thin women with carbohydrate intolerance or overt diabetes, hyperandrogenism, acanthosis nigricans, and severe target cell resistance to insulin.26 Usually, patients are first seen for evaluation of signs and symptoms of marked hyperandrogenism, acanthosis nigricans, or both. Typically, glucose tolerance in these patients is mildly impaired or even normal. Although patients typically are first seen at about the age of expected puberty, some cases have been discovered much later in life, and when younger siblings of affected patients have been evaluated, insulin resistance has been found at substantially younger ages. Insulin resistance has been observed in a brother of an affected female proband, indicating that the disorder can occur in men, and that androgen excess is not required for the development of insulin resistance.47 However, the apparent absence of hyperandrogenism in affected men removes one of the major presenting complaints in this disorder, and an accurate prevalence in both men and women is unknown. Habitus. The body habitus of these patients is noteworthy. Some patients are thin and others are remarkably muscular. Excessive muscular development may be due in part to hyper-androgenism, but it also may be due to the effects of high plasma concentrations of insulin, possibly acting through IGF-I receptors in muscle. Some patients have had acral enlargement, most notably involving the hands, as well as coarsening of the facial features (Fig. 146-3). A role for insulin, acting through both insulin and IGF-I receptors that are unable to signal increased glucose uptake, but are capable of stimulating other pathways, has been suggested as a cause of this “pseudoacromegaly” in one well-studied patient.48 The acanthosis nigricans has ranged from mild to moderately severe, and usually does not develop before the age of 7 to 10 years. Ovarian hyperandrogenism ranges from moderate to severe, and has tended to be refractory to all the usual therapeutic modalities. Pathologic examination of ovarian tissue in a few individuals has shown ovarian hyperthecosis and stromal hyperplasia.49 In one patient, complete ovariectomy with a resultant fall in androgen levels did not yield any change in insulin sensitivity or IR status. The relationship of this disorder to the more common clinical group with obesity and hyperandrogenism was discussed earlier.

FIGURE 146-3. Progressive coarsening of facial features over a 5-year period in a patient with severe insulin resistance and “pseudoacromegaly.”

CLINICAL PHYSIOLOGY Relatively little information is available on the clinical physiologic function of these patients. The use of euglycemic insulin clamps in several patients has disclosed a severely impaired ability of insulin to promote glucose utilization as well as a marked defect in insulin clearance.18 The latter probably arises because receptor-mediated pathways play an important role in the clearance of insulin. INSULIN ACTION AT THE CELLULAR LEVEL Insulin action at the cellular level has been studied intensively in a few patients with these syndromes. Although it was anticipated that the clinical phenotypes seen in individual patients with distinct insulin-resistant syndromes each would be associated with a unique abnormality at the level of the target cell, the current molecular understanding of these syndromes has not yet provided such a correlation. Three patients with distinct insulin-resistant syndromes have been shown to have different mutations in the IR gene.50,51 and 52 Studies of insulin action have involved IR binding on various cell types, studies of the IR kinase, and studies of insulin action on classic biochemical pathways. Insulin-binding studies can be divided into two types: those performed with freshly obtained cells (most commonly monocytes53 and red cells,54 but occasionally adipocytes) and those performed with cultured cells. Studies with fresh cells most closely reflect the in vivo milieu, and studies with cultured skin fibroblasts55,56 or Epstein-Barr virus– transformed lymphoblasts57 permit an assessment of the genetic component of the defect. At the level of IR binding, three categories of abnormalities have been observed: a markedly reduced number of IRs that are otherwise normal in affinity, receptors that bind insulin with altered affinity, and receptors that are normal in both number and affinity of insulin binding. The latter group may be the most common variety. With the discovery that the IR is a tyrosine protein kinase that is autophosphorylated when insulin binds, the kinase function of the IR in these patients became the subject of intense scrutiny. Studies have been performed in circulating monocytes and erythrocytes,54 as well as in cultured fibroblasts56 and Epstein-Barr virus–transformed lymphoblasts.58 As expected, patients with markedly decreased binding have decreased kinase activity. More interesting is that several patients with normal binding have reduced kinase activity, consistent with a role for this biochemical function in transduction of the insulin signal.56,59,60 The normalcy of insulin binding and kinase activity in other patients is consistent with a defect that is truly beyond the level of the IR. Studies of insulin action in freshly isolated adipocytes or cultured fibroblasts of these patients are limited. Defects in insulin-stimulated glucose transport or utilization have been demonstrated,61 but limited data are available on other pathways of insulin or of IGF action. Family studies indicate an autosomal dominant or autosomal recessive pattern of transmission of the type A syndrome, with variable penetrance.1,3,4 Genetic studies have revealed that many patients with the type A syndrome have mutations at the IR gene locus that typically alter the expression or function of one allele.13,14 Although several IR mutations have been previously associated with the type A syndrome,7,62 it currently appears that most patients with this syndrome do not possess such mutations, implying the presence of other critical primary defects in insulin signaling,60,61,62 and 63 as seems to be the case in patients with type 2 diabetes.64 In addition, a transmembrane glycoprotein named PC-1, which interacts with the a subunit of the IR and subsequently inhibits IR function, has been implicated in the pathogenesis of the type A syndrome and type 2 diabetes.7,65,66 However, its significance remains to be conclusively demonstrated.7 RABSON-MENDENHALL SYNDROME Another very rare syndrome associated with severe insulin resistance (initially described by Mendenhall)3,67 is currently known as the Rabson-Mendenhall syndrome. These patients present in childhood with severe insulin resistance and diabetes mellitus (commonly refractory to large doses of insulin), acanthosis nigricans, abnormal nails and dentition, short stature, protruberant abdomen, precocious pseudopuberty, and, ostensibly, pineal hyperplasia.3,67 Prognosis is generally poor, mainly due to the development of severe microvascular complications of diabetes.3 LEPRECHAUNISM Leprechaunism is a rare inherited disease characterized by an unusual facial appearance (Fig. 146-4), intrauterine and postnatal growth retardation, sparse subcutaneous fat, hirsutism, clitoromegaly, and early death.68 Patients have abnormalities of glucose homeostasis as well as fasting hypoglycemia associated with B-cell hypertrophy and marked endogenous hyperin-sulinemia. Additionally, affected female infants commonly have hirsutism and clitoromegaly, whereas affected boys commonly present with penile enlargement.3,68 Other features of this syndrome include dysmorphic lungs, renal disease, and breast hyperplasia.3 Few of these infants live beyond the first year of life, although some may survive until adolescence.3,68 Cultured cells from several patients have shown heterogeneous abnormalities of insulin action,69 IR binding,70 and IR function, as well as defects in IGF-I receptor pathways.69 All patients studied to date have had mutations affecting the expression or function of both alleles of the IR gene, 14 and patients with no functional IRs have been described.

FIGURE 146-4. Neonate with leprechaunism, displaying characteristic facies and muscle wasting.

LIPODYSTROPHIC STATES The lipodystrophic states are a phenotypically diverse group of syndromes characterized by either complete or partial lack of adipose tissue, often severe abnormalities of carbohydrate and lipid metabolism, and various associated somatic features. The disorders are considered here because of their frequent coexistence with severe tissue resistance to insulin, as well as a spectrum of clinical features that overlap with those seen in other syndromes of severe insulin resistance. CLINICAL FEATURES The lipodystrophy syndromes represent a diverse group of disorders characterized by severe insulin resistance and associated with severe hypertriglyceridemia, which lead to pancreatitis and fatty infiltration of the liver, thereby eventuating with cirrhosis.3,10 The extent of fat loss predicts the severity of metabolic complications. These syndromes have been conveniently subclassified according to the extent and location of the lipo-dystrophy and the age of onset3,10 (Table 146-4).

TABLE 146-4. Classification of Lipodystrophic Disorders

Congenital Forms of Total Lipodystrophy. In its congenital form, the lipoatrophy may be generalized (transmitted as an autosomal recessive trait) or partial (transmitted as an autoso-mal dominant trait). More than 40 cases of the congenital total lipoatrophy syndrome (also known as the Berardinelli-Seip syndrome) have been described.71 The locus for one gene associated with this syndrome was found to map to human chromosome 9q34.72 Parental consanguinity is high in these patients, in whom the disease occurs equally in boys and girls; it usually is diagnosed at birth or within the first 2 years of life. Newborns or infants with congenital generalized lipodystrophy (Berar-dinelli-Seip syndrome) lack adipose tissue completely in both subcutaneous and visceral locations but have normal adipose tissue in areas of mechanical adipose tissue, such as the orbits, hands, palms, and so forth.3 Patients have accelerated linear growth, accelerated genital maturation, muscle hypertrophy, and various other congenital defects but no structural brain abnormalities (Table 146-5). Insulin resistance, as assessed by elevated fasting insulin levels, develops or becomes manifest between the ages of 6 and 9 years, preceding the development of diabetes by several years. Hepatosplenomegaly, hypertriglyceridemia, and decreased low-density lipoprotein (LDL) are common, whereas high-density lipoprotein (HDL) is usually normal. Hepatic cirrhosis, which develops in the context of fatty liver, is a major cause of morbidity and mortality.

TABLE 146-5. Clinical Features of Patients with Lipodystrophic Disorders

Acquired Forms of Total Lipodystrophy. Lipoatrophy also can be either total or partial. Acquired total lipoatrophy is not known to be inherited, and can first develop in childhood or adulthood (Fig. 146-5). In contrast to the Berardinelli-Seip syndrome, patients with acquired total lipodystrophy (Lawrence syndrome) appear normal at birth, but develop lipoatrophy over days to weeks, sometimes after an infectious prodrome.3 Histologic evidence of panniculitis has suggested an inflammatory etiology for this syndrome, although this remains to be demonstrated.3 Although alterations in linear growth rate usually are not found in this disorder, the other associated features typical of congenital lipoatrophy are commonly seen, including insulin-resistant diabetes, acanthosis nigricans, muscle hypertrophy, hepatic cirrhosis, and an increased metabolic rate.

FIGURE 146-5. A patient with acquired total lipoatrophy, accompanied by muscular hypertrophy, cirrhosis, and characteristic curly hair.

Congenital and Acquired Forms of Partial Lipodystrophy. In addition to the above variants of generalized lipodystrophy, several forms of partial lipodystrophy have been recognized and affect specific body areas (Fig. 146-6). Thus, face-sparing lipodystrophy (Dunningan variety), initially reported as an X-linked but increasingly reported as an autosomal dominant condition, spares the face, which is typically full, in contrast to the lipoatrophic trunk and extremities.3,73 The gene for the autosomal dominant form of this syndrome is located on chromosome 1q21–22,74 and mutations of the LMNA gene encoding nuclear lamins A and C have been proposed to mediate this degenerative disorder of adipose tissue.75 Patients develop hypertriglyceridemia and hyperchylomicronemia that can result in pancreatitis. Another variety of the familial partial lipodystrophies is Kobberling syndrome, in which the loss of adipose tissue is restricted to the extremities. Patients may have normal amounts of visceral fat and may even have excessive amounts of subcutaneous truncal fat. Another form of partial lipodystrophy, which occurs in association with mandibuloacral dysplasia and joint contractures, is termed lipodystrophy with other dysmorphic features.3 Additionally, a sporadic form of partial lipodystrophy, named cephalothoracic lipodystrophy, which has been described predominantly in women, occurs in association with mesangiocapillary glomerulonephritis, presumably as a result of complement activation3 (Fig. 146-7). Several of these patients have autoimmune abnormalities (including a distinct alteration in complement metabolism), with accelerated catabolism of C3 as well as a serum IgG called C3NeF in ~90% of patients. The mechanistic link between this defect and the lipoatrophy remains an enigma. Patients infected with human immunodeficiency virus (HIV) being treated with the highly effective HIV-1 protease inhibitors develop lipodystrophy characterized by loss of subcutaneous adipose tissue from the extremities and face and deposition of excess fat in the neck and trunk. These patients have insulin resistance and develop hyperglycemia and hyperlipidemia sooner and more frequently than do patients who are on other regimens for HIV.76 The mechanism underlying the development of lipodystrophy and insulin resistance remains to be elucidated.

FIGURE 146-6. Posterior view of a patient with distal extremity lipo-hypertrophy and proximal extremity and truncal lipodystrophy.

FIGURE 146-7. Patient with acquired partial (face and upper extremity) lipoatrophy with hypocomplementemic nephritis.

Localized Lipodystrophies. Finally, localized lipodystrophies are characterized by a loss of subcutaneous adipose tissue from small areas or from small parts of a limb, but these patients do not develop insulin resistance or metabolic abnormalities. Drug-induced lipodystrophy was a frequent complication before the availability of purified human insulin, but is rather uncommon today. Other rare causes of localized lipodystrophy are due to repeated pressure and panniculitis or as part of a rare syndrome called lipodystrophia centrifugalis abdominalis infantilis. INSULIN RESISTANCE IN LIPODYSTROPHIC STATES The pathogenesis of the insulin resistance in lipoatrophic diabetes is poorly understood, and efforts to define it are complicated by the marked heterogeneity within this group of disorders. As in the type A and B syndromes of insulin resistance, initial efforts to understand the lipodystrophic states have focused on the IR or postreceptor molecules. Although an autosomal recessive mode of transmission has been suggested for the Berar-dinelli-Seip syndrome,3,10 the pathogenesis of associated insulin resistance is poorly understood, and it remains unclear whether insulin resistance is primary or occurs secondary to lipodystrophy. Linkage analysis in ten families with congenital lipodystrophy failed to implicate 14 candidate genes (including the IR, IRS-1, and IGF-I genes).77 Studies of cultured fibroblasts obtained from patients with congenital total lipoatrophy reveal either a modest reduction in insulin binding or normal insulin-binding characteristics.78,79 Likewise, studies of IRs on circulating monocytes and erythrocytes have yielded heterogeneous results, including normal binding, decreased binding associated with a reduced number of receptors, and reductions in receptor affinity.80 To further complicate our understanding, affected persons within the same family have different patterns of insulin binding, suggesting that the observed receptor abnormalities may be secondary to other unknown abnormalities. PSEUDOACROMEGALY Another rare syndrome of severe insulin resistance is associated with acromegaloidism.48 In addition to severe insulin resistance, these patients have features reminiscent of acromegaly, including coarse facies and bone thickening, despite a GH– IGF-I axis that appears to be normal.3,48 However, whether these physical findings result from high insulin levels signaling through the IGF-I receptor or, alternatively, the IR gene per se remains to be established.3,48 Selective impairment of insulin-stimulated phosphoinositide 3-kinase activity has been demonstrated in three patients with severe insulin resistance and pseudoacromegaly.81 OTHER COMPLEX SYNDROMES OF INSULIN RESISTANCE Finally, a number of rare genetic syndromes are associated with severe insulin resistance.82 Among them, Alstrom syndrome, an autosomal recessive disorder, which presents with retinitis pig-mentosa, sensorineural deafness, hypogonadism, and obesity, is commonly associated with severe insulin resistance and acanthosis nigricans.3 Myotonic dystrophy, an autosomal dominant condition that presents with progressive muscular dystrophy, myotonia, mild mental retardation, baldness, cataracts, and postpubertal testicular atrophy, has been associated with severe insulin resistance.3,82a Werner syndrome, a progeria syndrome, presents with bird-like facies, gray hair, cataract formation, slender extremities, and severe insulin resistance.83,83a SUBCUTANEOUS INSULIN DEGRADATION SYNDROME Among insulin-treated diabetics, a subgroup has been described in whom the subcutaneous administration of insulin in high doses is ineffective, whereas insulin administered by the intravenous route produces a normal response.84 A role for subcutaneous insulin-degrading activity in the etiology of this syndrome is suggested both by direct assay of such activity in subcutaneous tissue and by the therapeutic response observed when insulin is coinjected with the protease inhibitor aprotinin (Trasylol). Other causes for insulin resistance have been excluded in such patients, some of whom apparently have required prolonged administration of insulin by the intravenous route. The course is one of spontaneous exacerbations and remissions. The initial patient with this syndrome was extremely well documented. However, in a follow-up study of 20 patients referred for evaluation of this entity, subcutaneous insulin degradation could not be documented.17,85 Instead, these patients had problems with compliance or other emotional problems responsible for their difficulty with insulin therapy. Thus, such problems should be evaluated carefully in patients referred for evaluation of resistance to subcutaneous insulin with sensitivity to intravenous insulin.

TREATMENT Since the pathogenesis of the syndromes of severe insulin resistance is incompletely understood, available therapies are non-specific. Diet, which is the first-line treatment option for diabetes mellitus,86 was not effective in a small study of women with severe insulin resistance.86 Additionally, it is unknown whether exercise has a beneficial effect. Thus, the roles of diet and exercise in these syndromes need further study. Although it appears unlikely that the commonly lean, insulin-resistant individuals will significantly benefit from caloric restriction and exercise, it is prudent to recommend long-term caloric restriction to patients with obesity and polycystic ovary disease, in whom the insulin resistance probably is multifactorial and less severe. In several cases, this has caused improvement in the insulin resistance as well as the acanthosis nigricans and hyperandrogenism.25 Drugs for patients with severe insulin resistance syndromes are limited. Insulin, administered in very high doses, usually fails to provide adequate control.3,86 Similarly, administration of sulfonylureas to patients with severe insulin resistance does not result in significant benefits.86 Metformin, an insulin sensitizer biguanide that suppresses hepatic glucose output and increases insulin-mediated glucose disposal, improved glycemia in patients with the type B syndrome or lipoatrophic diabetes, but did not improve the insulin resistance in patients with myotonic dystrophy.86 In addition, insulin sensitizers (i.e., metformin and thiazolidinediones) are effective treatments for the insulin resistance associated with polycystic ovarian disease.86,87 and 88 Administration of IGF-I, which acts either through the IGF-I receptor or through a functioning IR, in patients with the type A or B syndromes, the Rabson-Mendenhall syndrome, leprechaunism, and lipodystrophy, has led to improvement in glycemic control and a decrease in fasting insulin levels in short-term studies.86,89,90 Some of these effects were not maintained in a 10-week trial, however.89 IGF-I administration is infrequently associated with acute side effects (e.g., fluid retention, carpal tunnel syndrome, jaw pain) and may exacerbate the development of microvascular complications (particularly retinopathy)86; increased endogenous IGF-I levels have been associated with several common malignancies (i.e., colon, breast, and prostate cancer). 91,92,93,94,95 and 96 Thus, the efficacy-safety profile of IGF-I requires further study in patients with severe insulin resistance. Thiazolidinediones, which improve insulin resistance in patients with Werner syndrome,97 are being studied in individuals with other syndromes of severe insulin resistance. Administration of vanadate or vanadium salts to patients with type 2 diabetes has had beneficial effects in insulin resistance,86,98 but their roles in patients with severe insulin resistance remain unclear. Limited data suggest an improvement in insulin sensitivity in response to administration of phenytoin to patients with the type A syndrome,86 and in response to administration of bezafibrate99 or dietary supplementation with n-3 fatty acid– rich fish oil86 in patients with lipodystrophy. Finally, immuno-suppressants and plasmapheresis have been tried with good results in patients with the type B syndrome.3 In addition, in patients with ovarian hyperandrogenism and insulin resistance caused by anti-IR antibodies, the ovarian lesion remits when antireceptor antibodies disappear, whether spontaneously or because of immunosuppressive therapy. Surgical treatment has been recommended for certain patients with insulin-resistant disorders. Women with clearly genetic syndromes of insulin resistance, in whom ovarian hyperandrogenism is often the most severe clinical complaint, have not responded to the usual therapies for polycystic ovary disease, but may respond to wedge resection. In a few patients, complete ovariectomy has been required, with a consequent marked reduction in androgen levels and, as expected, no change in the insulin action defect.100 Finally, cosmetic surgery is an option for patients with lipodystrophy.74 CHAPTER REFERENCES 1. Moller DE, Flier JS. Insulin resistance: mechanisms, syndromes, and implications. N Engl J Med 1991; 325:938. 2. Dunaif A, Xia J, Book CB, et al. Excessive insulin receptor serine phosphorylation in cultured fibroblasts and in skeletal muscle. A potential mechanism for insulin resistance in the polycystic ovary syndrome. J Clin Invest 1995; 96:801. 3. Mantzoros CS, Flier JS. Insulin resistance: the clinical spectrum. In: Maz-zaferi E, ed. Advances in endocrinology and metabolism, vol 6. St. Louis: Mosby–Year Book, 1995:193. 4. Kahn CR, Flier JS, Bar RS, et al. The syndromes of insulin resistance and acanthosis nigricans: insulin receptor disorders in man. N Engl J Med 1976; 294:739. 5. Rechler MM, Nissley SP. The nature and regulation of the receptors for insulin-like growth factors. Annu Rev Med 1985; 47:425. 6. Cheatham B, Kahn CR. Insulin action and the insulin signaling network. Endocr Rev 1995; 16:117. 7. Baynes KCR, Whitehead J, Krook A, O'Rahilly S. Molecular mechanisms of inherited insulin resistance. Q J Med 1997; 90:557. 8. Gavin JR III, Roth J, Neville DM Jr. Insulin dependent regulation of insulin receptor concentrations: a direct demonstration in cell culture. Proc Natl Acad Sci U S A 1974; 71:84. 9. Grunberger G, Taylor SI, Doris RF, Gorden P. Insulin receptors in normal and disease states. J Clin Endocrinol Metab 1983; 12:191. 10. Vidal-Puig A, Moller DE. Insulin resistance: classification, prevalence, clinical manifestations, and diagnosis. In: Azziz R, Nestler JE, Dewailly D, eds. Androgen excess disorders in women. Philadelphia: Lippincott–Raven, 1997:227. 11. Bergman RN. Toward physiological understanding of glucose tolerance: minimal model approach. Diabetes 1989; 38:1512. 12. Bergman RN, Prager R, Volund A, Olefsky JM. Equivalence of the insulin sensitivity model method and the euglycemic glucose clamp. J Clin Invest 1987; 79:790. 13. Flier JS. Syndromes in insulin resistance: mechanisms, syndromes and implications. N Engl J Med 1991; 325:935. 14. Taylor SI. Lilly Lecture: molecular mechanisms of insulin resistance. Lessons from patients with mutations in the insulin-receptor gene. Diabetes 1992; 41:1473. 15. O'Brien TD, Rizza RA, Carney JA, Butler PC. Islet amyloidosis in a patient with chronic massive insulin resistance due to antiinsulin receptor antibodies. J Clin Endocrinol Metab 1994; 79:290. 16. Taylor SI, Grunberger G, Marcus-Samuels B, et al. Hypoglycemia associated with antibodies to the insulin receptor. N Engl J Med 1982; 307:1422. 17. Duckworth WC, Bennett RG, Hamel FG. Insulin degradation: progress and potential. Endocr Rev 1998; 19(5):608. 18. Flier JS, Minaker KL, Landsberg L, et al. Impaired in vivo insulin clearance in patients with target cell resistance to insulin. Diabetes 1982; 31:132. 19. Haneda M, Polonsky KS, Tager HS, et al. Familial hyperinsulinemia due to a structurally abnormal insulin: definition of an emerging new clinical syndrome. N Engl J Med 1984; 310:1288. 20. Hotamisligil GS, Shargill NS, Spiegelman BM. Adipose expression of tumor necrosis factor-alpha: direct role in obesity-linked insulin resistance. Science 1993; 259:87. 21. Barosso I, Gurnell M, Crowley VEF, et al. Dominant negative mutations in human PPAR associated with severe insulin resistance, diabetes mellitus and hypertension. Nature 1999; 402:880. 22. Flier JS. The metabolic importance of acanthosis nigricans. Arch Dermatol 1985; 121:193. 23. Ober KP. Acanthosis nigricans and insulin resistance associated with hypothyroidism. Arch Dermatol 1985; 121:229. 24. Bishop JM. The molecular genetics of cancer. Science 1987; 235:305. 25. Flier JS, Eastman RC, Minaker KL, et al. Acanthosis nigricans in obese women with hyperandrogenism: characterization of an insulin-resistant state distinct from the type A and type B syndromes. Diabetes 1985; 34:101. 26. Kahn CR, Flier JS, Bar RS, et al. The syndromes of insulin resistance and acanthosis nigricans: insulin receptor disorders in man. N Engl J Med 1976; 294:739. 27. Peters EJ, Stuart CA, Prince MJ. Acanthosis nigricans and obesity: acquired and intrinsic defects in insulin action. Metabolism 1986; 35:807. 28. Poretsky L, Kalin MF. The gonadotropic function of insulin. Endocr Rev 1987; 8:132. 29. Poretsky L. On the paradox of insulin-induced hyperandrogenism in insulin resistant states. Endocr Rev 1991; 12:3. 30. Dunaif A. Insulin resistance and ovarian dysfunction in insulin resistance. In: Moller D, ed. Insulin resistance. New York: John Wiley and Sons, 1993:301. 31. Poretsky L, Grigorescu F, Seibel M, et al. Distribution and characterization of insulin and insulin-like growth factor I (IGF-I) receptors in normal human ovary. J Clin Endocrinol Metab 1985; 61:728. 32. Adashi EY, Resnick CE, Hernandez ER, et al. Characterization and regulation of a specific cell membrane receptor for somatomedin-C insulin-like growth factor I in cultured rat granulosa cells. Endocrinology 1988; 122:194. 33. Veldhuis JD, Kolp LA, Toaff ME, et al. Mechanisms subserving the trophic actions of insulin on ovarian cells. J Clin Invest 1983; 72:1046. 34. Veldhuis JD, Nestler JE, Strauss JF III. The insulin-like growth factor, somatomedin-C, modulates low density lipoprotein metabolism by swine granulosa cells. J Endocrinol 1987; 113:21. 35. Veldhuis JD, Rodgers RJ. Mechanisms subserving the steroidogenic synergism between follicle-stimulating hormone and insulin-like growth factor I (somatomedin C). Alterations in cellular sterol metabolism in swine granulosa cells. J Biol Chem 1987; 262:7658. 36. Flier JS, Bar RS, Muggeo M, et al. The evolving clinical course of patients with insulin receptor antibodies: spontaneous remission or receptor proliferation with hypoglycemia. J Clin Endocrinol Metab 1978; 47:985. 37. Taylor SI, Grunberger G, Marcus-Samuels B. Hypoglycemia associated with antibodies to the insulin receptor. N Engl J Med 1982; 307:1422. 38. Muggeo M, Flier JS, Abrams RA, et al. Treatment by plasma exchange of a patient with autoantibodies to the insulin receptor. N Engl J Med 1979; 300:477. 39. Tsokos GC, Gorden P, Antonovych T, et al. Lupus nephritis and other autoimmune features in patients with diabetes mellitus due to autoanti-body to insulin receptors. Ann Intern Med 1985; 102:176. 40. Flier JS, Kahn CR, Roth J, Bar RS. Antibodies that impair insulin receptor binding in an unusual diabetic syndrome with severe insulin resistance. Science 1975; 190:63. 41. Flier JS, Kahn CR, Jarrett DB, Roth J. Characterization of antibodies to the insulin receptor: a cause of insulin resistant diabetes in man. J Clin Invest 1976; 58:1442. 42. Zhang B, Roth RA. A region of the insulin receptor important for ligand binding (residues 450–601) is recognized by patients' autoimmune antibodies and inhibitory monoclonal antibodies. Proc Natl Acad Sci USA 1991; 88:9858. 43. Bloise W, Wajchenberg BL, Moncada VY, et al. Atypical antiinsulin antibodies in a patient with type B insulin resistance and scleroderma. J Clin Endocrinol Metab 1989; 68:227. 44. Kahn CR, Baird KL, Flier JS, Jarrett DB. Effect of anti-insulin receptor antibodies on isolated adipocytes. J Clin Invest 1977; 60:1094. 45. Taylor SI, Marcus-Samuels B. Anti-receptor antibodies mimic the effect of insulin to down regulate insulin receptors in cultured human lymphoblastoid cells. J Clin Endocrinol Metab 1984; 58:182. 46. Dons RF, Havlik R, Taylor SI, et al. Clinical disorders associated with autoantibodies to the insulin receptor. Stimulation by passive transfer of immunoglobulins to rats. J Clin Invest 1983; 72:1072. 47. Flier JS, Young JB, Landsberg L. Familial insulin resistance with acanthosis nigricans, acral hypertrophy and muscle cramps: a new syndrome. N Engl J Med 1980; 390:970. 48. Flier JS, Moller DE, Moses AC, et al. Insulin-mediated pseudoacromegaly: clinical and biochemical characterization of a syndrome of selective insulin resistance. J Clin Endocrinol Metab 1993; 76:1533. 49. Flier JS. Virilization and hyperpigmentation in a 15 year old girl. N Engl J Med 1982; 306:1537. 50. Kadowski T, Bevins CL, Cama A, et al. Two mutant alleles of the insulin receptor gene in a patient with extreme insulin resistance. Science 1988; 240:787. 51. Yoshimasa Y, Seino S, Whittaker J, et al. Insulin-resistant diabetes due to a point mutation that prevents proreceptor processing. Science 1988; 240:784. 52. Moller DE, Flier JS. Detection of an alteration in the insulin-receptor gene in a patient with insulin resistance, acanthosis nigricans, and the polycystic ovary syndrome. N Engl J Med 1988; 319:1526. 53. Bar RS, Muggeo M, Kahn CR, et al. Characterization of the insulin receptor in patients with syndromes of insulin resistance and acanthosis nigricans. Diabetologia 1980; 18:209. 54. Grigorescu F, Flier JS, Kahn CR. Characterization of binding and phosphorylation defects of insulin receptors in the type A syndrome of insulin resistance. Diabetes 1986; 35:127. 55. Podskalny JM, Kahn CR. Cell culture studies on patients with extreme insulin resistance. I. Receptor defects on cultured fibroblasts. J Clin Endocrinol Metab 1982; 54:261. 56. Grigorescu F, Flier JS, Kahn CR. Defect in insulin receptor phosphorylation in erythrocytes and fibroblasts associated with severe insulin resistance. J Biol Chem 1984; 259:15003. 57. Taylor SI, Samuels B, Roth J. Decreased insulin binding in cultured lymphocytes from two patients with extreme insulin resistance. J Clin Endocrinol Metab 1982; 54:919. 58. Whittaker J, Zick Y, Roth J, Taylor SI. Insulin-stimulated receptor phosphorylation appears normal in cultured Epstein-Barr virus-transformed lymphocyte cell lines derived from patients with extreme insulin resistance. J Clin Endocrinol Metab 1985; 60:381. 59. Grunberger G, Zick Y, Gorden P. Defect in phosphorylation of insulin receptors in cells from an insulin-resistant patient with normal insulin binding. Science 1984; 223:932.

60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82.

Moller DE, Cohen O, Yamaguchi Y, et al. Prevalence of mutations, in the insulin receptor gene in subjects with features of the type A syndrome of insulin resistance. Diabetes 1994; 43:247. Podskalny JM, Kahn CR. Cell culture studies on patients with extreme insulin resistance. II. Abnormal biological responses in cultured fibroblasts. J Clin Endocrinol Metab 1982; 54:269. Moller DE, Cohen O, Yamaguchi Y, et al. Prevalence of mutations of the insulin receptor gene in subjects with features of the type A syndrome of insulin resistance. Diabetes 1994; 43:247. Krook A, Kumar S, Laing I, et al. Molecular scanning of the insulin receptor gene in syndromes of insulin resistance. Diabetes 1994; 43:357. Krook A, O'Rahilly S. Mutant receptors in syndromes of insulin resistance. Clin Endocrinol Metab 1996; 10:97. Maddux BA, Goldfine ID. Membrane glycoprotein PC-1 inhibition of insulin receptor function occurs via direct interaction with the receptor alpha-subunit. Diabetes 2000; 49(1):13. Sbraccia P, Goodman PA, Maddux BA, et al. Production of an inhibitor of insulin receptor tyrosine kinase in fibroblasts from a patient with insulin resistance and NIDDM. Diabetes 1991; 40:295. Mendenhall EN. Tumor of the pineal gland with high insulin resistance. J Indiana State Med Assoc 1950; 43:32. Donohue WL, Uchida I. Leprechaunism: a euphemism for a rare familial disorder. J Pediatr 1954; 45:505. Knight AB, Rechler MM, Romanus JA, et al. Stimulation of glucose incorporation and amino acid transport by insulin and an insulin-like growth factor in fibroblasts with defective insulin receptors cultured from a patient with leprechaunism. Proc Natl Acad Sci U S A 1981; 78:2554. Taylor SI, Roth J, Blizzard RM, Elders MJ. Qualitative abnormalities in insulin binding in a patient with extreme insulin resistance: decreased sensitivity to alterations in temperature and pH. Proc Natl Acad Sci U S A 1981; 76:7157. Berardinelli W. An undiagnosed endocrinometabolic syndrome: report of 2 cases. J Clin Endocrinol Metab 1954; 14:193. Garg A, Wilson R, Barnes R, et al. A gene for congenital generalized lipo-dystrophy maps to human chromosome 9q34. J Clin Endocrinol Metab 1999; 84(9):3390. Kobberling J, Dunningan MG. Familial partial lipodystrophy: two types of an X linked dominant syndrome, lethal in the hemizygous state. J Med Genet 1986; 23:120. Peters JM, Barnes R, Bennett L, et al. Localization of the gene for familial partial lipodystrophy (Dunningan variety) to chromosome 1q21–22. Nat Genet 1998; 18:292. Cao H, Hegele RA. Nuclear lamin A/C R482Q mutation in Canadian kindreds with Dunningan-type familial partial lipodystrophy. Hum Molec Genet 2000; 9:109. Tsiodras S, Mantzoros C, Hammer S, Samore M. Effects of protease inhibitors on hyperglycemia, hyperlipidemia and lipodystrophy. A five-year cohort study. Arch Intern Med 2000; (in press). Vigouroux C, Khallouf E, Bourut C, et al. Genetic exclusion of 14 candidate genes in lipoatrophic diabetes using linkage analysis in 10 consanguineous families. J Clin Endocrinol Metab 1997; 82:3438. Oseid S. Decreased binding of insulin to its receptor in patients with congenital generalized lipodystrophy. N Engl J Med 1977; 296:245. Rosenbloom AL. Normal insulin binding to cultured fibroblasts from patients with lipoatrophic diabetes. J Clin Endocrinol Metab 1977; 44:803. Wachslicht-Rodbard H, Muggeo M, Kahn CR, et al. Heterogeneity of the insulin-receptor interaction in lipoatrophic diabetes. J Clin Endocrinol Metab 1981; 52:416. Dib K, Whitehead JP, Humphreys PJ, et al. Impaired activation of phospho-inositide 3 kinase by insulin in fibroblasts from patients with severe insulin resistance and pseudoacromegaly. J Clin Invest 1998; 101:1111. Tritos NA, Mantzoros CS. Clinical review 97: syndromes of severe insulin resistance. J Clin Endocrinol Metab 1998; 83:3025.

82a.Marchini C, Lonigro R, Verriello L, et al. Correlations between individual clinical manifestations and CTG repeat amplification in myotonic dystrophy. Clin Genet 2000; 57:74. 83. Uotani S, Yamaguchi Y, Yokota A, et al. Molecular analysis of insulin receptor gene in Werner's syndrome. Diabetes Res Clin Pract 1994; 26:175. 83a.Abe T, Yamaguchi Y, Izumino K, et al. Evaluation of insulin response in glucose tolerance test in a patient with Werner's syndrome: a 16-year follow-up study. Diabetes Nutr Metab 2000; 13:113. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100.

Freidenberg GR, White N, Cataland S, et al. Diabetes response to intravenous but not subcutaneous effectiveness of aprotinin. N Engl J Med 1981; 305:363. Schade DS, Duckworth WC. In search of the subcutaneous-insulin-resistance syndrome. N Engl J Med 1986; 315:147. Mantzoros CS, Moses AC. Treatment of severe insulin resistance. In: Azziz R, Nestler JE, Dewailly D, eds. Androgen excess disorders in women. Philadelphia: Lippincott–Raven, 1997:247. Dunaif A, Scott D, Finegood D, et al. The insulin-sensitizing agent troglita-zone improves metabolic and reproductive abnormalities in the polycystic ovary syndrome. J Clin Endocrinol Metab 1996; 81(9):3299. Ehrmann DA. Insulin-lowering therapeutic modalities for polycystic ovary syndrome. Endocrinol Metab Clin North Am 1999; 28(2):423. Vestergaard H, Rossen M, Urhammer SA, et al. Short and long-term metabolic effects of recombinant human IGF-I treatment in patients with severe insulin resistance and diabetes mellitus. Eur J Endocrinol 1997; 136:475. Nakae J, Kato M, Murashita M, et al. Long-term effect of recombinant human IGF-I on metabolic and growth control in a patient with leprechaunism. J Clin Endocrinol Metab 1998; 83:542. Mantzoros CS, Tzonou A, Signorello LB, et al. Insulin-like growth factor 1 in relation to prostate cancer and benign prostatic hyperplasia. Br J Cancer 1997; 76:1115. Shaneyfelt T, Husein R, Bubley G, Mantzoros C. Hormonal predictors of prostate cancer. A meta-analysis. J Clin Oncol 2000; 18:847. Wolk A, Mantzoros CS, Andersson SO, et al. Insulin-like growth factor 1 and prostate cancer risk: a population-based, case-control study. J Natl Cancer Inst 1998; 17;90(12):911. Bohlke K, Cramer DW, Trichopoulos D, Mantzoros CS. Insulin-like growth factor-I in relation to premenopausal ductal carcinoma in situ of the breast. Epidemiology 1998; 9(5):570. Manousos O, Souglakos J, Bosetti C, et al. IGF-I and IGF-II in relation to colorectal cancer. Int J Cancer 1999; 24;83(1):15. Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science 1998; 23;279(5350):563. Izumino K, Sakamaki H, Ishibashi M, et al. Troglitazone ameliorates insulin resistance in patients with Werner's syndrome. J Clin Endocrinol Metab 1997; 82:2391. Goldfine AB, Simonson DC, Folli F, et al. Metabolic effects of sodium metavanadate in humans with insulin-dependent and noninsulin-dependent diabetes mellitus: in vivo and in vitro studies. J Clin Endocrinol Metab 1995; 80:3311. Panz VR, Wing JR, Raal FJ, et al. Improved glucose tolerance after effective lipid-lowering therapy with bezafibrate in a patient with lipoatrophic diabetes mellitus: a putative role for Randle's cycle in its pathogenesis? Clin Endocrinol (Oxf) 1997; 46:365. Mantzoros CS, Lawrence WD, Levy J. Insulin resistance in a patient with ovarian stromal hyperthecosis and the hyperandrogenism, insulin resistance and acanthosis nigricans syndrome. Report of a case with a possible endogenous ovarian factor. J Reprod Med 1995; 40(6):491.

CHAPTER 147 CARDIOVASCULAR COMPLICATIONS OF DIABETES MELLITUS Principles and Practice of Endocrinology and Metabolism

CHAPTER 147 CARDIOVASCULAR COMPLICATIONS OF DIABETES MELLITUS KARIN HEHENBERGER AND GEORGE L. KING Epidemiology Pathology Atherosclerosis Cardiac Pathology Endothelial Cells Smooth Muscle Cells Hematogenous Factors Metabolic Factors that Contribute to the Development of Cardiovascular Disease Hyperglycemia Hormones Cardiologic Diseases in Diabetes Cardiomyopathy Coronary Artery Disease in Diabetes Myocardial Infarction Autonomic Neuropathy Management of Diabetic Ischemic Heart Disease Medical Management Diet Plan Exercise Drug Therapy Surgical Approaches Balloon Angioplasty Coronary Artery Bypass Surgery Diabetic Peripheral Vascular Disease Signs and Symptoms Laboratory Tests Therapy for Peripheral Vascular Disease Chapter References

Diabetes mellitus often causes many chronic complications, including cardiovascular disease, retinopathy, nephropathy, neuropathy, and chronic foot ulcers involving both micro- and macrovessels.1,2 The presence and severity of these microangiopathies and cardiovascular diseases are associated with the duration, the severity of metabolic disturbances, and modulation by genetic factors (Fig. 147-1). Due to these vascular diseases, diabetes is a major contributor to cardiovascular morbidity and mortality, to blindness, and to renal failure. The increase in cardiovascular mortality results from the acceleration of atherosclerosis of coronary and peripheral arteries, cardiomyopathy, and cardiac neuropathy. In addition, diabetes is associated with an unusual vascular sclerosis arising from calcification of the media of large arteries (termed Monckeberg sclerosis), which rarely occurs in the nondiabetic population.3 Many studies have suggested that the increased cardiovascular risks in diabetic patients are due to multiple factors, with alterations in many vascular cell functions. Among these factors are insulin resistance, hyperlipidemia, endothelial dysfunction, hypertension, hypercoagulation, and hyperglycemia.4,5,6,7 and 8 Prospective studies have demonstrated that lipid-lowering drugs and antihypertensive agents in combination with glycemic control can clearly decrease the mortality of diabetes patients due to cardiovascular diseases.8

FIGURE 147-1. Causes of death in insulin-dependent (type 1) diabetic patients. ( , cardiovascular; , renal; , other diabetes related; related.) (From Orchard TJ. From diagnosis and classification to complications and therapy. Diabetes Care 1994; 17:326.)

, nondiabetes

This chapter reviews the pathology of the various cardiovascular diseases previously stated and describes the possible mechanisms. Current treatment regimens are outlined at the end of the chapter.

EPIDEMIOLOGY An excess of coronary artery disease (CAD) in patients with insulin-dependent diabetes mellitus (type 1) is usually observed in patients older than 30 years of age.4,5 and 6 The cardiovascular risk in the diabetic population has been assessed by many studies5,6,8,9; among these, the Framingham Study (a longitudinal survey of >5000 patients with a follow-up of 18 years) demonstrated that the incidence of the major clinical manifestations of coronary disease was higher in diabetic patients.9 Type 1 diabetics followed for 20 to 40 years had a 33% incidence of death due to CAD between the ages of 30 and 55, whereas for nondiabetics in the same age group, the incidence of death due to CAD was only 8% among men and 4% among women.5 There was not only an excess of mortality among these patients, but also an excess of symptomatic and asymptomatic CAD among survivors. Among women, there was a significant twofold increase in incidence of the major clinical manifestations of coronary disease. The risks of CAD in men and women with type 1 diabetes are similar and increase at the same rate after age 30 regardless of whether the onset of diabetes was in early childhood or in late adolescence. In noninsulin-dependent diabetes (type 2), the prevalence of CAD is also increased. Furthermore, CAD and type 2 diabetes frequently cluster in families.5,6,7,8 and 9 Hyperinsulinemia and insulin resistance, which often precede type 2 diabetes, are also independent predictors of the development of elevated blood pressure and lipid abnormalities.7,10,11,12,13 and 14 In the presence of type 2 diabetes, risk factors for CAD become more prevalent and more intense. In the Framingham Study, it was found that the risk of CAD increases with duration; this reflects the effect of the aging process. In contrast, in patients with diabetes, the increased risk may reflect the combined effects of aging and duration of diabetes.9 More than 50% of mortality in diabetic patients is related to cardiovascular diseases—the risk for developing cardiac or cerebrovascular disease is two to four times higher in diabetic patients than in the general population. The effect of diabetes on CAD risk may be greater for women than for men (five-fold increase in women compared to a two-fold increase in men). Several studies have shown that type 2 diabetics treated with insulin have a higher risk of CAD than do noninsulin-treated type 2 diabetics, indicating that the severity of the disease, the loss of islet cell functions, or exogenous insulin treatment may have a greater impact on CAD.7,10,11 Diabetes also has a significant adverse impact on peripheral vascular disease and congestive heart failure. For CAD, not only is diabetes synergistic with other risk factors (age, hypercholesterolemia, hypertension, and smoking), but it also is an independent risk factor.9,12,13,14,15,16,17,18,19,20,21 and 22 The extent and complexity of coronary artery lesions in diabetic patients are greater than those in the general population. (It is not certain whether atherosclerosis is more diffuse within the coronary vasculature in diabetic patients as compared with the general population. If so, this could be clinically important, because diffuse coronary lesions limit the ability to treat coronary disease by surgical means.) Hyperinsulinemia may be an important risk factor for ischemic heart disease. Among >2000 men who did not have ischemic heart disease, those persons who

eventually experienced a coronary event had higher fasting insulin levels, even after controlling for lipid levels. High fasting insulin concentrations appeared to be an independent predictor of ischemic heart disease in men.7 Myocardial infarction is the leading cause of death in both type 1 and 2 diabetic patients.5,6,9,14,17 This results not only from the higher incidence of CAD, but also from the higher acute and long-term mortality in diabetic patients who have had a myocardial infarction.23,24,27 Several prospective studies have been performed to determine whether a more intensive insulin treatment protocol can decrease the morbidity and mortality due to cardiovascular events in diabetic patients. In the University Group Diabetes Program (UGDP), a more intensive insulin strategy was not associated with a change in cardiovascular event rates when compared to standard insulin or diet therapy.25 Intensive insulin therapy in newly diagnosed nonobese, insulin-sensitive, type 2 diabetic Japanese patients slowed the progression of retinopathy, nephropathy, and neuropathy, but had too few cardiovascular events to detect the effect of intensive glucose management.26 A Swedish study comparing intensive insulin with standard antidiabetic therapy in patients with a recent myocardial infarction23,24 showed that the very high death rates in the first 3.4 years after myocardial infarction in diabetes (44%) can be significantly lowered by early intensive insulin management (33%). Most of these patients (82%) had type 2 diabetes. This effect was observed within a few months after infarction, and may have been the result of the action of insulin on platelet function, thrombosis, or myocardial dysfunction after coronary occlusion, rather than of an effect on long-term atherosclerotic disease. In fact, the risk of myocardial infarction in diabetic patients without a history of myocardial infarction is equal to that of nondiabetic patients with a history of myocardial infarction.27 The United Kingdom Prospective Diabetes Study (UKPDS) in newly diagnosed type 2 diabetic patients examined the issue of glycemic control and complications.8 This study showed that intensive glucose control is associated with a 12% reduction in the risk of pooled macrovascular and microvascular events. However, the effect is primarily due to the slowing of progression of microvascular rather than macrovascular events.

PATHOLOGY ATHEROSCLEROSIS In general, the atherosclerotic lesions in diabetic patients are similar, but they appear earlier and in greater number as compared to the nondiabetic population. The earliest lesions are fatty streaks in focal areas of the intima, characterized microscopically by the presence of macrophages, lipid deposits, and smooth muscle cells. Later lesions include fibrofatty plaques. The latest, complicated lesions are characterized by fibrosis, calcium deposits, plaque fissuring, hemorrhage, and thrombosis.12,13,14,15,16,17,18,19,20,21,22 and 22a CARDIAC PATHOLOGY Histologic abnormalities in the hearts of diabetic patients can occur at all levels of the myocardium, extending from the basement membrane to major intramural arteries.13,14,28,29 and 30 Specifically, the thickness of the capillary basement membrane is increased in the myocardium, as it is in all tissues. In addition, small capillaries and venules exhibit aneurysmal formation and, occasionally, intense vasospasm, which can lead to a state of transient myocardial ischemia. The arterial vasculature also exhibits changes, consisting of medial hypertrophy, endothelial thickening, and a thickened extracellular matrix of the intramural arterioles. Autonomic nerve pathologies can be found in diabetic patients with reduced myelinated fiber density, degeneration and regeneration of unmyelinated fibers, and capillary basement membrane thickening.30 In the myocardium, fibrosis, hypertrophy, and increased extracellular matrix are commonly observed.28 Changes in cardiac function and histology also occur in animal models of diabetes. There is an impairment in the development of left ventricular pressure and a delay in the rate of muscle relaxation.31 However, some differences have been reported between type 1 and type 2 diabetic animal models. First, the heart of the type 1 animal has a greater impairment in generating tension against an elevated preload. Second, the hearts of type 2 animals exhibit a greater decrease in cardiac compliance than do those of type 1 diabetic animals; these findings have been observed in diabetic patients as well.31,32,33,34 and 35 Increases in fatty acids and triglycerides have been noted in the cardiac tissues of diabetic animals, possibly as a result of elevated plasma levels. Increases in fatty acid metabolism elevate the accumulation of citric acid cycle intermediates, which are potent inhibitors of phosphofructokinase, a rate-limiting enzyme of glycolysis. Glycolysis and glucose oxidation are hindered further by decreases in glucose uptake and in activities of pyruvate dehydrogenase, which are regulated by insulin.36 These changes in carbohydrate metabolism may be partly responsible for the dysfunction in the heart under stress. Besides abnormal glucose metabolism, abnormalities of calcium transport between the sarcolemma and the sarcoplasmic reticulum have also been reported.37,38 Since Ca2+ fluxes are important for cardiac contractility, the decrease in the Ca2+ pool may be responsible for some of the reduction of the contractility observed in the cardiac muscles of diabetic animals and humans. The decreases in contractility in the heart could also be due to changes of the predominant active form of myosin adenosine triphosphatase (ATPase) in the heart to a less active type.38 The a1-adrenoceptor is decreased in diabetic rats; this change could be the result of protein kinase C (PKC) activation.38,39 Furthermore, other changes (e.g., cardiac lipo-protein lipase abnormalities, altered G-protein actions, increased NADPH/NADP [nicotinamide adenine dinucleotide phosphate] ratio, increased type IV collagen, and decreased [g-2-(Na+)-K+-ATPase] activities) have all been reported.40,41 and 42 At the molecular level, the expression of multiple genes in the heart is altered by diabetes. Glucose transporter (GLUT4), which is expressed mostly in insulin-sensitive tissue (e.g., muscle, fat, and the heart), is reduced in diabetic rats, leading to the hypothesis that the decrease in glucose transport could be partially responsible for the decrease in cardiac work. The expression of cardiac myosin heavy chain shifts from predominantly a to predominantly b, with the onset of diabetes; this effect is reversed by insulin administration. The expression of Ca2+ ATPase mRNA is reduced in diabetic as well as in insulin-resistant obese mice, providing an explanation for the delay in diastolic relaxation.41,42 The expression of core 2 transferase, an enzyme involved in O-linked glycosylation, is increased in the myocardium of diabetic animals. This finding could explain the increased enzymatic glycosylation observed in the hearts of diabetic patients.43 Thus, cardiac dysfunction in diabetic patients can be due to atherosclerotic changes in the coronary macrovessels or due to metabolic alterations of the myocardium resulting from exposure to the abnormal metabolic milieu of diabetic patients. Although glycemic control may be able to reverse the latter changes, the data are not clear concerning any beneficial effects in preventing or reversing atherosclerotic pathologies. ENDOTHELIAL CELLS Endothelial cells have an active role in maintaining vascular function involving anticoagulation, contractility, leukocyte trafficking, and vascular permeability. Endothelial cells can produce large numbers of vasoactive agents (e.g., platelet-derived growth factor [PDGF], endothelin, prostaglandins, and various other cytokines; see Chap. 173) that have significant effects on the metabolism and proliferation of vascular cells. Moreover, large polypeptides (e.g., insulin-like growth factors [IGFs] and low-density lipoproteins [LDL]) have receptors on the endothelial cells, and may be transported across the cells in a receptor-mediated transcytosis pathway.44,45 Injury to the endothelium can result in alteration of its many functions, thereby decreasing its nonthrombogenic surfaces and releasing factors that affect smooth muscle cells.18,19 and 20,46,47 In the diabetic environment, many factors, such as hyperglycemia, insulin resistance, increased plasma LDL, decreased high-density lipoproteins (HDL), abnormal rheologic factors, platelet aggregation, and coagulation, can contribute to endothelial cell injury.13,14,15,16,17,18,19,20,21,22,47,48 and 49 Much interest has been focused on insulin's effects on endothelial cells, especially the blunting of insulin's vasodilatory effects in insulin-resistant states.50 Infusion of insulin can cause vasodilation in several local vascular beds (e.g., leg, arm, and retina). However, insulin does not appear to induce any acute systemic vascular changes (i.e., consistent changes in blood pressure or pulse rates have not been observed). The vasodilatory effect of insulin is mediated by increases in nitric oxide (NO) production, either by activating or by enhancing the expression of NO synthase.50,51 The physiologic importance of insulin's acute vasodilatory effect is unclear; it is minor in comparison to other vasodilators, such as acetylcholine. Chronically, insulin may modulate the level of endothelial nitric oxide (eNO) and regulate endothelial functions. Hyperglycemia alters endothelial functions (e.g., increased production of oxidants, increased production of adhesion molecules, and decreased production of NO). As a result of these factors, many functions mediated by the endothelium are abnormal in the diabetic state, leading to conditions favoring atherosclerosis.52,52a SMOOTH MUSCLE CELLS Smooth muscle cells, which play a major role in the development of atherosclerosis, are found in fibrous plaques and fatty streaks, together with macrophages and monocytes. In general, smooth muscle cells can incorporate a variety of lipid particles, migrate into the intima, and proliferate. Among the many growth factors that have been implicated in this process are PDGF, IGFs, angiotensin, growth hormone (GH), and endothelin. The most-studied of these factors in diabetic conditions is insulin, because multiple epidemiologic studies have suggested a connection between hyperinsulinemia, insulin resistance, and atherosclerosis.7,10,51,52,53 and 54 Since insulin can enhance the migration and proliferation of smooth muscle cells, it has been postulated that the hyperin-sulinemia observed in insulin resistance states may enhance the atherosclerotic process. However, the mitogenic effects of insulin on smooth muscle cells are relatively weak and require insulin concentrations 10 to 50 times above the physiologic levels. Nonetheless, insulin resistance appears to be consistently associated with an increased risk of atherosclerosis. Thus, it is likely that insulin has antiatherosclerotic effects, such as the activation of eNO activity. In insulin-resistant states, insulin's effects are decreased, thereby accelerating atherosclerosis.51,52,53 and 54 Studies on insulin's direct effects on the vasculature of insulin-resistant animals have shown that there is a selective inhibition of insulin's

acute actions (e.g., on NO production), but its chronic effects are not impaired.55 HEMATOGENOUS FACTORS In the blood, platelets and macrophages play important roles in the development of atherosclerosis. They secrete vasoactive factors, including PDGF, fibroblast growth factor (FGF), tumor necrosis factor (TNF), and other cytokines.56,57,58,59 and 60 Abnormalities in the platelet function of diabetic patients13 include increased sensitivity to aggregation and augmented synthesis of pro-thrombotic prostaglandins such as thromboxane A2.60 These abnormalities enhance the release of growth factors and other cytokines both in platelets and vascular cells. The plasma of diabetics also contains another group of atherogenic components, namely abnormal lipids and lipopro-teins (see Chap. 166). In more than half of diabetic patients, especially type 2 diabetic patients, a decrease in HDL cholesterol and hypertriglyceridemia occurs.12,61,62 LDLs are the main carrier of cholesterol to tissues. Elevation of LDL levels is associated with an early onset of atherosclerosis. In type 2 diabetics and in poorly controlled type 1 diabetics, LDL cholesterol levels are reported to be elevated. Furthermore, LDLs are modified in diabetes; hyperglycemia can increase levels of glycated or oxidized LDL,63,64 which are internalized less and, hence, less degraded. The LDLs of diabetic patients may also be enriched in triglycerides, thus increasing the modified or small dense LDLs that are known to be more atherogenic.65 These altered LDLs can bind to circulating macrophages, which can then interact with the vascular cell wall, releasing vasoactive factors and becoming foam cells. The most common lipid abnormality in diabetic patients is hypertriglyceridemia.65,66,67 and 68 This is likely due to a decrease in the lipoprotein lipase (LPL) activity, which is necessary for the breakdown of chylomicrons and triglycerides.66 Thus, the plasma levels of chylomicrons and very low-density lipoproteins (VLDL) are elevated.65,66,67 and 68 The elevation of triglycerides has been reported mainly in type 2 diabetic patients, since insulin is a major regulator of LPL activity.69 Increases of VLDL levels accelerate the atherosclerotic process in a number of ways: (a) VLDL may be toxic for the metabolism and growth of endothelial cells,70 and (b) VLDL from diabetic animals may deposit more lipids in macrophages, which can be precursors of foam cells in the arterial walls.13,14,15,16,17,18,19 and 20 Plasma HDL levels may also be decreased in diabetic patients,67,68 especially those with insulin resistance. The role of HDL is to remove cholesterol from peripheral tissue; hence, a low HDL level is associated with the premature onset of atherosclerosis. HDL levels can be elevated by improved plasma glucose control, by using either insulin or by decreasing insulin resistance.55 The decrease in HDL is probably due to the diminution of insulin-sensitive LPL, which indirectly causes a decrease in the transfer of phospholipids and proteins from LDL and chylomicrons to HDL3, resulting in a decrease of HDL2A, which is important for transferring cholesterol. HDL catabolism may be accelerated because of the increased nonenzymatic glycosylation of HDL.64 Some reports indicate that this decrease in HDL levels is much greater in women than in men, perhaps explaining the increased risk of cardiovascular events in women with diabetes.

METABOLIC FACTORS THAT CONTRIBUTE TO THE DEVELOPMENT OF CARDIOVASCULAR DISEASE HYPERGLYCEMIA Hyperglycemia has been postulated to have a major role in the development of complications in diabetic patients.8,71 In both type 1 and type 2 diabetic patients, strict metabolic control can reduce cardiovascular and microvascular complications.8 Several theories have been proposed to explain the adverse effects of hyperglycemia on vascular cells (Table 147-1). One theory is based on the finding that intracellular levels of sorbitol are increased, due to an augmented conversion via aldose reductase within the cells.72,73 The high concentrations of sorbitol may alter cellular functions by osmotic changes or by altering levels of myoinositol and other enzymes, such as Na+-K+-ATPase (see Chap. 148). The metabolism of sorbitol may also alter the NAD/NADH ratio, thus potentially altering vascular cellular metabolism.

TABLE 147-1. Possible Mechanisms of Hyperglycemia's Adverse Effects

A second theory suggests that in patients with poorly controlled diabetes, hyperglycemia may interact with primary amines of proteins to form nonenzymatic glycation products with cellular and matrix proteins.74,75 Glycated proteins (e.g., albumin or LDL) may interact differently with vascular cells to cause injury to endothelial cells and to increase the proliferation of smooth muscle cells.63,64 In addition, the degradation products of cross-linked proteins may interact with specific receptors on macrophages, which then release vasoactive substances (e.g., PDGF or TNF). Possibly, hyperglycemia may also enhance the effects of other risk factors (e.g., lipoproteins). Nonenzymatic glycation of proteins can also lead to increased formation of oxidative products that can react with lipids and proteins to cause vascular changes and damage.74,75 and 76 Another likely candidate for causing the adverse effects of hyperglycemia is an increased production of oxidants. As stated previously, hyperglycemia can increase oxidant production in at least two ways: by the nonenzymatic glycation process and by enhanced production of H2O2 via an increase in the mitochondrial flux.76,77 Evidence in support of an increase in oxidative stress can be observed in vascular tissues, as measured by elevations in lipid peroxidation and levels of antioxidative enzymes. Treatment with antioxidants, such as vitamins E and C, has been reported to prevent early changes in the cardiovascular system in diabetic animals, although the plasma levels of these vitamins are probably not significantly reduced in diabetes.78,79 Lastly, changes in signal transduction of the vascular cells induced by hyperglycemia can also cause cardiovascular complications. One of the best-documented changes in vascular tissues is the activation of PKC.80,81,82,83 and 84 The activation of PKC appears to regulate a number of vascular and hematologic functions, including vascular permeability, contractility, cellular proliferation, basement membrane synthesis, platelet aggregation, macrophage activation, and signaling mechanisms for various cytokines and hormones.85,86 Since the vascular pathologies of diabetes mellitus and atherosclerosis exhibit changes in all of these properties, activation of PKC is likely to play a role. Elevated levels of glucose in diabetes mellitus can activate PKC by increasing the formation of diacylglycerol (DAG), one of two physiologic regulators of PKC (calcium is the other). In animals with chemically or genetically induced diabetes, intra-cellular DAG levels and PKC activity are elevated in many vascular tissues (i.e., the retina, aorta, heart, and renal glomeruli80,81,82,83 and 84), as well as in noncardiovascular tissues (i.e., the liver87; Fig. 147-2). Elevated PKC has been reported in monocytes, platelets, and (possibly) the myocardium of diabetic patients. Oxidants and glycation products have also been reported to activate PKC in vascular cells, suggesting that alteration in signal transduction in the vascular cells could be the common downstream mechanism by which hyperglycemia causes many of its adverse effects.88,89

FIGURE 147-2. Schematic description of insulin's potential vascular actions in physiologic and insulin-deficient or resistant state. (MAPK, mitogen-activated protein kinase; NO, nitric oxide.)

Functional abnormalities in diabetes, such as changes in blood flow and contractility, can be observed in many organs (e.g., retina, kidney, skin, large vessels, and microvessels of peripheral nerves) of diabetic patients and animals. In animal models of diabetes, decreases in neuronal and retinal blood flow can be normalized by intravitreous injection of an isoform-specific PKC inhibitor (LY333531).81,82,83,84,85,86,87,88,89 and 90 In the kidney, an elevated glomerular filtration rate is a common finding in patients with short-term diabetes mellitus91 and in experimental animal models of diabetes.92 This may be the result of hyperglycemia-induced decreases in afferent arteriolar resistance.91,92 Transgenic mice overexpressing the PKCb isoform specifically in the myocardium develop cardiac hypertrophy and fibrosis, which are likewise observed in diabetic cardiomyopathy.93 HORMONES The levels of many hormones are altered in the plasma of diabetic patients; some of these substances have been reported to be vasoactive or trophic. Those that have been associated with an increased risk of cardiovascular disease are described briefly. Elevated plasma insulin levels (either in fasting states as may occur in type 2 diabetes or in patients treated with exogenous insulin) may enhance the proliferation of arterial smooth muscle cells, especially in the presence of other growth factors.56,57,58 and 59 Hyperinsulinemia and insulin resistance have been associated with the development of hypertension and an increased risk of macrovascular diseases in diabetic or insulin-resistant patients. However, only insulin resistance appears to be consistently associated across all ethnic groups. The molecular explanation of how insulin resistance can cause acceleration of atherosclerosis has been studied intensively without any clear conclusion. To summarize briefly, insulin has multiple effects on the endothelial and smooth muscle cells, including activation of NO, protein synthesis, expression of cytokines and extracellular proteins, and growth. In general, there appears to be a selective resistance to insulin's actions in the vascular tissue. For example, insulin's vasodilation effect, as mediated by increasing NO production, is blunted in the insulin-resistant state and in type 2 diabetes.50,51 Yet, insulin's mitogenic effect on smooth muscle cells appears to be unaffected (Fig. 147-2). These findings indicate that insulin's physiologic effects (e.g., on NO production) are mainly antiatherogenic. The loss of these effects in insulin-resistant states creates conditions that can enhance atherosclerosis without the requirement for hyper-insulinemia. If hyperinsulinemia is present, then the migration and growth of smooth muscle cells may also be enhanced. However, it is unclear whether insulin's growth-promoting effects have any physiologic significance, since this would require plasma insulin concentrations that are 10 to 100 times above the levels normally encountered. IGF-I, IGF-II, and GH also increase the synthesis of DNA and induce cellular proliferation of smooth muscle cells.94,95 However, these growth factors probably do not have a major role in initiating atherosclerosis, because in conditions in which GH and IGF-I levels are elevated (e.g., acromegaly), the prevalence of atherosclerosis is not significantly increased. Some reports indicate that GH replacement therapy may decrease atherosclerosis in GH-deficient individuals. Levels of counter-regulatory hormones (e.g., cortisol and catecholamines) are also elevated. In this regard, there is some suggestive evidence that CAD is increased in Cushing syndrome and in patients treated with prednisone.96 Increased angiotensin actions have also been suggested, since treatment with angiotensin-converting enzyme inhibitors can decrease the risk of nephropathy and possible cardiovascular events in diabetic patients. However, angiotensin levels are not demonstrably elevated in the plasma or vascular tissues of diabetic patients.97

CARDIOLOGIC DISEASES IN DIABETES Cardiomyopathy without significant CAD has also been described in diabetics, including up to 20% of diabetics who have congestive heart failure. In addition, autonomic neuropathy can cause cardiac arrhythmia. CARDIOMYOPATHY Although the causes of diabetic cardiomyopathy may be multiple, it is clear that the acute metabolic derangements that occur in diabetes can produce alterations in cardiac myofibrillar performance. Numerous studies examining the myocardium of experimental animals and of diabetic patients have demonstrated that altered glucose and free fatty acid metabolism results in a derangement of myocardial performance.36,41 Tissue culture studies have shown a decrease in the velocity of contraction of myocardial cells when exposed to a high glucose environment. In addition, in the presence of high glucose concentrations, alterations are often found in the relaxation pattern of myocardial cells. Normalization of the glucose concentrations results in restoration of normal contractile relaxation indices.98 Diabetic cardiomyopathy has been reported for both type 1 and type 2 diabetics.31,32 and 33 With the use of gated-blood pool studies after vigorous exercise, both type 1 and type 2 diabetic patients have demonstrated an abnormal ejection fraction response. In catheterization studies, consistent abnormalities of cardiac function have been demonstrated in diabetic patients, even those with no or minimal evidence of CAD. These abnormalities have included elevated left ventricular and diastolic pressures, decreased ejection fractions, and increased ventricular wall stiffness. Hemodynamic functional abnormalities, including lower cardiac output and lower left ventricular compliance, have also been found in newly diagnosed diabetic patients under the stress of exercise. Studies of systolic time intervals and M-mode echocardiography in asymptomatic diabetic patients have revealed an increase of the pre-ejection period (PEP)/left ventricular ejection time (LVET), which is an index of decreased left ventricular contractility and compliance. Other evidence suggests that myocardial dysfunction is related to metabolic control. In one study, 3 months of intensive insulin therapy reversed the previously noted abnormalities.33,99 Despite a preponderance of theoretical evidence, it has been difficult to make a specific diagnosis of diabetic cardiomyopathy, because of other potential etiologies that can lead to poor cardiac function in diabetic patients. In addition, the prevalence of CAD in diabetics could contribute significantly to the systolic or diastolic cardiac impairment. Another confounding factor in diabetics is a history of hypertension, which, by itself, can lead to the development of cardiac abnormalities. Other factors (e.g., renal failure with resultant anemia, hypertension, and volume overload) can impose additional stress on the myocardium, unmasking an underlying diabetic myopathic state. Thus, diabetic cardiomyopathy is frequently the silent partner of a more clinically obvious form of diabetic heart disease. Its effect may influence both the presenting clinical symptoms and the response to therapy. CORONARY ARTERY DISEASE IN DIABETES Multiple large population-based studies have demonstrated that the major clinical complication of long-standing diabetes is atherosclerotic cardiovascular disease.4,5 and 6 The Framingham Study demonstrated a marked prevalence of CAD in both men and women with diabetes.9 Diabetic men have a two-fold greater risk of dying of ischemic heart disease than do nondiabetic controls. Female diabetic patients are particularly vulnerable to the effects of atherosclerotic heart disease; their incidence of ischemic heart disease is five-fold that of the nondiabetic female population.9,14 A review of the coronary angiograms of cardiac patients revealed a higher incidence of atherosclerosis in diabetic patients; they also had had a more diffuse pattern than did non-diabetic patients.100 Similar conclusions were reached in an angiographic study of patients with juvenile-onset type 1 diabetes. In addition, female juvenile diabetic patients were particularly prone to have a diffuse pattern of coronary artery involvement. Although classic manifestations of CAD are often observed in the diabetic patient, there may be a discrepancy between the clinical symptoms and the severity of

underlying heart disease. Not uncommonly, rather mild symptoms are associated with marked, and potentially life-threatening, coronary artery lesions. Although this phenomenon occurs in many nondia-betic patients with ischemic heart disease, it appears to be more prevalent in people with diabetes. Not uncommonly, diabetic patients with ischemic heart disease may not experience chest pain; they may complain of nonspecific or ambiguous findings such as exertional dyspnea, mild diaphoresis, nausea, and vomiting, or generalized weakness. Thus, in assessing the diabetic patient with CAD, it is important to have a high index of suspicion of underlying atherosclerotic cardiovascular disease and not to be misled by what appear to be mild or stable clinical symptoms.100 Because of the subtlety of presentation and the difficulty in relying on symptomatology to assess the clinical severity of coronary atherosclerosis, an objective evaluation using some modality of exercise testing to assess both the presence and the severity of CAD should be performed. Indications for coronary angiography are the same for diabetic and nondiabetic patients. The outcome of coronary bypass graft surgery in diabetic patients is excellent in terms of survival and symptomatic relief.100 The hazards of angiography per se in diabetic patients are increased only in those patients with abnormal renal function, due to an increased risk of compromising renal function when using radiographic dyes. This risk can be normalized by reducing the dose of the dye, by hydrating the patients, and by the possible use of mannitol to augment urine output and to avoid hypotension. On the day of the procedure, good glucose control should be maintained—either by continuous insulin pump or by splitting the morning dose of insulin (half before the procedure and the remaining half afterward). The blood glucose value should be checked frequently during the patient procedure. A cautionary note should be raised in the care of insulin-taking diabetic patients who are scheduled to undergo cardiac surgery after angiography. There is an increased danger of anaphylaxis after the injection of protamine, which is used to neutralize the actions of administered heparin. The increased rate of reaction is probably due to previous exposure to protamine in the neutral protamine Hagedorn (NPH) insulin used by diabetic patients.84 MYOCARDIAL INFARCTION In a previous review of morbidity and mortality in diabetics after myocardial infarction,5,100 an overall mortality of 30% was found; there was a high frequency of congestive heart failure. In addition, 30% of the diabetic patients presenting to the coronary care unit had no pain, compared with 20% mortality within 2 years.5,23,28,100 Analysis of multiple cardiovascular parameters did not elucidate the cause of the increased mortality in the diabetic group. Specifically, the extent of myocardial infarction as assessed by global left ventricular function (quantitated by radionuclide ventriculogram) failed to explain the marked discrepancy found between diabetic and nondiabetic patients. In addition, the results of baseline Holter monitoring and low-grade exercise tolerance tests before discharge from the hospital were not statistically different between the two groups. The mode of death in the diabetic group usually consisted of progressive congestive heart failure or sudden death. Although the underlying cause for the increased mortality is not definitely known, the unique characteristics of diabetic heart disease may be important factors. For example, there is an increased propensity for lethal ventricular arrhythmias, presumably caused by more extensive fibrosis and, hence, less responsiveness to antiarrhythmic agents. In addition, the cardiomyopathic changes unique to diabetes may result in more dramatic abnormalities because of the altered diastolic compliance of the left ventricle. Although indices of systolic performance may be comparable in diabetic and nondiabetic groups, increased stiffness of the diabetic ventricle may aggravate hemodynamic changes and thereby lead to more severe and progressive congestive heart failure.25 Autonomic neuropathy, particularly the loss of parasympathetic innervation, may predispose to a more vulnerable ventricular myocardium in terms of arrhythmia potential.101 There is a greater propensity for coronary vasoconstriction to occur in diabetic patients with parasympathetic denervation. Although speculative, some or all of these unique characteristics of diabetic heart disease may be responsible for the increased initial and first-year mortality found in diabetics with acute myocardial infarction. In the Diabetes Insulin-Glucose in Acute Myocardial Infarction (DIGAMI) study, it was found that intensive metabolic treatment of diabetic patients with acute myocardial infarction with insulin-glucose infusion followed by multidose insulin treatment improved the prognosis.23,24 The effect was most apparent in patients who had not previously received insulin treatment and who were at low cardiovascular risk. The overall mortality after 1 year was 19% in the insulin group compared to 26% among controls. The most frequent cause of death in all patients was congestive heart failure, but cardiovascular mortality tended to be decreased in insulin-treated patients.23,24 AUTONOMIC NEUROPATHY Abnormalities in cardiac innervation have been associated with two types of cardiac dysfunctions, involving both sensory and autonomic nerve fibers. Injury to sensory fibers could result in the phenomenon of silent, painless infarctions. Various abnormalities in heart rate occur in the diabetic population. These include persistent tachycardia, absence of a rate variation with Valsalva maneuver, and a blunting of the normal variation of the heart rate that occurs during deep breathing.101,102 and 103 These abnormalities are mainly due to dysfunctions of the vagus nerve, although sympathetic activity in the heart may also be altered in patients with severe autonomic neuropathy. Loss of sympathetic modulation may be important during exercise, when maximizing heart rate may be functionally significant.101,102 and 103 An evaluation of the parasympathetic regulation of the heart rate can be clinically useful. Several methods have been used to evaluate cardiac parasympathetic function, including heart rate or RR interval variation during deep breathing, heart rate response to Valsalva maneuver, and heart rate response to standing. The sensitivity for detecting a clinically relevant abnormality is increased if two of these tests are used together. The principle behind these assays is to measure the vagal regulation of the heart rate, which is reflected in the RR intervals.103a The variation in the heart rate change depends on the blood flow back to the heart. The tachycardia normally observed during the Valsalva maneuver is induced by the lack of vagal tone, whereas the bradycardia that occurs after the maneuver is due mainly to an increase of vagal tone. Thus, a deficiency of vagal activity will lead to a decrease in the variation of heart rate during these maneuvers. In one study, the variation in heart rate during one deep breath was reduced significantly in 62 of 64 diabetic patients with other autonomic symptoms, and in 30% of diabetic patients who had peripheral neuropathy but no autonomic symptomatology.101,102 and 103 Prolonged follow-up of up to 5 years did not show any improvement and, in some patients, demonstrated deterioration. Indeed, the rate of mortality differed remarkably between diabetic patients with and without abnormal cardiovascular reflex tests. In a study of 73 diabetic patients, the mortality in those with abnormal cardioreflex testing was three- to four-fold higher; in 20%, the death was sudden, suggesting cardiac arrhythmia as a possible cause.102 Similar findings of a decrease in respiratory variations of the electrocardiographic RR interval in diabetic patients have also been reported; however, these differences are not as marked as the heart rate variation. The high rate of mortality is mostly due to other serious illnesses. Nevertheless, diabetic patients with severe cardiac autonomic dysfunction should be carefully followed, cardiac dysfunction should be treated appropriately, and blood glucose should be strictly controlled.103b

MANAGEMENT OF DIABETIC ISCHEMIC HEART DISEASE MEDICAL MANAGEMENT DIET PLAN The diet plan is one of the most important aspects of therapy for both type 1 and type 2 diabetic patients. First, with proper diet and weight loss the peripheral sensitivity, insulin secretion, or both, can be improved in type 2 diabetes. This improvement in blood-glucose control can reduce some of the risk factors by lowering LDL and increasing HDL levels. Second, it is important for patients on insulin to consume similar amounts of calories during the same time period every day, since this will require fewer changes in insulin dose. The initial goal in the diet plan is to construct a diet that will enable the diabetic patient, if obese, to lose weight. Without unusual physical activity, patients who are taking in 35 kcal/kg will, in general, maintain their weight. Because a deficit of 3500 kcal is needed to lose 1 lb, in patients with normal physical activity, a diet program allowing 1200 kcal per day will result in a loss of 1 to 2 lb per week. Although this may not appear to be particularly significant, many obese diabetic patients have a clear improvement in insulin sensitivity after losing only a few pounds. An increase in fiber content of the diet also may be helpful for this, owing to its effect in delaying enteric absorption (see Chap. 124 and Chap. 141). The second goal of the diet plan is to attempt to decrease risk factors such as hyperlipidemia, hypertension, and nephropathy—all of which will accelerate the rate of atherosclerosis. Thus, a diet low in saturated fat should be strictly followed. Other factors that can accentuate the lipid abnormalities, such as alcoholic beverages, should be avoided, although moderate intake of red wine has been suggested to decrease cardiovascular risk by increasing HDL in both diabetic and nondiabetic populations. In addition, familial hyperlipidemia, which will greatly accentuate the lipid abnormalities (especially the elevation of plasma triglyceride levels), may be found.20,104,105,106 and 107 Drug treatment with triglyceride- and cholesterol-lowering agents may be required (see Chap. 164). However, glycemic control should be tried

first since this can significantly lower both plasma VLDL and LDL levels and decrease the adverse effects of hyperglycemia. The frequency of hypertension is also increased in the diabetic population even before the development of clinical renal disease. Hypertension is a strong risk factor for the development and progression of diabetic, cardiac, renal, retinal, and peripheral vascular complications108; therefore, it should be treated vigorously. A low-salt diet in combination with drug therapy is often required. Also, animal studies suggest that a low-protein diet along with plasma glucose control in diabetic patients may improve or stabilize deteriorating renal function, although the effect in humans may not be as significant. It should be emphasized that cigarette smoking in diabetic patients must be strongly discouraged because the increase of risk for cardiovascular morbidity and mortality is enhanced in a multiple fashion (see Chap. 234). EXERCISE Exercise in diabetic patients can be helpful in several ways. First, moderate exercise has been demonstrated to improve insulin sensitivity and glucose tolerance, even in the absence of weight loss. Second, in combination with a proper diet program, exercise can be helpful in promoting weight loss and thereby lead to reduced cardiovascular risks. Exercise may also reduce the risk of vascular complications, not only by improving plasma glucose control but by increasing plasma HDL, which can lessen the risk of cardiovascular disease. In addition, risk factors for macrovascular disease in diabetic patients, such as the levels of LDL, VLDL, and fasting insulin levels, may also be lowered. Although the advantages of an exercise program can be substantial for diabetic patients, it should not be initiated without proper planning since there are potential risks. The occurrence of hypoglycemia may be increased in patients taking oral hypoglycemia agents because of an augmented sensitivity to insulin or an increased absorption of exogenously administered insulin. Exercise, which may increase retinal intravascular pressure, such as that associated with straining (e.g., in weight lifting), should also be limited. Patients who experience severe sensory neuropathy have an increased risk of injuring their lower extremities. For type 2 diabetic patients, who are generally older than type 1 diabetics, care should be taken to avoid precipitating cardiac ischemia, which may lead to arrhythmia and myocardial infarction. Also, the high prevalence of peripheral vascular involvement in diabetic patients may result in easily bruised skin, which could lead to abscess formation and osteomyelitis. Nevertheless, a supervised and planned exercise program can usually avoid these potential hazards109 (see Chap. 132 and Chap. 141). Exercise Regimen. In the initial medical evaluation, a detailed examination of the cardiac and vascular systems, in addition to ophthalmoscopic studies, is necessary to rule out cardiac lesions or proliferative retinopathy, which may be exacerbated by strenuous exercise programs. Laboratory evaluations are necessary to evaluate the control of plasma glucose levels and to determine the patient's working capacity. Evaluation of cardiac and working capacities should include resting and exercise electrocardiograms in the patient older than 40 or with a history of cardiac symptoms. Endurance-type exercise, such as walking, cycling, jogging, and aerobic workouts, is generally recommended over strength building (e.g., weight lifting) because of the increased potential of the latter to raise blood pressure transiently. The exercise plan usually begins with a short warm-up period with stretching routines, followed by 10 to 30 minutes of endurance-building activities. These exercises should stimulate the heart rate to 50% to 75% of the maximal rate, depending on the persistence in the exercise program. These exercise plans need to be performed at least three times a week to achieve beneficial effects, such as lower cardiovascular risks and decreased insulin resistance. DRUG THERAPY Specific pharmacologic therapy for cardiac dysfunctions among the diabetic and nondiabetic populations do not differ greatly. However, several general points should be stressed. As discussed previously, in patients with type 1 and 2 diabetes, normalized glucose excursions can prevent or delay the development of cardiovascular and myocardiovascular complications.8,26 Questions have been raised concerning the safety of the sulfonylureas in diabetic patients with a history of cardiac disease, as a result of the findings of the UGDP, which concluded that there was an increase of cardiovascular deaths in the tolbutamide-treated group compared with the insulin-treated group.25 These differences became significant after 3.5 years. However, the UKPDS has laid these fears to rest since intensive treatment with sulfonylurea, insulin, or metformin was equally effective in reducing fasting plasma glucose concentrations in type 2 diabetic patients.8 They also found that intensive glucose control with metformin appears to reduce the risk of diabetes-related cardiovascular end points in overweight diabetic patients as compared to insulin or sulfonylurea groups. Moreover, treatment with metformin is associated with less weight gain and fewer hypoglycemic attacks than are either insulin or sulfonylureas, suggesting that it may be the first-line pharmacologic therapy of choice in these patients. Even after starting oral hypoglycemic agents, diet and weight-control programs should be continued. These drugs should be discontinued if further weight reduction alone can result in satisfactory control of plasma glucose. The insulin-sensitizing agents, thiazolidinediones, exert direct effects on the PPARg receptors, resulting in improved insulin action and reduced hyperinsulinemia. These glitazone compounds can improve insulin action and reduce glycemia and insulin requirements in type 2 diabetics.110 Other symptoms such as dyslipidemia and hypertension could also improve following a decrease in insulin resistance. In addition to the novel mechanism of action through binding and activation of PPARgs, the glitazones are potent antioxidants since it contains vitamin E within its structure. However, the side effect of liver damage can be severe (even if uncommon) and has led to the development of alternative drugs. Additionally, these compounds are less suitable for obese patients, since there is a weight gain of up to 5 kg during treatment. Thiazolidinedione compounds provide an important additional resource for the health care provider in the management of type 2 diabetes and other aspects of the insulin resistance syndrome.110 The drug therapy of angina pectoris in diabetic patients differs slightly from that of the nondiabetic population. This primarily pertains to the use of b-adrenergic blocking agents. Although b-blockers are effective in diabetic patients without risk factors (e.g., hypoglycemia unawareness111), long-standing diabetics with impaired autonomic function often have a poor sympathetic response to hypoglycemia for which some of the warning signs may be blunted. In addition, nonselective b-adrenergic blockers inhibit gluconeogenesis, thereby prolonging the hypoglycemia recovery phase because of the requirement for mobilization of glucose from the liver. Clinically, these problems are best dealt with by warning diabetic patients that their symptoms of hypoglycemia may be more subtle. When the symptoms of hypoglycemia are finally perceived, the patients will have a shorter time to respond before unconsciousness may occur. The complete absence of symptoms when hypoglycemia occurs usually constitutes a contraindication to the use of b-adrenergic blocking drugs. The advantage of a selective b-blocker is limited to a faster recovery phase from hypoglycemia; it would appear that this minor advantage may be clinically worthwhile. Another problem encountered with the use of b-adrenergic blocking drugs in diabetic patients is the lack of efficacy among those patients with significant autonomic neuropathy. A major benefit of b-adrenergic blockers is the slowing of heart rate and a subsequent decrease of myocardial oxygen requirements, thereby minimizing the possibility of myocardial ischemia. However, in autonomic neuropathy, the parasympathetic control of heart rate is often lost. Consequently, the resting heart rate is higher than in those persons with a normally functioning vagus nerve. Patients with parasympathetic denervation do not experience the same magnitude of decrease in heart rate after full b-adrenergic blocking doses. This is due to the relatively greater importance of the parasympathetic tone, which results in only minor changes in heart rate response and, therefore, only minor improvement of myocardial oxygen requirements. Combined with the propensity of b-adrenergic blockers to aggravate hypoglycemic episodes, this narrows the therapeutic/toxic ratio of these agents. In individuals with autonomic dysfunction, it is sometimes best to select a calcium-channel antagonist as a primary mode of therapy, since these drugs cause a decrease in heart rate by a direct nonadrenergic-mediated depression of sinus mode function. Frequently, calcium-channel antagonists can result in a more impressive negative chronotropic response than do b-adrenergic blockers, especially when there is substantial autonomic dysfunction. Lastly, angiotensin-converting enzyme inhibitors (ACEI) are widely used for treatment of hypertension in diabetes since multiple studies have shown that they may delay the onset of diabetic nephropathy and decrease cardiac mortality in diabetic patients.112,113 and 114 It is likely that angiotensin receptor antagonists will have effects similar to ACEI. In nondiabetic patients (and most likely in diabetic patients), ACEI have cardiac protective effects in addition to antihypertensive effects. Hyperlipidemia is very common in both type 1 and 2 diabetics, and, as stated previously, diet and glycemic control can decrease both cholesterol and triglyceride levels. Hypercholesterolemia is a condition that occurs more often in diabetic patients than in the general population. The targeted level of cholesterol for diabetic patients should be lower than that of the general population (e.g., ideally to the range of 150 mg/dL). In the randomized trial of cholesterol lowering in patients with coronary heart disease, the Scandinavian Simvastatin Survival Study (4S) showed that long-term treatment with simvastatin, an HMG-reductase (hydroxymethylglutaryl-CoA reductase) inhibitor, is safe and improves survival in patients with coronary heart disease (Fig. 147-3).115 Over the 5.4-year median follow-up period, simvastatin produced mean changes in total cholesterol, LDL cholesterol, and HDL cholesterol of –25%, –35%, and +8%, respectively. There were fewer adverse effects among the diabetic patients in the simvastatin group (8% died, compared with 12% in the placebo group; Fig. 147-3). There were 111 coronary deaths in the simvastatin group and 189 in the placebo group (relative risk 0.58, 95% cumulative index [CI] 0.46–0.73), while noncardiovascular causes accounted for 46 and 49 deaths, respectively. One or more major coronary events occurred in 622 patients in the placebo group (28%) and in 431 patients in the simvastatin group (19%). The relative risk was 0.66 (95% CI 0.59– 0.75, p 150 mL, and recurrent urinary tract infections. Management of diabetic cystopathy emphasizes the facilitation of bladder emptying and the treatment of urinary incontinence. The Credé method and intraabdominal straining (Valsalva technique) increase intravesicular pressure and may allow for passive voiding in milder cases. Medications, including cholinergic stimulants (e.g., bethanechol chloride), which facilitate bladder contractions, and a-adrenergic agonists (e.g., phenoxybenzamine, prazosin, and terazosin), which relax the internal sphincter, may help facilitate bladder emptying. In advanced conditions, chronic catheterization may be necessary. If incontinence develops because of an incomplete sphincter, placement of an artificial urinary sphincter bladder, bladder neck reconstruction, or a fascial sling bladder neck suspension may be of benefit.1,3 Neuropathic sexual dysfunction affects 50% of diabetic men, producing impotence or retrograde ejaculation. Erectile impotence must be distinguished from psychogenic impotence and organic impotence of other causes by a detailed sexual history and appropriate laboratory tests. Yohimbine, an a2-adrenergic blocker, may help increase penile rigidity by increasing blood flow to the corpus of the penis. Other vasoactive medications, including phentolamine, papaverine, or prostaglandins, may be injected directly into the corpora cavernosa. Several mechanical devices are available that help produce erections by creating a vacuum to draw blood into the penis. If these treatments are ineffective, a penile prosthesis may be efficacious.1,3 Retrograde ejaculation, caused by the uncoordinated closure of the internal vesicle sphincter and relaxation of the external vesicle sphincter, is diagnosed by an absent ejaculate plus the recovery of live, mobile sperm in the postcoital urine (see Chap. 117 and Chap. 118). Anatomic, nondiabetic causes of retro-grade ejaculation must be excluded. Artificial insemination of sperm recovered from fresh urine samples may circumvent the resulting infertility.1,3 The blunted epinephrine response to hypoglycemia produces hypoglycemic unawareness that greatly complicates intensive insulin treatment; it is ascribed to autonomic neuropathy of the adrenal medulla, although it may occur in patients without apparent neuropathy. Gastrointestinal Neuropathy. Gastrointestinal neuropathy, which can involve the entire gastrointestinal tract, probably explains the high frequency of complaints involving this system in diabetics1,3 (see Chap. 149). Esophageal motility disorders, although usually asymptomatic, are common in long-standing diabetes.1,3 Diabetic gastropathy diminishes gastric acid secretion and motility, explaining both the reduced frequency of duodenal ulceration in diabetics and the presence of anorexia, nausea, vomiting, postprandial fullness, and early satiety in the absence of demonstrable intrinsic lesions. Delayed gastric emptying may complicate diabetic control because of retarded postprandial caloric absorption. Gastric emptying may be improved by erythromycin, which increases gastric motility through simulating the action of motilin; by bethanechol, which increases gastric contractibility through stimulation of muscarinic receptors; or by metoclopramide, which inhibits central and peripheral dopaminergic pathways and allow endogenous parasympathetic tone to act uninhibited.1,3 In refractory cases of gastroparesis, surgical procedures including jejunostomy, pyloroplasty, or partial gastrectomy may be beneficial. Diabetic diarrhea may reflect intestinal hypermotility from decreased sympathetic inhibition, hypomotility with consequent bacterial overgrowth, bile salt malabsorption, diabetic “sprue” (steatorrhea with mucosal histology resembling gluten sensitivity), and pancreatic insufficiency.1,3 Cardiovascular Autonomic Neuropathy. Cardiovascular autonomic neuropathy (see Chap. 147) produces exercise intolerance and orthostatic hypotension, the former from reduced augmentation of cardiac output and impaired constriction of the visceral vascular bed and the latter from an impaired sympathetically mediated compensatory increase in peripheral vascular resistance.1,3 Initial treatment of symptomatic orthostatic hypotension includes the use of waist-high compressive stockings, keeping the head of the bed elevated at night, and use of salt supplements, all of which act to increase blood volume and cardiac filling. If non-pharmacologic measures are ineffective, medications including fludrocortisone or short-acting pressor agents (e.g., ephedrine, midodrine, yohimbine, ergotamine, clonidine, and phenylephrine hydrochloride [used as an atomized nasal spray]) may be of benefit.1,3 Some studies suggest that erythropoietin may increase standing blood pressure in diabetic patients with orthostatic hypotension. Octreotide has also been used with some success. Cardiac denervation is marked by an invariant pulse of 80 to 90 beats per minute that is unresponsive to stress, exercise, or sleep; it is associated with painless myocardial infarction, coronary artery spasm, and sudden cardiac death.1,3

PATHOGENESIS AND THERAPEUTIC IMPLICATIONS OF DIABETIC NEUROPATHY A pathogenetic role for the metabolic alterations resulting from insulin deficiency is now generally accepted for diabetic neuropathy; however, the nature, extent, and importance of that role remain controversial.6 There is an array of interrelated metabolic alterations in diabetic nerves resulting from elevated ambient glucose concentrations (Fig. 148-8), several of which appear to interact, in a self-reinforcing or synergistic fashion, with a negative impact on nerve function. These changes include (a) increased activity of the polyol pathway leading to accumulation of sorbitol and fructose and imbalances in nicotinamide adenine dinucleotide phosphate (NADP)/NADPH,17,18 and 19 (b) formation of reactive oxygen species via glucose autooxidation,20,20a (c) nonenzymatic glycation of proteins resulting in production of “advanced glycation end products” (AGEs),18,19,19a and (d) inappropriate activation of protein kinase C.18,19 One unifying mode of injury among these different metabolic impairments lies in the production and decreased scavenging of reactive oxygen species, thereby promoting cellular oxidative stress and mitochondrial dysfunction that can lead to programmed cell death of nervous system tissues.12,17,18,19,20,21,22,23 and 24

FIGURE 148-8. Increased glucose is thought to initiate a cascade of putatively cytotoxic metabolic events through autooxidation, glycation, and formation of advanced glycation end products (AGEs), and through increased sorbitol pathway activity. Autooxidation and/or AGEs are thought to promote the generation of toxic free radicals, including a variety of reactive oxygen species (ROS). Sorbitol production produces compensatory depletion of other organic osmolytes such as myo-inositol and taurine, and endogenous antioxidant, with resulting attenuation of oxidative defense. Sorbitol accumulation and/or reciprocal osmolyte depletion also produce osmotic stress, which may damage mitochondria and stimulate protein kinase C (PKC). PKC may also be activated by shifts in triose phosphates toward diacylglycerol production as a result of increased sorbitol pathway activity. Both ROS and PKC activation have been implicated in microvascular dysfunction in peripheral nerve and retina, producing tissue ischemia. Ischemia impairs mitochondrial function and generates additional ROS. Mitochondrial dysfunction and ROS further impair oxidative defense mechanisms. Neurotrophic support through nerve growth factor (NGF) and perhaps other neurotrophins that regulate oxidative defense mechanisms may be impaired by ROS. Mitochondrial damage is thought to lead to the release of cytochromes, activation of caspases, and induction of apoptosis or programmed cell death (PCD).

POLYOL PATHWAY The polyol pathway refers to the enzymatic conversion of glucose to sorbitol and then to fructose. Because this reaction is dependent on the concentration of glucose, sorbitol and fructose are formed in high concentrations in many tissues of the diabetic, including nerves. Conversion of glucose to sorbitol is dependent on aldose reductase, while sorbitol is converted to fructose via sorbitol dehydrogenase. Both reactions change the oxidation/reduction state of the cell, decreasing NADPH/NADP+ and NADH/NAD+ ratios, respectively. This alters the normal redox potential of the cell and not only decreases the ability of the cell to detoxify reactive oxygen species but also promotes nerve ischemia and free radical injury of mitochondria.17,18,19 and 20,23,24 Mitochondrial dysfunction further impairs the cell's ability to function in the presence of unchecked reactive oxygen species by decreasing the available adenosine triphosphate (ATP), which is needed for the de novo synthesis of free radical scavengers.25 Left unchecked, reactive oxygen species produce (a) lipid, DNA, and protein peroxidation17,18,19 and 20; (b) further ischemia and reduced nerve blood flow21,22,23 and 24; and (c) cellular apoptosis.12,26 These collective metabolic and vascular impairments result in peripheral nervous system injury and clinical neuropathy. Animal studies have revealed that indicators of oxidative stress parallel neuropathy, and antioxidants that block formation of reactive oxygen species prevent neuropathy.6 In addition, insulin and aldose reductase inhibitors (which block the accumulation of sorbitol) prevent formation of reactive oxygen species and experimental diabetic neuropathy.17,18,19,20,21,22,23 and 24 AUTOOXIDATION OF GLUCOSE AND ADVANCED GLYCATED END PRODUCT FORMATION During diabetes, trace amounts of free transition metals (e.g., iron and copper) promote autooxidation of glucose, leading to further formation of reactive oxygen species.17,18 and 19,23 In experimental diabetes, metal-chelating agents preserve both nerve function and blood flow. Reactive oxygen species also may link autooxidation and AGE formation. High ambient glucose results in glycation of proteins and oxidation by reactive oxygen species, with the final end products known as AGEs. Reactive oxygen species enhance AGE formation, which, in turn, further accelerates reactive oxygen species (ROS) formation (a process known as autooxidative glycosylation). AGE accumulation correlates with endothelial dysfunction, impaired nerve blood flow, and ischemia.17,18 and 19,23 NERVE GROWTH FACTORS Once diabetic neuropathy becomes clinically evident, structural and functional abnormalities within the peripheral nervous system presumably are widespread and of long standing. To what degree these structural abnormalities are reversible is unclear; however, the studies described earlier have shown greater improvement after treatment in patients with early neuropathy. This suggests that interventions should be introduced as early as possible and should be geared toward slowing the progression of neuropathy rather than reversing the clinical signs and symptoms. Studies evaluating other neurotropic growth factors, which induce axonal sprouting in injured neurons, and may be a means to overcome some of the structural damage in diabetic neuropathy, are either planned or are under way.27. CHAPTER REFERENCES 1. Greene DA, Feldman EL, Stevens MJ, et al. Diabetic neuropathy. In: Porte D Jr, Sherwin R, eds. Diabetes mellitus. East Norwalk, CT: Appleton & Lange, 1997:1009. 2. DCCT Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329:977. 3. Feldman EL, Stevens JJ, Greene DA. Clinical management of diabetic neuropathy. In: Veves A, ed. Clinical management of diabetic neuropathy. Totowa: Humana Press, 1998:89. 4. Salpeter MM, Spanton S, Holley K, Podleski TR. Brain extract causes acetylcholine receptor redistribution which mimics some early events at developing neuromuscular junctions. J Cell Biol 1982; 93:417. 5. Stevens MJ, Feldman EL, Thomas T, Greene DA. Pathogenesis of diabetic neuropathy. In: Veves A, ed. Clinical management of diabetic neuropathy. Totowa: Humana Press, 1998:13. 6. Feldman EL, Russell JW, Sullivan KA, Golovoy D. New insights into the pathogenesis of diabetic neuropathy. Curr Opin Neurol 1999; 12:553. 7. Pirart J. Diabetes mellitus and its degenerative complications; a prospective study of 4,400 patients observed between 1947 and 1973. Diabetes Care 1978; 1:168. 8. Fedele D, Comi G, Coscelli C, et al. A multicenter study on the prevalence of diabetic neuropathy in Italy. Diabetes Care 1997; 20:836. 9. Young MJ, Boulton AJM, Macleod AF, et al. A multicentre study of the prevalence of diabetic peripheral neuropathy in the United Kingdom hospital clinic population. Diabetologia 1993; 36:150. 10. Dyck PJ, Kratz KM, Karnes JL, et al. The prevalence by staged severity of various types of diabetic neuropathy, retinopathy, and nephropathy in a population-based cohort: the Rochester Diabetic Neuropathy Study. Neurology 1993; 43:817. 11. Maser RE, Steenkiste AR, Dorman JS, et al. Epidemiological correlates of diabetic neuropathy. Report from Pittsburgh Epidemiology of Diabetes Complications Study. Diabetes 1989; 38:1456. 12. Russell JW, Sullivan KA, Windebank AJ, et al. Neurons undergo apoptosis in animal and cell culture models of diabetes. Neurobiol Dis 1999; 6:347. 13. Stevens MJ, Feldman EL, Greene DA. Diabetic peripheral neuropathy. In: DeFronzo RA, ed. Current therapy of diabetes mellitus. St. Louis: Mosby– Year Book, 1997:160. 14. Greene DA, Feldman EL, Stevens MJ. Neuropathy in the diabetic foot: new concepts in etiology and treatment. In: Levin M, O'Neal L, eds. The diabetic foot. St. Louis: Mosby, 1993:135. 15. Pfeifer MA, Ross DR, Schrage JP, et al. A highly successful and novel model for treatment of chronic painful diabetic peripheral neuropathy. Diabetes Care 1993; 16:1103. 16. The Capsaicin Study Group. Treatment of painful diabetic neuropathy with topical capsaicin: a multicenter, double-blind, vehicle controlled study. Arch Intern Med 1991; 151:2225. 17. Feldman EL, Stevens MJ, Greene DA. Pathogenesis of diabetic neuropathy. Clin Neurosci 1997; 4:365. 18. Greene DA, Stevens MJ, Obrosova I, Feldman EL. Glucose-induced oxidative stress and programmed cell death in diabetic neuropathy. Eur J Pharmacol 1999; 375:217. 19. Greene DA, Obrosova IG, Stevens MJ, Feldman EL. Pathways of glucose-mediated oxidative stress in diabetic neuropathy. In: Packer L, Tritschler HJ, King GL, Azzi A, eds. Antioxidants in diabetes management. New York: Marcel Dekker, Inc., 2000: 111. 19a. Rahbar S, Hadler JL. A new rapid method to detect inhibition of Amador; product generated by delta-gluonolactone. Clin Chim Acta 1999; 287:123. 20. van Dam PS, Bravenboer B. Oxidative stress and antioxidant treatment in diabetic neuropathy. Neurosci Res Commun 1997; 21:41. 20a. Stevens MJ, Obrosova I, Cao X, et al. Effects of DL–alpha–lipoic acid on peripheral nerve conduction, blood flow, energy metabolism, and oxidative stress in experimental diabetic neuropathy. Diabetes 2000; 49:1006. 21. 22. 23. 24. 25. 26. 27.

Low PA, Nickander KK, Tritschler HJ. The roles of oxidative stress and antioxidant treatment in experimental diabetic neuropathy. Diabetes 1997; 46(Suppl 2):S38. Tomlinson DR. Future prevention and treatment of diabetic neuropathy. Diabetes Metab 1998; 24(Suppl 3):79. Zochodne DW. Diabetic neuropathies: features and mechanisms. Brain Pathol 1999; 9:369. Cameron NE, Cotter MA. Metabolic and vascular factors in the pathogenesis of diabetic neuropathy. Diabetes 1997; 46(Suppl 2):S31. Heiden MG, Chandel NS, Schumacker PT, Thompson CB. Bcl-xL prevents cell death following growth factor withdrawal by facilitating mitochondrial ATP/ADP exchange. Mol Cell 1999; 3:159. Anderson KM, Seed T, Ou D, Harris JE. Free radicals and reactive oxygen species in programmed cell death [In Process Citation]. Med Hypotheses 1999; 52:451. Feldman EL, Windebank AJ. Growth factors and peripheral neuropathy. In: Dyck PJ, Thomas PK, eds. Diabetic neuropathy. Philadelphia: WB Saunders, 1998:377.

CHAPTER 149 GASTROINTESTINAL COMPLICATIONS OF DIABETES Principles and Practice of Endocrinology and Metabolism

CHAPTER 149 GASTROINTESTINAL COMPLICATIONS OF DIABETES FREDERIC D. GORDON AND KENNETH R. FALCHUK Pathogenesis Esophagus Stomach Small Intestine Celiac Disease Large Intestine Gallbladder Disease Liver Disease Chapter References

Diabetes mellitus can affect many organs in the gastrointestinal tract. The metabolic derangements caused by diabetes can have a significant impact on the function of the digestive and hepatobiliary systems. Rundles, in 1945, was the first to draw attention to the effects of diabetes on the gut.1 Subsequently, others have confirmed the observation of altered motility in diabetics with autonomic neuropathy. Dysphagia, nausea, vomiting, diarrhea, constipation, and fecal incontinence are common complaints among these patients. Differentiating the various causes of these symptoms can be difficult and should not necessarily be attributed to diabetes alone. Recent developments have led to a better understanding of the pathophysiology of diabetic gastroenteropathy. These advances have broadened the therapeutic options. In this chapter the focus is on the pathophysiology, clinical features, diagnosis, and treatment of the common gastrointestinal complications of diabetes.

PATHOGENESIS Despite recent scientific advances, the pathogenetic mechanisms causing diabetic gastrointestinal symptoms remain poorly understood. The most widely accepted theory is that an autonomic visceral neuropathy plays an important role in the development of symptoms. This is supported by the observation that diabetics with impaired gastric emptying have the same clinical constellation and radiographic findings as postvagotomy patients.2 The delay in gastric emptying may result from impaired myogenic contraction and altered cholinergic pathways. Nitric oxide may play a critical role in mediating gastric wall tone and compliance. In diabetic animal models, nitric oxide synthase activity is decreased, resulting in gastrointestinal motor dysfunction.3 In addition, electron microscopic studies have demonstrated a decreased density of unmyelinated axons in the vagus nerve of diabetics with gastroparesis.4 Furthermore, additional manifestations of autonomic neuropathy such as orthostatic hypotension, urinary bladder dysfunction, and impotence are often present in these patients. Other mechanisms have also been proposed. Hyperglycemia and electrolyte imbalance (e.g., hyperkalemia, hypokalemia) may affect nerve function and delay gut motility.5 Diabetics may also have altered production of various hormones such as motilin, pancreatic polypeptide, somatostatin, vasoactive intestinal peptide, calcitonin gene-related peptide, glucagon, and gastrin, which are known to influence gastrointestinal tract motility, absorption, and secretion.6,7 Finally, diabetic microangiopathy may play a causal role in the development of gastrointestinal tract complications.8 Surprisingly, none of the abnormalities caused by these proposed mechanisms correlate with the duration of diabetes, the severity of the neuropathy, the degree of blood sugar control, the presence of other diabetic complications, or the age or sex of the patient.9

ESOPHAGUS Although diabetics rarely complain of symptoms referable to the esophagus, the most typical symptoms are heartburn, and less frequently dysphagia and chest pain. Esophageal motor dysfunction in diabetics has been extensively studied using various techniques, including barium studies, nuclear radiographic methods, and manometry. Delayed esophageal transit and emptying have been demonstrated in up to 35% of both type 1 and type 2 diabetics.10,11 and 12 Esophageal manometry can reveal multiple aberrant motor patterns, including reduced or absent peristalsis, tertiary contractions, and decreased lower esophageal sphincter pressures.13 The correlation between investigative studies and clinical symptoms is inconsistent. Diabetics with dysphagia warrant complete investigation to rule out gastroesophageal reflux, benign or malignant esophageal strictures, and infectious esophagitis. Endoscopy is the preferred method of evaluation. Biopsy samples can be obtained and therapeutic maneuvers can be performed. Gastroesophageal reflux disease can be treated with H2-receptor antagonists or proton pump inhibitors alone or in combination with a prokinetic agent. If esophageal biopsy samples document candidal esophagitis, antifungal therapy can be instituted.

STOMACH The most common gastric disorder of diabetes is gastroparesis, which is manifested by postprandial fullness, vague epigastric pain, nausea, vomiting, heartburn, and anorexia. The onset of symptoms is usually insidious, but symptoms may arise acutely, especially in association with ketoacidosis. There is a strong correlation between the onset of gastroparesis and the development of other diabetic complications, such as peripheral neuropathy, retinopathy, and nephropathy. It is important to exclude other causes of impaired gastric emptying such as gastric outlet obstruction, peptic ulcer disease, side effects of pharmacologic agents, and uremia. Barium studies in patients with gastroparesis can demonstrate an elongated stomach with sluggish, ineffective, or absent peristalsis. At times, solid residue can be found in the stomach of these patients (Fig. 149-1). Because gastric emptying of liquids is usually normal, barium contrast studies are insensitive in assessing gastric dysmotility. By using solid, non-digestible radiopaque markers, the gastric emptying of nondigestible solids is found to be more prolonged than the emptying of digestible solids.14 Results of radioisotope gastric-emptying studies are more accurate because both liquid- and solid-phase motility are assessed.15 Because of the impaired gastric emptying, these patients are predisposed to the formation of bezoars, consisting of nondigestible foods. Gastroparesis leads to the irregular emptying of solids into the small intestine, resulting in erratic serum glucose control that may, in turn, worsen gastric emptying. To break this cycle, frequent small meals and close monitoring of the serum glucose concentration are advised.

FIGURE 149-1. Barium upper gastrointestinal radiograph in a patient with diabetic gastroparesis. The stomach is elongated and filled with food debris.

Should symptoms persist, various antiemetic and prokinetic agents can be used (Table 149-1). Metoclopramide is an antidopaminergic drug that inhibits dopamine-induced gastric smooth muscle relaxation.16 It also has cholinergic activity that probably acts by increasing the release of acetylcholine from postganglionic neurons or by enhancing acetylcholine receptor sensitivity in the gastric wall.17 In addition to its cholinergic effects, metoclopramide binds to medullary chemoreceptors in the brain, preventing emesis. As a result of its ability to penetrate the blood–brain barrier, side effects such as sedation, tremulousness, headache, galactorrhea, and amenorrhea due to hyperprolactinemia can occur. Several studies have shown that while metoclopramide is effective with short-term use, tachyphylaxis may develop.18 Domperidone is a potent dopamine antagonist lacking cholinergic activity. It enhances gastric emptying and has central antiemetic properties. Side effects of domperidone are infrequent because of its limited penetration of the blood–brain barrier, although hyperprolactinemia can occur. It is beneficial for long-term treatment and has an excellent safety profile.19,20 Erythromycin is another drug that is effective in relieving symptoms of gastroparesis. It is a macrolide antibiotic that has prokinetic properties attributed to its ability to bind to motilin receptors in the antrum and duodenum. Numerous studies have shown improvement in gastric emptying.21,22

TABLE 149-1. Prokinetic Agents Commonly Used for the Treatment of Diabetic Gastroparesis: Their Mechanisms of Action and Side Effects

Gastroparesis can be a difficult disorder to manage; symptoms may persist despite aggressive medical treatment. Patients who have failed prokinetic therapy may benefit from a percutaneous endoscopic gastrostomy and jejunostomy that allows gastric decompression and adequate nutritional support. This nonpharmacologic approach is well tolerated.23 Gastric pacing has been shown to improve gastric emptying. This technique entrains and improves gastric slow-wave activity and enhances slow-wave amplitude. This leads to a normalization of gastric dysmotility.24

SMALL INTESTINE The incidence of diarrhea in the diabetic population is ~22%, occurring primarily in patients with poorly controlled diabetes.25 The clinical features are nonspecific, consisting of chronic intermittent diarrhea alternating with normal bowel movements or constipation. Despite the chronicity, weight loss is distinctly uncommon and should elicit a search for other causes of diarrhea. The etiology of diarrhea in diabetes is multifactorial. Autonomic dysfunction plays a role in diabetic diarrhea. Specifically, a2-adrenergic receptors in enterocytes of the small and large intestines stimulate absorption of sodium chloride and inhibit bicarbonate secretion, resulting in a net influx of fluids and ions into the gut.26 A second cause may be bacterial overgrowth. Up to 4% of diabetics have achlorhydria,27 and many have delayed gastric and small bowel transit. These features alone or in combination can allow overgrowth of Escherichia coli, enterococci, and staphylococci. These bacteria deconjugate bile salts, preventing micelle formation and resulting in the development of fat malabsorption. There is also evidence that pancreatic insufficiency is a contributing factor.28 The evaluation of diabetics with diarrhea should be directed by their clinical presentation. Stool output should be quantified to separate true diarrhea from fecal incontinence. Stool cultures for enteric pathogens and 72-hour fecal fat studies should be performed. Patients with documented steatorrhea need to be evaluated for exocrine pancreatic insufficiency, bacterial overgrowth, or a primary malabsorptive disorder such as celiac disease. Pancreatic insufficiency can be diagnosed by a secretin/cholecystokinin stimulation test or a clinical response to a trial of pancreatic enzyme replacement. Bacterial overgrowth can be diagnosed by various methods, including breath tests and quantitative small bowel bacterial cultures. Alternatively, a response to a 2-week trial of antibiotics, such as tetracycline, can assist in the evaluation of this condition. Malabsorptive processes often require biopsy of the small bowel for diagnosis. Patients without steatorrhea or evidence of enteric infection fall into the category of “idiopathic diabetic diarrhea” and may benefit from a trial of clonidine, conventional antidiarrheal agents (loperamide, diphenoxylate, codeine), anticholinergic agents, or psyllium. CELIAC DISEASE Celiac disease has been reported to occur in diabetics with greater frequency than in the general population.29,30 The histo-compatibility antigens HLA-DR3 and HLA-DQB1*0201 have been linked to celiac disease and type 1 diabetes, suggesting a genetic predisposition to both conditions. The clinical manifestations of celiac disease and diabetic diarrhea are similar, although celiac disease may be distinguished by the absence of peripheral neuropathy, the onset of diarrhea before the initial diagnosis of diabetes, and steatorrhea, which is rare in diabetic diarrhea. In these patients, biopsy of the small bowel is essential. A presumptive diagnosis of celiac disease can be made if the biopsy shows flattened villi and increased cellularity in the lamina propria. A gluten-free diet will improve both the symptoms and histology and thus confirm the diagnosis. Serum immunoglobulin A (IgA) antiendomysial and IgA antigliadin antibodies may also be useful in diagnosing and monitoring patients with celiac disease. Several studies have shown 80% to 100% sensitivity and specificity of these antibodies for celiac disease, although their use cannot substitute for a histologic diagnosis.31,32 With successful treatment, titers of these antibodies decrease or even disappear and can, therefore, be used to monitor compliance with a gluten-free diet.

LARGE INTESTINE The most common gastrointestinal complaint among diabetics is chronic constipation, occurring in 20% to 60% of patients.24,33 The pathogenesis is poorly understood, although neurogenic dysfunction is thought to be the primary cause. Diminished or delayed postprandial myoelectric colonic activity is observed in diabetics without constipation, and absent activity is observed in those with constipation.34 Pharmacologic challenge with neostigmine or metoclopramide induces a return of the myoelectric response. The evaluation of chronic constipation in the diabetic should be guided by the patient's age, associated symptoms, and family history. In patients with rectal bleeding, occult blood in the stool, age older than 40, or a family history of colon cancer, malignancy should be ruled out by appropriate studies such as flexible sigmoidoscopy, barium enema, or colonoscopy. Treatment of constipation should include dietary fiber, stool softeners, and rarely laxatives. Fecal incontinence is common and often coincides with the onset of diarrhea.34a There is a strong association of incontinence with autonomic neuropathy. The pathogenesis is complex and incompletely understood. Initially it was thought that diminished rectal sensation with intact anal sphincter function allowed incontinent passage of stool.35 Later, anorectal manometry was used to demonstrate that fecal incontinence is caused by aberrant sphincter function; however, it is unclear whether autonomic neuropathy or abnormal intrinsic sphincter smooth muscle underlies the manometric findings.36 Therapy for fecal incontinence is difficult but responds to the control of diarrhea. If this fails, biofeedback and sphincter tone improving techniques can be used, although the results are not always successful.

GALLBLADDER DISEASE The management of gallstones in diabetics is a controversial subject because of the vast number of studies and the opposing conclusions reached. In autopsy studies, diabetics have approximately twice the incidence of gallstones as compared with the nondiabetic population.36a,37 Although no specific cause has been determined, factors that may predispose to cholelithiasis include impaired gallbladder emptying, altered glucose metabolism, hyperinsulinemia, and supersaturation of gallbladder bile levels with decreased bile acid concentrations.38 Other studies, however, indicate that comorbid conditions, such as obesity and hyperlipidemia, are confounding variables and that diabetics are not at increased risk of gallstone formation.39 Asymptomatic gallstones in the general population rarely lead to life-threatening complications.40 It is not clear whether this applies to diabetic patients. Several studies in diabetics have shown that when gallstones become symptomatic, morbidity and mortality are increased, primarily due to infectious complications.41 Previously, emergency cholecystectomy in the diabetic population was reported to have a fiveto ten-fold increase in mortality when compared with the nondiabetic population.42 It has been demonstrated that diabetes itself does not increase morbidity and mortality, but rather its associated conditions (e.g., renal failure, vascular disease) are responsible.43 Additionally, complication rates have decreased significantly because of the early detection of gallstones and cholecystitis by ultrasonography, the more effective use of antibiotics, and improved intraoperative hemodynamic monitoring.44 Because diabetics often have comorbid conditions, patients with biliary colic or cholecystitis should be advised to undergo cholecystectomy as soon as possible. In the past, it has been recommended that diabetics with asymptomatic gallstones also have surgery to reduce the chance of potential morbidity in the future. This concept has been carefully reevaluated in the past 10 years. Expectant management of diabetics with asymptomatic gallstones has been shown to be safer and less costly than prophylactic cholecystectomy.44,45 The introduction of laparoscopic cholecystectomy, while potentially less invasive than conventional surgery, should not influence the decision to operate on a patient with asymptomatic cholelithiasis until this concept is studied in more detail.

LIVER DISEASE The incidence of liver dysfunction in diabetics ranges from 28% to 39%.46 Viral hepatitis is seen in diabetics two to four times more frequently than in the general population.47 Previously, this was thought to be due to needle sticks; however, the association remains constant even in the era of disposable needles. More likely, it is caused by increased nosocomial exposures or diminished resistance to infection.48 Many of the medications used to treat diabetes and its complications can cause liver function test abnormalities. These drugs include the oral hypoglycemic sulfonylureas, many anti-hypertensive agents, and antibiotics. Both transaminase elevations and cholestatic jaundice can be seen as side effects of medication. Rarely does the side effect manifest clinically, and stopping the offending agent usually leads to a clinical and biochemical resolution. Increased glycogen deposition is the most common hepatic disorder found in diabetics. Excessive glycogen stores have been documented in up to 80% of persons with type 1 and 2 diabetes and is thought to be stimulated by exogenous hyperinsulinemia.47 A second liver lesion associated with diabetes is nonalcoholic steatohepatitis (NASH).49 This condition must be distinguished from hepatic steatosis, a noninflammatory process, which occurs in up to 50% of type 2 diabetics. The diagnosis of NASH requires a liver biopsy showing moderate to gross macrovesicular steatosis in association with inflammation. Additionally, there must be negligible alcohol consumption and absence of active viral infection. The mechanism of fat accumulation in type 1 diabetics is thought to be due to chronic hyperglycemia coupled with inadequate insulin levels that stimulate the release of fatty acids from adipose tissue. These fatty acids are then transported to the liver, where they are deposited into the hepatocyte. Fat accumulates because of increased hepatocellular triglyceride production and diminished secretion in the form of very low-density lipoprotein. In type 2 diabetes the mechanism of fat deposition is different and probably accounts for the increased incidence of NASH when compared with type 1 diabetes. In the type 2 diabetic population, obesity rather than hyperglycemia and hypoinsulinemia seems to be the critical factor. Intake of excessive dietary fats and carbohydrates leads to high levels of free fatty acids that are stored in hepatocytes.46 Accumulation of fatty acids may be responsible for hepatic inflammation because they are highly reactive and can damage biomembranes. Physical examination of the diabetic patient with NASH can be normal or reveal an enlarged, tender liver. The alkaline phosphatase value is typically normal or slightly elevated. Serum transaminase levels can be slightly elevated, but more significant elevations should prompt an evaluation of other etiologies of hepatitis (e.g., infections, drugs). Radiologic evaluation may be normal or show evidence of fatty infiltration of the liver. The diagnosis of NASH, however, can be confirmed only by liver biopsy. The course of NASH is typically indolent, but nearly 50% of patients develop hepatic fibrosis and up to 15% will develop cirrhosis. There is no proven therapy for NASH although gradual and sustained weight loss is recommended.50 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

Rundles RW. Diabetic neuropathy. Medicine 1945; 24:111. Zitomer BR, Gramm HF, Kozak GP. Gastric neuropathy in diabetes mellitus: radiologic observations. Metabolism 1968; 17:199. Takahashi T, Nakamura K, Itoh H, et al. Impaired nitric oxide synthase in the gastric myenteric plexus of spontaneously diabetic rats. Gastroenterology 1997; 113:1535. Guy R, Dawson J, Garrett J, et al. Diabetic gastroparesis from autonomic neuropathy: surgical considerations and changes in vagus nerve morphology. J Neurol Neurosurg Psychiatry 1984; 47:686. DeBoer SY, Masclee AAM, Lamers CBHW. Effect of hyperglycemia on gastrointestinal and gallbladder motility. Scand J Gastroenterol 1992; 27(Suppl 194):13. Hilsted J. Pathophysiology in diabetic autonomic neuropathy: cardiovascular, hormonal, and metabolic studies. Diabetes 1982; 31:730. Belai A, Lincoln, J, Burnstock G. Lack of vasoactive intestinal polypeptide and calcitonin gene-related peptide during electrical stimulation of enteric nerves in streptozotocin diabetic rats. Gastroenterology 1987; 93:1034. Liberski S, Koch K, Atnip R, et al. Ischemic gastroparesis: resolution after revascularization. Gastroenterology 1990; 99:252. Keshavarzian A, Iber F, Nasrallah S. Radionuclide esophageal emptying and manometric studies in diabetes mellitus. Am J Gastroenterol 1987; 82:625. Westin L, Lilja B. Oesophageal scintigraphy in patients with diabetes mellitus. Scand J Gastroenterol 1986; 21:1200. Borgstrom P, Olsson R, Sundkvist G, et al. Pharyngeal and oesophageal function in patients with diabetes mellitus and swallowing complaints. Br J Radiol 1988; 61:817. Horowitz M, Harding P, Maddox A, et. al. Gastric and oesophageal emptying in patients with type 2 diabetes mellitus. Diabetologia 1989; 32:151. Mandelstam P, Lieber A. Esophageal dysfunction in diabetic neuropathy gastroenteropathy. JAMA 1967; 201:88. Feldman M, Smith H, Simon T. Gastric emptying of solid radiopaque markers: studies in healthy subjects and diabetic patients. Gastroenterology 1984; 87:895. Siegel JA, Urbain JL, Adler LP, et al. Biphasic nature of gastric emptying. Gut 1988; 29:85. Peringer E, Jenner P, Donaldson JM, et al. Metoclopramide and dopamine receptor blockade. Neuropharmacology 1976; 15:463. Hay AM, Man WK. Effect of metoclopramide on guinea pig stomach: critical dependence on intrinsic stores of acetylcholine. Gastroenterology 1979; 76:492. Schade RR, Dugas MC, Lhostky DM, et al. Effect of metoclopramide on gastric liquid emptying in patients with diabetic gastroparesis. Dig Dis Sci 1985; 30:10. Soykan I, Sarosiek I, McCallum RW. The effect of oral domperidone therapy on gastrointestinal symptoms, gastric emptying, and quality of liver in patients with gastroparesis. Am J Gastroenterol 1997; 92(6):976. Silvers D, Kipnes M, Broadstone V, et al. Domperidone in the management of symptoms of diabetic gastroparesis. Clin Ther 1998; 20:438. Altomare DF, Rubini D, Pilot M-A, et al. Oral erythromycin improves gastrointestinal motility and transit after subtotal but not total gastrectomy for cancer. Br J Surg 1997; 84:1017. Janssens J, Peeters TL, Vantrappen G, et al. Erythromycin improves delayed gastric emptying in diabetic gastroparesis. N Engl J Med 1990; 322:1028. Kim CH, Nelson DK. Venting percutaneous gastrostomy in the treatment of refractory idiopathic gastroparesis. Gastrointest Endosc 1998; 47(1):67. McCallum RW, Chen JDZ, Lin Z, et al. Gastric pacing improves emptying and symptoms in patients with gastroparesis. Gastroenterology 1998; 114:456. Feldman M, Schiller L. Disorders of gastrointestinal motility associated with diabetes mellitus. Ann Intern Med 1983; 98:378. Ogbonnaya K, Arem R. Diabetic diarrhea: pathophysiology, diagnosis, and management. Arch Intern Med 1990; 150:262. Ungar B, Stocks AE, Martin FIR, et al. Intrinsic-factor antibody, parietal-cell antibody, and latent pernicious anemia in diabetes mellitus. Lancet 1968; 2:415. Newihi H, Dooley C, Saad C, et al. Impaired exocrine pancreatic function in diabetics with diarrhea and peripheral neuropathy. Dig Dis Sci 1988; 33:705. Walsh CH, Cooper BT, Wright AD, et al. Diabetes mellitus and celiac disease: a clinical study. Q J Med 1978; 47:89. Rensch MJ, Merenich JA, Lieberman M, et al. Gluten-sensitive enteropathy in patients with insulin-dependent diabetes mellitus. Ann Intern Med 1996; 124(6):564. Corroccio A, Iacono G, Montalto G, et al. Immunologic and absorptive tests in celiac disease: can they replace intestinal biopsy? Scand J Gastroenterol 1993; 28:673. McMillan SA, Haughton DJ, Biggart JD, et al. Predictive value for coeliac disease of antibody to gliadin, endomysium, and jejunum in patients attending for jejunal biopsy. Br Med J 1991; 303:1163. Goyal RK, Spiro HM. Gastrointestinal manifestations of diabetes mellitus. Med Clin North Am 1971; 55:1031. Battle WM, Snaper WJ Jr, Alavi A, et al. Colonic dysfunction in diabetes mellitus. Gastroenterology 1980; 79:1217.

34a. Folnaczny C, Riepl R, Tschop M, Landgraf R. Gastrointestinal involvement in patients with diabetes mellitus: part I. Epidemiology, pathophysiology, clinical findings. Z Gastroenterol 1999; 37:803. 35. Katz LA, Kaufman HJ, Spiro HM. Anal sphincter pressure characteristics. Gastroenterology 1967; 52:513. 36. Schiller LR, Santa Ana CA, Schmulen AC, et al. Pathogenesis of fecal incontinence in diabetes mellitus. N Engl J Med 1982; 307:1666. 36a. Ruhl CE, Everhart JE. Association of diabetes, serum insulin and C-peptide with gallbladder disease. Hepatology 2000; 31:299. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50.

Lieber MM. The incidence of gallstones and their correlation with other diseases. Ann Surg 1952; 135:394. DeSantis A, Attili AF, Corradini SG, et al. Gallstones and diabetes: a case-control study in a free-living population sample. Hepatology 1997;25:787. Persson GE, Thulin AJG. Prevalence of gallstone disease in patients with diabetes mellitus: a case-control study. Eur J Surg 1991; 157:579. Gracie WA, Ransohoff DF. The natural history of silent gallstones. N Engl J Med 1982; 307:798. Hickman MS, Schwesinger WH, Page CP. Acute cholecystitis in the diabetic. Arch Surg 1988; 123:409. Pellegrini C. Asymptomatic gallstones: does diabetes mellitus make a difference? Gastroenterology 1986; 91:245. Haff RC, Butcher HR, Ballinger WF. Factors influencing morbidity in biliary tract operations. Surg Gynecol Obstet 1971; 132:195. Friedman LS, Roberts MS, Brett AS, et al. Management of asymptomatic gallstones in the diabetic patient. Ann Intern Med 1988; 109:913. Ransohoff DF, Gracie MD, Wolfenson LB, et al. Prophylactic cholecystectomy or expectant management for silent gallstones. Ann Intern Med 1983; 99:199. Falchuk KR, Fiske S, Haggitt R, et al. Pericentral hepatic fibrosis and intracellular hyalin in diabetes mellitus. Gastroenterology 1980; 78:535. Stone B, Van Thiel D. Diabetes mellitus and liver disease. Semin Liver Dis 1985; 5:8. Khuri KG, Shamma MH, Abourizk N. Hepatitis B virus markers in diabetes mellitus. Diabetes Care 1985; 8:250. Luyckx FH, Lefebvre PJ, Scheen AJ. Non-alcoholic steatohepatitis: association with obesity and insulin resistance, and influence of weight loss. Diabetes Metab 2000; 26:98. Sheth SG, Gordon FD, Chopra S. Nonalcoholic steatohepatitis. Ann Intern Med 1997; 126(2):137.

CHAPTER 150 DIABETIC NEPHROPATHY Principles and Practice of Endocrinology and Metabolism

CHAPTER 150 DIABETIC NEPHROPATHY RALPH A. DEFRONZO Diabetic Nephropathy Glomerulosclerosis Vascular Involvement Tubulointerstitial Disease Pathogenesis of Diabetic Glomerulosclerosis Acquired (Metabolic/Hemodynamic) Theory Genetic Theory Clinical Course Early Changes in Renal Function Late Changes in Renal Function Staging of Diabetic Nephropathy Laboratory Abnormalities Proteinuria Glomerular Filtration Rate Glucosuria Hyperkalemia Metabolic Acidosis Clinical Correlations Diabetic Retinopathy Diabetic Neuropathy Hypertension Treatment of Diabetic Nephropathy Hypertension Urinary Tract Infection Neurogenic Bladder Intravenous Pyelography Blood Glucose Control Protein Restriction Dyslipidemia Dialysis Renal Transplantation Pathogenesis of Diabetic Nephropathy Chapter References

Of the ~1 million people with type 1 (previously called insulin-dependent) diabetes mellitus in the United States, 30% to 40% eventually develop end-stage renal failure,1 and there is little evidence that the incidence of diabetic nephropathy in type 1 diabetics has changed within the last decade.2 Type 2 (previously called non–insulin-dependent) diabetes mellitus is much more common, affecting some 15 million people.3 Among type 2 diabetic patients, the incidence of renal disease is lower (~5–10%).4,5 However, the incidence of nephropathy in type 2 diabetics varies considerably among ethnic groups, being three to four times higher in blacks and Hispanics and seven times higher in Native Americans when compared to whites.6,7 and 8 Nevertheless, in absolute terms, more type 2 than type 1 diabetic patients eventually progress to renal insufficiency. The magnitude of the problem can be readily appreciated when it is realized that every third person who is started on dialysis is diabetic,9 and that the Medicare payment for diabetics is ~$51,000 per year per patient.10

DIABETIC NEPHROPATHY Since the original description of diabetic nephropathy by Kim-melstiel and Wilson11 in 1936, numerous studies have reported the renal histologic changes and the clinical course, which is characterized by hypertension, edema, heavy albuminuria, and varying degrees of renal insufficiency.5 Three major histopathologic alterations have been described in the diabetic kidney: glomerulosclerosis, vascular involvement, and tubulointerstitial disease.12,13 and 14 GLOMERULOSCLEROSIS Glomerular involvement is the most characteristic feature of diabetic nephropathy and includes three distinctive lesions: diffuse intercapillary glomerulosclerosis, nodular glomerulosclerosis, and glomerular basement membrane (GBM) thickening (Fig. 150-1).

FIGURE 150-1. A, Diffuse diabetic glomerulosclerosis with marked mesangial matrix hyperplasia (thick arrows) and thickened glomerular basement membranes (thin arrows). No hypercellularity is evident. B, Nodules (thick arrow) usually are observed in the peripheral capillary loops and rarely are seen without the diffuse lesion. Note the prominent arteriolosclerosis in the efferent and afferent arterioles. C, Interstitial fibrosis, tubular atrophy (thick arrow), and thickening of the tubular basement membranes. (From DeFronzo RA. Diabetes and the kidney: an update. In: Olefsky JM, Sherwin RS, eds. Diabetes mellitus: management and complications. New York: Churchill Livingstone, 1985:161.)

DIFFUSE AND NODULAR LESIONS Diffuse intercapillary glomerulosclerosis, the most frequent histologic abnormality, is characterized by increased, periodic acid–Schiff–positive, eosinophilic material within the mesangial region (see Fig. 150-1A). The process is diffuse, involving the entire glomerulus, and generalized, affecting all glomeruli throughout the kidney. The earliest change is a widening of the mesangial matrix. With time, this mesangial material expands and coalesces, encroaching on adjacent capillary lumina. As this process progresses, entire glomeruli may become hyalinized. There is no increase in mesangial cell number, but mesangial cell volume is increased.15 In 40% to 50% of patients, the increase in mesangial matrix forms large acellular nodules at the center of peripheral glomerular lobules (see Fig. 150-1B). This nodular lesion is invariably associated with the diffuse lesion and is pathognomonic of diabetes mellitus. The nodular lesion correlates poorly with the severity of clinical renal disease; the best predictor of clinical renal disease is the diffuse lesion.13,15 Immunofluorescent staining reveals diffuse linear deposition of immunoglobulin G (IgG) along the GBM. Although this linear pattern of immunoglobulin deposition is reminiscent of anti-GBM nephritis, eluted IgG has no affinity for the GBM and most likely represents nonspecific trapping of filtered proteins. BASEMENT MEMBRANE THICKENING

Thickening of the GBM is an early and characteristic change of diabetic glomerulosclerosis (see Fig. 150-1A). Basement membrane thickening is not limited to the glomerulus but can be observed in capillaries throughout the body, including muscle, skin, and retina. The GBM is a collagen-like protein that is synthesized by visceral epithelium of the glomerulus.12 The integrity of the GBM is maintained by the continuous addition of new material from the epithelial aspect and by the simultaneous removal from the endothelial side by mesangial cells. The half-life of the GBM is ~100 days. Both increased epithelial synthesis and impaired removal by mesangial cells contribute to the basement membrane thickening. Biochemical analysis of the normal GBM has demonstrated significant differences in the collagen and protein content compared with basement membranes from nonrenal tissues.16 These differences in chemical composition account for the high glomerular filtration coefficient, which permits a water flux that is 50 to 100 times greater than in other capillaries, yet totally restricts the passage of serum proteins. The ability to restrict proteins selectively also is related to the strong negative charge provided by heparin sulfate and sialic acid residues within the GBM. A number of biochemical alterations in the composition of the GBM occur in diabetic nephropathy.16,17 Moreover, studies in alloxan diabetic rats have demonstrated increased activity of glucosyltransferase, the enzyme involved in the assembly of hydroxylysine-linked carbohydrate units. The restoration of normoglycemia with insulin normalized the activity of this enzyme. These results suggest that the accelerated synthesis of hydroxylysine-glucose-galactose subunits contributes to the GBM thickening; other investigators have been unable to demonstrate a significant alteration in amino acid or glucosylgalactose content in diabetic GBM. A polyantigenic expansion, involving all of the intrinsic components of the GBM and mesangium, has been demonstrated.18 This suggests an overall increase in the synthetic rate of normal basement membrane or a decrease in the rate of degradation. In diabetics with advanced nephropathy, the sialic acid and heparin sulfate proteoglycan content of the GBM is uniformly diminished. This decrease in anionic charge contributes to the increased clearance of negatively charged macromolecules, such as albumin.19 In vitro studies using isolated glomeruli have shown that glucose stimulates its own incorporation into capillary basement membrane in a dose-dependent fashion. Excessive glycosylation of the GBM may render collagen more resistant to degradation and contribute to the GBM thickening. VASCULAR INVOLVEMENT Accelerated renal arteriosclerosis and arteriolosclerosis are more characteristic of the diabetic than the nondiabetic kidney (see Fig. 150-1B). In the larger arteries, atheromatous changes are often advanced and may contribute to renal failure by causing ischemic parenchymal atrophy. In the smaller renal arterioles, hyaline thickening involves the afferent and efferent vessels. Although arteriosclerosis and arteriolosclerosis may be extensive, neither process correlates with the severity of glomerular change. TUBULOINTERSTITIAL DISEASE Although not usually appreciated, tubulointerstitial changes (see Fig. 150-1C) are common in the diabetic kidney, and in advanced cases, there is marked tubular atrophy, thickening of the tubular basement membrane, and interstitial fibrosis.14,20,21 Such changes are typically observed with renal ischemia, but in the diabetic, the tubular changes correlate poorly with the degree of vascular involvement and may be seen in their absence. The involved tubules often show thickening of the tubular basement membrane, and immunofluorescent studies have demonstrated deposition of IgG along the tubular basement membrane. The interstitial area surrounding the involved tubules is fibrotic and may contain a cellular infiltrate of lymphocytes and plasma cells. These changes progress with increasing duration of the diabetes.13,14 The interstitial reaction does not imply infection, such as pyelonephritis, although asymptomatic bacteriuria and pyelonephritis are twice as common in diabetics, particularly women, than in nondiabetics.22 This propensity to urinary tract infection probably results from various factors, including impaired renal blood flow, bladder dysfunction, interstitial scarring, impaired polymorpho-nuclear leukocyte function, defective leukocyte chemotaxis, and glucosuria, which enhance bacterial growth. It has been difficult to demonstrate any relationship between the frequency or severity of interstitial changes and urinary tract infection, and patients whose bacteriuria was not eradicated fared no worse than those in whom the bacteriuria was cured.23 The interstitial changes likely represent a direct representation of the diabetic state. The only tubular lesion that is characteristic of the diabetic state is the Armanni-Ebstein lesion, a vacuolization of the proximal tubular cells of the pars recta. These vacuoles contain glycogen and are most often observed in poorly controlled diabetics. Papillary necrosis, a special type of interstitial lesion, is observed in diabetics; indeed, diabetes mellitus accounts for over half of all cases.24 This is the result of ischemic damage to the inner medulla with eventual infarction and sloughing of part or all of the papilla. These patients may present with an acute fulminant illness, with fever, shock, flank pain, gross hematuria, pyuria, oliguria, and renal failure, or they may have a more subacute form characterized by microscopic hematuria, pyuria, and indolent renal failure. Urinary tract obstruction and pyelonephritis often complicate the diabetes and cause renal papillary necrosis. Because phenacetin has commonly been associated with papillary necrosis, chronic use of this analgesic should be avoided in the diabetic population.

PATHOGENESIS OF DIABETIC GLOMERULOSCLEROSIS Two theories have been proposed to explain the development of diabetic renal disease (Table 150-1). The first states that the renal changes are caused by a combination of metabolic and hemodynamic changes that result from a lack of insulin. The second argues that genetic factors are responsible for the alterations in renal structure and function.

TABLE 150-1. Metabolic and Genetic Theories of Diabetic Nephropathy

ACQUIRED (METABOLIC/HEMODYNAMIC) THEORY A large body of data now indicates that diabetic nephropathy is acquired. First, renal involvement is absent in newly diagnosed diabetics (Fig. 150-2). Basement membrane thickening, expansion of the mesangium, and vascular changes do not become evident until 2 to 3 years after the onset of diabetes.12,13,25 Second, typical changes of diabetic nephropathy occur in patients with pancreatic diabetes (e.g., hemochromatosis, chronic pancreatitis, pancreatectomy).26 Third, diabetic nephropathy can be produced in various animal models, including the rat, mouse, dog, and monkey, irrespective of the means of induction of diabetes. Fourth, transplantation of pancreatic islets into diabetic rats prevents the development of renal lesions. Fifth, when diabetic rat kidneys are transplanted into nondiabetic rats, the renal lesions resolve. Conversely, when normal kidneys are transplanted into diabetic rats, diabetic nephropathy ensues. Sixth, when kidneys from nondiabetic persons are transplanted into diabetics, all of the typical histologic changes of diabetic nephropathy develop within 2 to 5 years.25 The Diabetes Control and Complications Trial (DCCT), a study carried out in 1441 type 1 diabetics, has conclusively shown that intensive insulin therapy can prevent the development of nephropathy.27 Tight glycemic control reduced the occurrence of microalbuminuria (40–300 mg per day) and of overt albuminuria (>300 mg per day) by 39% and 54%, respectively. Intensified insulin therapy was effective both in primary and secondary prevention of nephropathy27 (Fig. 150-3).

FIGURE 150-2. Natural course of diabetic nephropathy. At the initial diagnosis, renal function and glomerular histology are normal. Within 2 to 3 years, increased mesangial matrix and basement membrane thickening are observed. Renal function remains normal until ~15 years, when proteinuria develops. This is an ominous sign and usually indicates advanced diabetic glomerulosclerosis. Within 5 years after the onset of proteinuria, elevation of the serum urea nitrogen and creatinine levels is observed, and within 3 to 5 years after the development of azotemia, half of the patients have advanced to end-stage renal insufficiency. (GFR, glomerular filtration rate.) (Modified from DeFronzo RA. Diabetes and the kidney: an update. In: Olefsky JM, Sherwin RS, eds. Diabetes mellitus: management and complications. New York: Churchill Livingstone, 1985:169.)

FIGURE 150-3. Effect of intensive versus conventional glycemic control with insulin on the albumin excretion rate in type 1 diabetics. Tight glycemic control significantly reduced the risk of developing microalbuminuria and macroalbuminuria in normoalbuminuria patients. (Reproduced from DCCT Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329:977.)

The importance of the metabolic and hemodynamic environment is underscored by a case report, in which kidneys were transplanted from a type 1 diabetic with established proteinuria and biopsy-proven diabetic nephropathy into two nondiabetic patients.28 After 7 months, renal function in both recipients was normal, proteinuria had disappeared, and renal biopsy showed resolution of the diffuse diabetic glomerulosclerosis. Similarly, others have demonstrated the regression of established diabetic nephropathy in diabetic patients after pancreatic transplantation. Lastly, pancreatic transplantation has reversed established lesions of diabetic nephropathy.29 GENETIC THEORY Proponents of the genetic theory cite the poor correlation between diabetic glomerulosclerosis and the degree of diabetic control. Evidence is available to support or to negate the role of metabolic control in the pathogenesis of diabetic complications. In short-term studies with a follow-up of 1 to 2 years, the achievement of good metabolic control has failed to reverse or to ameliorate established nephropathy or to prevent the emergence of new cases of diabetic renal disease.30,31 and 32 This information has been offered in favor of inheritance as the primary determinant of diabetic nephropathy. However, the follow-up period was 300 mg per day of albumin) or a diminished GFR within a period of 10 years.58,59 and 60 Patients with high-grade microalbuminuria (>100 mg per day) are at particularly high risk of developing renal impairment.47,48 and 49 Microalbuminuria also predicts the development of diabetic nephropathy in type 2 diabetics,47,48,49,52,62,63 and 64 although it is not as strong a predictor for these patients as it is for type 1 diabetics. In both type 1 and type 2 diabetics, microalbuminuria is a very strong predictor of macrovascular complications (i.e., heart attacks, stroke)48,62,65,66 and 67 (Fig. 150-4).

TABLE 150-2. Definition of Microalbuminuria*

FIGURE 150-4. Overall survival in 76 type 2 diabetic patients with varying degrees of microalbuminuria based on the urinary albumin concentration. Patients with microalbuminuria in the moderate to high range (two right columns) demonstrated a markedly shortened survival compared to an age-matched control population. (From Morgensen CE. Microalbuminuria predicts clinical proteinuria and early mortality in maturity-onset diabetes. N Engl J Med 1984; 311:89.)

Some caution must be exercised concerning the clinical interpretation of microalbuminuria. The upper limit of normal should not be considered as absolute, and common sense must be used in judging borderline cases. A number of factors (e.g., exercise, high blood pressure, urinary tract infection, and very poor diabetic control) elevate the urinary albumin excretion rate. Obviously, if such confounding factors are present, the finding of microalbuminuria does not necessarily imply incipient diabetic nephropathy. Although several authors have suggested that exercise-induced microalbuminuria is a harbinger of diabetic renal disease, there are few experimental data to support this. Another concern about the clinical interpretation of microalbuminuria centers on its development during the first 5 years in type 1 diabetics. In these patients, microalbuminuria is unlikely to have the same ominous prognostic significance as microalbuminuria that occurs 10 to 15 years after the onset of diabetes. This statement is not applicable to type 2 diabetics, who may have had their disease for many years before the diagnosis of diabetes was established. The interpretation of microalbuminuria in white type 2 diabetics is further complicated by the observation that ~25% of such patients have microalbuminuria, yet the incidence of renal disease in this population is only 5% to 10%.48,62,63 and 64,67 Although microalbuminuria predicts the development of renal disease in type 2 diabetics, it is not as strong a predictor as in type 1 diabetics. Thus, type 2 diabetics with microalbuminuria have a five-fold (compared to 20-fold in type 1 diabetics) increased risk of developing proteinuria over a 10-year period. However, in certain ethnic populations with a high incidence of diabetic nephropathy, including Native Americans, Mexican-Americans, and blacks, microalbuminuria is a very strong predictor of proteinuria and impaired renal function.6,49,66,68 From the routine laboratory standpoint, the earliest detectable manifestation of diabetic glomerulosclerosis is Albustix-positive proteinuria (see Fig. 150-2), which begins ~15 to 18 years after the diagnosis of diabetes.47,49,52,58,59 and 60,62,66 At this time, the GFR may still be normal or even elevated, but within a mean of 5 years after the onset of overt proteinuria, the GFR begins to decline, and the serum urea nitrogen and creatinine concentrations increase. Within 3 to 5 years after the first elevation of serum creatinine concentration, approximately half of the patients have progressed to end-stage renal insufficiency. Heavy proteinuria (>3 g per day) and the nephrotic syndrome are common, occurring in more than half of those who progress to end-stage renal failure.

LABORATORY ABNORMALITIES PROTEINURIA Proteinuria is the earliest laboratory manifestation of diabetic renal disease. When the urine albumin excretion exceeds 300 mg per day (corresponding to a urine protein excretion >500 mg per day), the subject is said to have overt diabetic nephropathy. Using fractional dextran clearances and urinary albumin and IgG excretion, the increased transglomerular flux of proteins has been shown to result from an augmented transglomerular ultrafiltration pressure gradient and an alteration in the molecular charge of the glomerular barrier that results from the loss of anionic charges (i.e., heparin sulfate and sialic acid residues) from the GBM.63,66,69,70,71 and 72 Neither the number of pores in the GBM nor the pore size is altered early in the course of diabetic nephropathy. At this stage, loss of barrier charge selectivity, without alteration of pore-size diameter, appears to be the primary factor that allows the escape of albumin without change in the clearance of other macromolecules, and improved glycemic control with intensive insulin therapy decreases the albumin excretion rate.73 With progressive proteinuria and the development of impaired renal function, glomerular charge and size selectivity is lost, there is an increase in pore size with the appearance of large molecular weight proteins in the urine, and enhanced transglomerular passage of plasma proteins contributes to progressive kidney damage. GLOMERULAR FILTRATION RATE Early in the natural history of diabetic nephropathy, before the onset of microalbuminuria, the GFR is increased47,48,49,62,63,64,65,66 and 67,70,72,74,75; this is believed to result

from increased intraglomerular pressure and increased glomerular capillary surface area.47,48,70,72 The GFR often remains elevated, although marked renal histologic changes are present. Thus, a decline in GFR from elevated to normal values without an improvement in metabolic control represents an ominous finding. The most widely used laboratory tests for GFR are the serum creatinine and the serum urea nitrogen concentrations. However, both are influenced by prerenal (i.e., extracellular fluid volume depletion) and postrenal (i.e., urinary tract obstruction) factors. More importantly, the GFR may decline by 40% to 50% before either test increases into the abnormal range. Therefore, many authors have advocated serial determinations of the creatinine clearance. In patients with renal failure of diverse causes, a plot of the reciprocal of the serum creatinine concentration as a function of time is useful in following the progression of renal disease. Pragmatically, the author advocates following the creatinine clearance in diabetic patients with normal serum creatinine levels. Once the serum creatinine concentration becomes elevated, either the reciprocal of the serum creatinine or the creatinine clearance can be followed. The GFR is determined by two factors: the net transmembrane ultrafiltration pressure and the glomerular ultrafiltration coefficient. The ultrafiltration pressure is governed by the balance between the transmembrane hydraulic pressure, which increases GFR, and the intraglomerular oncotic pressure, which decreases GFR. In animals and in humans, experimental diabetes elevates the transmembrane ultrafiltration pressure.63,70 Because the glomerular capillary oncotic pressure is normal or reduced in patients with diabetic nephropathy, this factor cannot explain the decrease in GFR. By exclusion, therefore, it has been concluded that the glomerular ultrafiltration coefficient must be reduced. The ultrafiltration coefficient is determined by the surface area available for filtration and the hydraulic permeability. Changes in the latter term have not been evaluated. Early in the course of type 1 diabetes mellitus, the total glomerular capillary surface area is increased and contributes to the glomerular hyperfiltration. However, in diabetics with clinically manifest renal disease, the mesangial matrix is greatly expanded, with obliteration of many capillary lumina. Anatomically, this reduction in surface area for filtration plays a significant role in the decline in GFR in patients with advanced diabetic nephropathy. GLUCOSURIA In normal persons, the maximum tubular reabsorptive capacity for glucose (TmG) varies inversely with the GFR, and similar observations have been made in recent-onset, type 2 diabetics.76 However, in long-term diabetics with reduced GFR and diabetic glomerulosclerosis, the TmG is raised and glucose in the urine becomes an unreliable means of monitoring the adequacy of diabetic control. Even in diabetic patients without renal disease, glucosuria correlates poorly with the plasma glucose concentration. HYPERKALEMIA Normally, potassium homeostasis depends on both renal and extrarenal mechanisms.77 Many factors predispose the diabetic to the development of hyperkalemia. The primary hormones that regulate the distribution of potassium between intracellular and extracellular compartments are insulin, epinephrine, and aldosterone. Because all of these hormones may be deficient, it is not surprising that hyperkalemia is so prevalent in the diabetic population.77 Furthermore, as the plasma glucose concentration rises, the increased tonicity causes an osmotic shift of fluid and electrolytes, primarily potassium, out of cells. Metabolic acidemia, common in diabetic patients, also predisposes to hyperkalemia. During the development of metabolic acidemia, about half of a hydrogen ion load is buffered within cells; this occurs in exchange for potassium ions.77 Renal mechanisms also contribute to the development of hyperkalemia in the diabetic patient. When the GFR falls to less than 15 to 20 mL per minute, the ability of the kidney to excrete potassium may become impaired.77 Many diabetic patients demonstrate a marked interstitial nephritis with prominent tubular atrophy and tubular basement membrane thickening.14,20,21 Because most urinary potassium is derived from distal and collecting tubular cell secretion, renal potassium excretion becomes impaired. Hypoaldosteronism is common in diabetics, particularly in those with evidence of impaired renal function.77,78 Most patients with hypoaldosteronism have no clinical symptoms and are diagnosed on routine laboratory screening or during evaluation for some other, unassociated illness. The baseline plasma aldosterone concentration is low and fails to increase normally after volume contraction. In most patients, the plasma renin level also is reduced and accounts for the hypoaldosteronism. However, in ~20% of patients with hypoaldosteronism, normal basal renin values have been reported.77,78 and 79 Moreover, all patients with the syndrome of hyporeninemic hypoaldosteronism have clinically significant hyperkalemia, which is the most potent stimulus to aldosterone secretion. This suggests that, along with hyporeninemia, there may be a primary adrenal defect in aldosterone secretion. This hypothesis is supported by the observation that the aldosterone response to angiotensin II and adrenocorticotropic hormone, agents that directly stimulate aldosterone secretion by the zona glomerulosa, is markedly impaired in persons with hypoaldosteronism.77,79 The cause of the defect in aldosterone secretion is unknown. A number of mechanisms have been suggested. Hyporeninemia could result from damage to the juxtaglomerulosa apparatus, impaired conversion of the biologically inactive renin precursor (“big renin”) to renin, decreased circulating cate-cholamine levels or autonomic neuropathy, diminished circulating prostaglandin levels, or chronic extracellular fluid volume expansion secondary to hyperglycemia and sodium retention. However, none of these explanations can satisfactorily account for the development of hyporeninemia in most diabetics. Impaired aldosterone secretion may be caused by intracellular potassium depletion within the zona glomerulosa of the adrenal gland secondary to insulin deficiency or insulin resistance.77,78 Several enzymatic steps involved in aldosterone biosynthesis are potassium dependent.77 A disruption of normal tubuloglomerular feedback mechanisms may also suppress aldosterone secretion.77 In diabetic patients with fasting plasma glucose levels in excess of 180 mg/dL, there is a large increase in the filtered load of glucose. The resultant osmotic diuresis would be expected to impair sodium and chloride reabsorption by all segments of the nephron, disrupting the normal tubulo-glomerular feedback control mechanism and leading to suppression of renin and aldosterone secretion. The complexity of the alterations in the renin-aldosterone axis in the diabetic is underscored by a report in which the aldosterone axis was characterized in 59 normokalemic diabetics with normal GFR.80 In half of the patients, both the renin and aldosterone responses were normal; 10% demonstrated diminished aldosterone secretion despite a normal renin response, 20% had impaired renin release with a normal aldosterone response, and 20% manifested impaired secretion of both renin and aldosterone. These results suggest that there are multiple defects in the renin-aldosterone axis and that the origin of hyperkalemia is multifactorial. Many drugs impair aldosterone secretion,77 and, if used in the diabetic, the plasma potassium concentration should be closely monitored. Of these, the b-blockers are the most widely known. Captopril, a converting enzyme inhibitor, has been advocated for the treatment of diabetic nephropathy. This agent infrequently causes hyperkalemia, and this complication should be monitored closely at the start of therapy.77 Another widely used class of drugs, the nonsteroidal antiinflammatory prostaglandin inhibitors, causes hyporeninemic hypoaldosteronism and hyperkalemia.77 The diuretic agents (e.g., spironolactone, triamterene, and amiloride) block potassium secretion by the renal tubular cell.77 METABOLIC ACIDOSIS Metabolic acidosis commonly is observed in diabetic patients.77,78,81 Metabolic acidosis can be divided into two broad categories: anion gap and nonanion gap or hyperchloremic. The anion gap is calculated by subtracting the concentrations of the major anions (chloride plus bicarbonate) from the major cation (sodium). The difference should not exceed 12 ± 2 mEq/L. If the value is >14 mEq/L, an anion gap acidosis is present. In the diabetic, there are three causes of an anion gap acidosis: renal insufficiency, ketoacidosis, and lactic acidosis. The latter two are discussed in Chapter 155. In diabetic nephropathy, as in other causes of chronic renal disease, when the GFR declines to 20 to 25 mL per minute, the ability of the kidney to excrete titratable acid becomes impaired, and the laboratory manifestation is an anion gap acidosis. The causes of an anion gap acidosis are well known to the clinician. Less well recognized is the frequent occurrence of a hyperchloremic (or nonanion gap) metabolic acidosis. The syndrome of hypoaldosteronism is commonly observed in diabetic patients, particularly those with renal impairment.77,78 and 79 Aldosterone is an important regulator of ammonia production and hydrogen ion secretion by the distal nephron, and more than half of the reported cases of hyporeninemic hypoaldosteronism present with a hyperchloremic metabolic acidosis.77,78 and 79 Because aldosterone does not affect urinary acidification, urine pH remains acidic (pH 5.4) despite systemic acidosis. This defect may be related to the prominent interstitial nephritis, to the tubular basement membrane thickening, or to an intracellular abnormality in any of the steps involved in hydrogen ion secretion. Other diabetics may present with a hyperchloremic metabolic acidosis due to widespread chronic interstitial nephritis with decreased ammonia production. Their urine pH is maximally acidic, indicating that the ability of the distal tubule and collecting duct to generate a steep pH gradient is intact. In these patients, the primary problem is impaired ammonia production, leading to a decrease in the total amount of hydrogen ion that can be excreted. In the absence of ammonia, which is one of the major urinary buffers, the urine pH is maximally acidic. Hyperkalemia also may cause a hyperchloremic metabolic acidosis by inhibiting ammonia synthesis by the renal tubular cell.

CLINICAL CORRELATIONS It often is stated that diabetic nephropathy is unusual in the absence of retinopathy, neuropathy, and hypertension. These associations were first popularized by Root

and colleagues,82 who referred to the triopathy of diabetes: nephropathy, retinopathy, and neuropathy. The occurrence of this triad was supported by other studies.83 However, the validity of this association has been challenged,84 and several reviews85,86 and 87 suggest a more variable association between the three microvascular complications. Diabetic patients with nephropathy invariably have retinopathy; thus, the diagnosis of diabetic nephropathy should not be made in the absence of retinopathy. This association has been referred to as the retinal syndrome. However, fewer than half of diabetic patients with retinopathy have clinically evident renal disease. The associations between diabetic neuropathy and retinopathy/nephropathy are even more variable. DIABETIC RETINOPATHY Diabetic retinopathy (e.g., hemorrhages, exudates, proliferative retinopathy) is present in most diabetic patients with end-stage renal failure who are admitted into dialysis or transplantation programs.87,88 However, when diabetic nephropathy is first diagnosed (i.e., documentation of persistent proteinuria or elevation of the serum creatinine concentration), 30% to 40% of patients do not have evidence of diabetic retinopathy by routine ophthalmologic examination,83,84,87,89 even though fluorescein angiography will demonstrate typical diabetic abnormalities in essentially 100% of patients. As the renal disease progresses, however, diabetic retinopathy appears to accelerate.86,87 and 88 In one study, evidence of retinopathy was noted in only 61 of 150 diabetic patients when nephropathy was first diagnosed, but retinal involvement developed in another 42 during the follow-up.89 In diabetic patients placed on hemodialysis, a rapid progression of diabetic retinopathy often ensues. A dissociation between retinopathy and nephropathy is evident if one examines diabetic patients with established retinopathy. After 15 to 20 years, >80% of diabetic patients have evidence of retinopathy, but as many as 30% to 50% have no laboratory evidence of renal disease. DIABETIC NEUROPATHY The association between diabetic nephropathy and neuropathy is much less impressive than between diabetic nephropathy and retinopathy. In patients who have had diabetes for 20 years, approximately one-half have some evidence of neuropathic involvement.87,90,91 and 92 In diabetics with end-stage renal failure, the incidence of neuropathy varies from 50% to 90%, depending on how carefully one looks for evidence of diabetic neuropathic involvement. Peripheral neuropathy is more common than autonomic neuropathy. In one study, symptomatic diabetic gastroparesis was observed in ~50% of diabetic patients treated with dialysis.93 However, the specificity of these neuropathologic abnormalities, especially those relating to peripheral neuropathy, must be questioned because dialysis and transplantation often lead to reversal of the abnormalities,94 suggesting a uremic etiology. As with retinopathy, uremia appears to exacerbate the progression of diabetic neuropathy. When diabetic nephropathy is first diagnosed, fewer than half of patients have clinically evident diabetic neuropathy. HYPERTENSION The incidence of hypertension, as well as the relationship between hypertension and renal disease, are very different in type 1 and type 2 diabetes mellitus. In newly diagnosed type 2 diabetics, ~50% to 60% have hypertension, whereas 2 mg/dL) or heavy proteinuria are at greatest risk. With the judicious use of ultrasound, radio-nuclide studies, and computed tomographic scanning without contrast, most of the information necessary to ensure adequate diagnosis and treatment can be obtained. If diabetic patients must receive radiocontrast dye, hydration with normal saline should be started 12 to 24 hours before the procedure.140 BLOOD GLUCOSE CONTROL GLYCEMIC CONTROL (TABLE 150-5)

TABLE 150-5. Recommended Levels of Glycemic Control in Diabetic Individuals

Diabetic nephropathy and other microvascular complications (retinopathy and neuropathy) do not occur in the absence of poor glycemic control. The product of the mean day-long blood glucose concentration (as reflected by the hemoglobin A1c [HbA1c]) and the duration of diabetes provide the total hyperglycemic burden and are the best predictors of microvascular complications.87,141 Poor glycemic control is a major risk factor for diabetic nephropathy in both type 1 and type 2 diabetics,47,87,119,141,142 and 143 and diabetic nephropathy is uncommon when the glycosylated hemoglobin is maintained at 300 mg per day of albumin) or renal insufficiency has uniformly been ineffective in halting the relentless progression to end-stage renal failure in type 1 diabetes.47,89,154,155,156 and 157 These findings suggest that intensive insulin therapy in humans is most effective when started early, that is, before or during the phase of microalbuminuria (30–300 mg per day; see Fig. 150-2), before the advanced lesions of diabetic nephropathy have become manifest. From the available evidence, it is recommended that the HbA1c in diabetic patients with microalbuminuria and without microalbuminuria be maintained at 50% of cases). Gas in the bladder wall is produced by bacterial fermentation. Gross hematuria occurs in approximately half of the cases. In most patients the infection responds to appropriate antibiotic therapy.46 Autopsy studies from the 1930s indicated that pyelonephritis was more common in diabetic patients than in the general population. Pyelonephritis may be complicated in rare cases by renal papillary necrosis, and the sloughed papillae may subsequently obstruct urine outflow. A rare, severe form of renal parenchymal infection, emphysematous pyelonephritis, occurs more frequently in diabetic patients (Fig. 152-1).47 Gas is seen in the renal tissue, and the kidney may be completely destroyed. Infection is most often caused by Escherichia coli or Klebsiella. Failure to respond to appropriate antibiotic therapy may necessitate nephrectomy. Even after nephrectomy, mortality in these severely ill patients remains high. At least one-third of patients with perinephric abscesses have diabetes. The advent of computed tomography has facilitated delineation of such processes (Fig. 152-2).

FIGURE 152-1. A middle-aged diabetic woman presented with fever and gross pyuria. A coned-down anteroposterior radiograph of the right kidney shows gas within the parenchyma of the kidney (K) and subcapsular gas (arrows). A nephrectomy was performed within 24 hours. The kidney was diffusely involved with multiple small, gas-filled microabscesses. (Courtesy of Dr. Michael Hill.)

FIGURE 152-2. Computed tomographic demonstration of a perinephric abscess (A) arising from a transplanted kidney (K). (P, psoas muscle.)

Although the development of a neurogenic bladder is a frequent complication of diabetes, ascribing any increased tendency toward urinary infections to this condition has been difficult except as related to the need for instrumentation.48 Additional factors such as nonneurogenic bladder outlet disorders probably contribute to the risk of infection. Animal studies suggest that osmotic diuresis, as seen in uncontrolled diabetes, predisposes to ascending infections, possibly because of vesi-coureteral reflux or dilution of urinary bacteriostatic activity.49 EMPHYSEMATOUS CHOLECYSTITIS Diabetes is present in approximately one-third of cases of emphysematous cholecystitis, in which gas is found within the gallbladder lumen, gallbladder wall, or pericholecystic tissues.50 Organisms recovered at surgery include streptococci, clostridia, and gram-negative bacilli. In contrast to the more common form of cholecystitis, this type of cholecystitis occurs more often in men, and gangrene of the gallbladder is more frequent.

FUNGAL INFECTIONS RHINOCEREBRAL MUCORMYCOSIS Diabetic ketoacidosis is a major predisposing factor in rhinocerebral mucormycosis, a rare infection caused by fungi of the order Mucorales (e.g., Mucor, Rhizopus, and Absidia).51 The process begins on the palate or in the nasal passages, where a black eschar may be seen. Infection spreads to the paranasal sinuses, orbits, or brain. Necrosis is prominent because of blood vessel invasion. Fever, lethargy, headache, and facial swelling are often present. Orbital invasion causes proptosis, decreased ocular motion, and vision loss. Thrombosis of the venous sinuses or carotid system may occur. Diagnosis is made by biopsy of affected tissues. Broad, nonseptate hyphae that tend to branch at right angles are characteristic (Fig. 152-3).

FIGURE 152-3. Broad, nonseptate hyphae with right-angle branching seen in mucormycosis.

Treatment requires prompt control of diabetes and correction of acidosis, aggressive surgical debridement of infected tissues, and the systemic administration of amphotericin B aggressively. Studies suggest that diabetic ketoacidosis induces a defect in the ability of macrophages to inhibit spore germination and impairs both leukocyte chemotaxis to fungal products and leukocyte-mediated hyphal injury. These organisms may also cause a fungal pneumonia in patients with diabetes mellitus, which is associated with high fatality rates.13 FUNGAL URINARY TRACT INFECTION Funguria with Candida species is common in diabetic patients, particularly with urethral catheterization, with the prior use of antibiotics, and in the presence of glucosuria.52 Yeast infections may be asymptomatic or may produce symptoms that are indistinguishable from those of bacterial infection. Occasionally, such infections ascend to the renal parenchyma, cause ureteral obstruction, or disseminate hematogenously.53 Systemic therapy is required with upper tract disease.54 Candida may even cause emphysematous pyelonephritis.51a CHAPTER REFERENCES 1. Sentochnik DE, Eliopoulos GM. Infection and diabetes. In: Kahn CR, Weir GC, eds. Joslin's diabetes mellitus, 13th ed. Philadelphia: Lea & Febiger, 1994:867. 1a. Joshi N, Caputo GM, Weitekamp MR, Karchmer AW. Infections in patients with diabetes mellitus. N Engl J Med 1999; 341:1906. 2. Coopan R. Infection and diabetes. In: Marble A, Krall LP, Bradley RF, et al., eds. Joslin's diabetes mellitus, 12th ed. Philadelphia: Lea & Febiger, 1985:737. 3. Margarit C, Rimola A, Gonzalez-Pinto I, et al. Efficacy and safety of oral low-dose tacrolimus treatment in liver transplantation. Transpl Int 1998;11 Suppl 1:S260. 4. Walli R, Herfort O, Michl GM, et al. Treatment with protease inhibitors associated with peripheral insulin resistance and impaired oral glucose tolerance in HIV-1 infected patients. AIDS 1998; 12(15):F167. 5. Gunther SF, Gunther SB. Diabetic hand infections. Hand Clin 1998; 14(4):647. 6. Borger MA, Rao V, Weisel RD, et al. Deep sternal wound infection: risk factors and outcomes. Ann Thorac Surg 1998; 65(4):1050. 7. Pomposelli JJ, Baxter JK 3rd, Babineau TJ, et al. Early postoperative glucose control predicts nosocomial infection rate in diabetic patients. J Parenter Enteral Nutr 1998; 22(2):77. 8. Babineau TJ, Bothe A Jr. General surgery considerations in the diabetic patient. Infect Dis Clin North Am 1995; 9(1):183. 9. Zerr KJ, Furnary AP, Grunkemeier GL, et al. Glucose control lowers the risk of wound infection in diabetics after open heart operations. Ann Thorac Surg 1997; 63(2):356. 10. Chandler PI, Chandler SD. Pathogenic carrier rate in diabetes mellitus. Am J Med Sci 1977; 273:259. 11. Mackowiak PA, Martin RM, Jones SR, Smith JW. Pharyngeal colonization by gram-negative bacilli in aspiration-prone persons. Arch Intern Med 1978; 138:1224. 12. Breen JD, Karchmer AW. Staphylococcus aureus infections in diabetic patients. Infect Dis Clin North Am 1995; 9(1):11. 13. Koziel H, Koziel MJ. Pulmonary complications of diabetes mellitus. Infect Dis Clin North Am 1995; 9(1):65. 14. Brayton RG, Stokes PE, Schwartz MS, Louria DB. Effect of alcohol and various diseases on leukocyte mobilization, phagocytosis, and intracellular bacterial killing. N Engl J Med 1970; 282:123. 15. Molenaar DM, Palumbo PJ, Wilson WR, Ritts RE Jr. Leukocyte chemotaxis in diabetic patients and their non-diabetic first-degree relatives. Diabetes 1976; 25:880. 16. Bagdade JD, Stewart M, Walters E. Impaired granulocyte adherence: a reversible defect in host defense in patients with poorly controlled diabetes. Diabetes 1978; 27:677. 17. Davidson NJ, Sowden JM, Fletcher J. Defective phagocytosis in insulin controlled diabetics: evidence for a reaction between glucose and opsonising proteins. J Clin Pathol 1984; 37:783. 18. Repine JE, Clawson CC, Goetz FC. Bactericidal function of neutrophils from patients with acute bacterial infections and from diabetics. J Infect Dis 1980; 142:869. 19. Jubiz W, Draper RE, Gale J, Nolan G. Decreased leukotriene B 4 synthesis by polymorphonuclear leukocytes from male patients with diabetes mellitus. Prostaglandins Leukot Med 1984; 14:305. 20. Qvist R, Larkins RG. Diminished production of thromboxane B2 and prostaglandin E by stimulated polymorphonuclear leukocytes from insulin-treated subjects. Diabetes 1983; 32:622. 21. Delamaire M, Maugendre D, Moreno M, et al. Impaired leucocyte functions in diabetic patients. Diabet Med 1997; 14(1):29. 22. Geisler C, Almdal T, Bennedsen J, et al. Monocyte functions in diabetes mellitus. Acta Pathol Microbiol Immunol Scand 1982; 90:33. 23. Stewart J, Collier A, Patrick AW, et al. Alterations in monocyte receptor function in type 1 diabetic patients with ketoacidosis. Diabetic Med 1991; 8:213. 24. Alexiewicz JM, Kumar D, Smogorzewski M, et al. Polymorphonuclear leukocytes in non-insulin dependent diabetes mellitus: abnormalities in metabolism and function. Ann Intern Med 1995; 123:919. 25. Tebbs SE, Lumbwe CM, Tesfaye S, et al. The influence of aldose reductase on the oxidative burst in diabetic neutrophils. Diabetes Res Clin Pract 1992; 15:121. 26. Tebbs SE, Gonzales AM, Wilson RM, et al. The role of aldose reductase inhibition in diabetic neutrophil phagocytosis and killing. Clin Exp Immunol 1991; 84:482. 27. Sato N, Kashima K, Uehara Y, et al. Epalrestat, an aldose reductase inhibitor, improves an impaired generation of oxygen-derived free radicals by neutrophils from poorly controlled NIDDM patients. Diabetes Care 1997; 20(6):995. 28. Wykretowicz A, Wierusz-Wysocka B, Wysocki J, et al. Impairment of the oxygen-dependent microbicidal mechanisms of polymorphonuclear neutrophils in patients with type 2 diabetes is not associated with increased susceptibility to infection. Diabetes Res Clin Pract 1993; 19:195. 29. Speert DP, Silva J Jr. Abnormalities of in vitro lymphocyte response to mitogens in diabetic children during acute ketoacidosis. Am J Dis Child 1978; 132:1014. 30. Casey J, Strum C Jr. Impaired response of lymphocytes from non–insulin-dependent diabetics to staphage lysate and tetanus antigen. J Clin Microbiol 1982; 15:109. 31. Kolterman OG, Olefsky JM, Kurahara C, Taylor K. A defect in cell-mediated immune function in insulin-resistant diabetic and obese subjects. J Lab Clin Med 1980; 96:535. 32. Li YM. Glycation ligand binding motif in lactoferrin. Implications in diabetic infection. Adv Exp Med Biol 1998; 443:57. 33. Cooper G, Platt R. Staphylococcus aureus bacteremia in diabetic patients: endocarditis and mortality. Am J Med 1982; 73:658. 34. Butler JC, Breiman RF, Campbell JF, et al. Pneumococcal polysaccharide vaccine efficacy. An evaluation of current recommendations. JAMA 1993; 270(15):1826. 34a. Centers for Disease Control and Prevention. Influenza and pneumococcal vaccination rates among persons with diabetes mellitus—United States, 1997. MMWR Morb Mort Wkly Rep 1999; 48:961. 35. Pablos-Mendez A, Blustein J, Knirsch CA. The role of diabetes mellitus in the higher prevalence of tuberculosis among Hispanics. Am J Public Health 1997; 87(4):574. 36. Olmos P, Donoso J, Rojas N, et al. Tuberculosis and diabetes mellitus: a longitudinal retrospective study in a teaching hospital. Rev Med Chile 1989; 117:979.

37. 38. 39. 40. 41. 42. 43.

Doroghazi RM, Nadol JB Jr, Hyslop NE Jr, et al. Invasive external otitis: report of 21 cases and review of the literature. Am J Med 1981; 71:603. Johnson MP, Ramphal R. Malignant external otitis: report on therapy with ceftazidime and review of therapy and prognosis. Rev Infect Dis 1990; 12:173. Lang R, Goshen S, Kitzes-Cohen R, et al. Successful treatment of malignant external otitis with ciprofloxacin: report of experience with 23 patients. J Infect Dis 1990; 161:537. Giamarellou H. Malignant otitis externa: the therapeutic evolution of a lethal infection. J Antimicrob Chemother 1992; 30:745. LeFrock JL, Mulavi A. Necrotizing skin and subcutaneous infections. J Antimicrob Chemother 1982; 9:183. Davies HD, McGeer A, Schwartz B, et al. Invasive group A streptococcal infections in Ontario, Canada. Ontario Group A Streptococcal Study Group. N Engl J Med 1996; 335(8):547. Patterson JE, Andriole VT. Bacterial urinary tract infections in diabetes.Infect Dis Clin North Am 1995; 9(1):25.

43a. Geerlings SE, Stolk RP, Camps MJ, et al. Asymptomatic bacteriuria may be considered a complication in women with diabetes. Diabetes Care 2000; 23:744. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

Pometta D, Rees SB, Younger D, Kass EH. Asymptomatic bacteriuria in diabetes mellitus. N Engl J Med 1967; 276:1118. Lindberg U, Bergström AL, Carlsson E, et al. Urinary tract infection in children with type I diabetes. Acta Paediatr Scand 1985; 74:85. Lee JB. Cystitis emphysematosa. Arch Intern Med 1960; 105:150. Evanoff GV, Thompson CS, Foley R, Weinman EJ. Spectrum of gas within the kidney: emphysematous pyelonephritis and emphysematous pyelitis. Am J Med 1987; 83:149. Sobel JD. Pathogenesis of urinary tract infections. Infect Dis Clin North Am 1987; 1:751. Levison ME, Pitsakis PG. Effect of insulin treatment on the susceptibility of the diabetic rat to Escherichia coli-induced pyelonephritis. J Infect Dis 1984; 150:554. Sarmiento RV. Emphysematous cholecystitis. Arch Surg 1966; 93:1009. Nussbaum ES, Hall WA. Rhinocerebral mucormycosis: changing patterns of disease. Surg Neurol 1994; 41:152. Kauffman CA, Vasquez JA, Sobel JA, et al. Prospective multicenter surveillance study of funguria in hospitalized patients. Clin Infect Dis 2000; 30:14. Hildebrand TS, Nibbe L, Frei U, Schindler R. Bilateral emphysematous pyelonephritis caused by Candida infection.Am J Kidney Dis1999; 33:E10. Vazquez JA, Sobel JD. Fungal infections in diabetes.Infect Dis Clin North Am 1995; 9(1):97.

CHAPTER 153 DIABETES AND THE SKIN Principles and Practice of Endocrinology and Metabolism

CHAPTER 153 DIABETES AND THE SKIN ROBERT J. TANENBERG AND RICHARD C. EASTMAN Noninfectious Complications of Diabetes Xanthomas Scleredema Waxy Skin and Reduced Joint Mobility Perforating Dermatoses Yellow Skin Diabetic Dermopathy Necrobiosis Lipoidica Diabeticorum Granuloma Annulare Bullosis Diabeticorum Acanthosis Nigricans Acrochordons Necrolytic Migratory Erythema Other Associated Dermatoses Infections of the Skin and Soft Tissues Necrotizing Fasciitis Malignant External otitis Rhinocerebral Mucormycosis Erythrasma Dermatologic Complications of the Treatment of Diabetes Immunologic Reactions to insulin Complications of Continuous Subcutaneous insulin Administration Reactions to Oral Antidiabetic Agents Dermatologic Findings in Secondary Diabetes Chapter References

The skin of the patient with diabetes may manifest various abnormalities and is prone to the development of certain infections.1,2 Additionally, the injection or infusion of insulin and the administration of oral hypoglycemic agents may lead to cutaneous reactions. Occasionally, skin and soft tissue changes may alert the clinician to the presence of diabetes.

NONINFECTIOUS COMPLICATIONS OF DIABETES Various noninfectious skin diseases may occur in the patient with diabetes. Some conditions, such as xanthomas in patients with hypertriglyceridemia, carotenoderma, and reduced skin elasticity, are directly related to metabolic abnormalities, while other conditions, such as diabetic dermopathy, necrobiosis lipoidica diabeticorum, granuloma annulare, and bullosis diabeticorum, are of uncertain etiology and may occur in nondiabetics, in patients with impaired glucose tolerance, and in patients with overt diabetes. In these conditions, subtle metabolic factors, environmental factors, or genetic factors probably influence the development of the skin abnormality. The relationship between metabolic control and these cutaneous manifestations is not clear.1,3 XANTHOMAS Xanthomas due to hypertriglyceridemia may occur in the uncontrolled diabetic. Triglyceride levels are usually >1000 mg/dL when xanthomas are present. Concomitant lipemia retinalis is usually noted on funduscopic examination. Eruptive xanthomas characteristic of hypertriglyceridemia occur in clusters primarily on the extensor surfaces of the extremities and over the buttocks. They manifest as red-yellow, 2- to 5-mm maculopapular lesions (Fig. 153-1). A red inflammatory border is characteristic. Initially, the lesions contain more triglycerides and free fatty acids than cholesterol; later, as the xanthomas begin to resolve, cholesterol predominates.4

FIGURE 153-1. Eruptive xanthomas on the buttocks of a patient with hypertriglyceridemia and uncontrolled diabetes mellitus.

Eruptive xanthomas secondary to uncontrolled diabetes usually clear with improved metabolic control. Patients with familial dyslipidemias may require therapy with drugs to reduce triglyceride levels to a satisfactory range (see Chap. 164). SCLEREDEMA Scleredema (also known as scleredema diabeticorum) is asymmetric nonpitting induration of the skin on the posterolateral aspects of the neck, shoulders, and upper back.5,6 Although the condition is often preceded by a streptococcal infection in the nondiabetic, this is rarely the case in the patient with diabetes.1,3,5 Males with diabetes are four times more likely to develop this lesion than are diabetic females. Patients are frequently obese and often have microvascular and macrovascular complications of diabetes. Spontaneous resolution in 6 months to 2 years is common in the nondiabetic, but rarely occurs in patients with diabetes.5 There is no known treatment. WAXY SKIN AND REDUCED JOINT MOBILITY Reduced skin elasticity in patients with diabetes has long been recognized; it has a “hidebound” quality, resembling the skin of patients with scleroderma.5 The condition is associated with limited joint mobility and with poor metabolic control. The correlation of these abnormalities with concentrations of glycosylated hemoglobin suggests that metabolic factors influence the development of the skin and joint changes. Improvement may occur with improved metabolic control5 (see Chap. 211). This condition is distinct from Dupuytren contracture of the palmar fascia, which, although increased in incidence in diabetic individuals, is extremely common in the nondiabetic population. PERFORATING DERMATOSES Kyrle disease, perforating folliculitis, and reactive perforating collagenosis are disorders that may be more common in patients with diabetes, particularly in those with chronic renal failure.1 In a review of acquired reactive perforating collagenosis, 72% of the patients had diabetes with micro- or macrovascular complications.7 These disorders are characterized by papules with a central plug that extrudes keratin or collagen. Patients typically complain of pruritus, which may be severe. Lesions have a predilection for the extensor surfaces of the upper and lower extremities, trunk, and buttocks. The lesions are often refractory to therapy, although retinoic acid with or

without ultraviolet light may improve some cases.7 YELLOW SKIN The elastic tissue of patients with diabetes may demonstrate yellow discoloration possibly secondary to deposition of carotenoids, which are pigments present in green and yellow vegetables. The frequency may be as high as 10% in patients with diabetes.1,3 Blood carotene levels are usually elevated. Hypothyroidism, hypogonadism, hypopituitarism, bulimia, and anorexia nervosa are also associated with carotenoderma and should be considered in the differential diagnosis. Brown discoloration of the nail fold due to advanced glycosylation end products also may occur in patients with diabetes.3 DIABETIC DERMOPATHY Dermopathy is a common complication of diabetes, occurring twice as often in males.1,3 The occurrence of the lesions in patients without diabetes is evidence that factors other than hyperglycemia are causative. The high frequency of other complications of diabetes in affected patients suggests that microangiopathy is an important factor in the pathogenesis. Half of the patients have concurrent retinopathy, neuropathy, and nephropathy. Basement membrane thickening is found in areas of dermopathy, yet adjacent areas of unaffected skin show similar changes. The predominant occurrence of the lesions on the lower extremities implicates trauma, neuropathy, and basement membrane thickening. However, the lesions are not reliably reproduced by trauma. Heat or cold traumatization of the extremities in diabetics of more than 10 years' duration may induce lesions that resemble those of dermopathy.8 Although the lesions occur mostly on the lower extremities (“shin spots”), they also may occur on the arms, thighs, trunk, and scalp; they begin as dull red macules or papules, which are 5 to 12 mm in diameter.3 Crops of four to five lesions may appear over a period of a week and then persist or slowly resolve (Fig. 153-2). As the lesions evolve, they become shallow, depressed, hyperpigmented scars. The lesions are asymptomatic and require no treatment.

FIGURE 153-2. Diabetic dermopathy on the anterior tibial surfaces. The lesions are pigmented and slightly depressed below the surrounding skin.

NECROBIOSIS LIPOIDICA DIABETICORUM Necrobiosis (degeneration of collagen) is an uncommon complication of diabetes of unknown etiology. Although it occurs in only 0.3% of all patients with diabetes, 65% of the patients presenting with this lesion have diabetes. Of the remaining patients, many have impaired glucose intolerance or a positive family history of diabetes. It is three times more common in women. A review of the literature concluded that “necrobiosis lipoidica diabeticorum is usually associated with poor glucose control and that tighter glucose control, as currently practiced, might improve or prevent the disorder.”9 Pathologically the lesions show degeneration of collagen, granulomatous inflammation of subcutaneous tissues and of blood vessels, capillary basement membrane thickening, and obliteration of vessel lumina.The yellow color in the central area of the lesions is most likely secondary to thinning of the dermis, making subcutaneous fat more visible. The lesions of necrobiosis occur most commonly in the pretibial areas and are bilateral in 75% of the cases. They begin as red-brown papules that enlarge and coalesce to form irregular lesions, which may be several centimeters in diameter (Fig. 153-3). The lesions ulcerate in 35% of the cases; although usually painless, they may be painful and become infected when ulcerated. For this reason, trauma to the anterior tibial area should be avoided. As the lesions evolve, they appear as atrophic plaques with a thin, translucent surface manifesting a “porcelain-like sheen.”

FIGURE 153-3. Necrobiosis lipoidica diabeticorum. These typical lesions were initially noted in this 9-year-old girl several months before the diagnosis of type 1 diabetes. (Courtesy of Dr. William Burke.)

Various therapies have been used: fluorinated corticosteroids (applied topically, under occlusive dressings, or injected into the lesion), salicylates with dipyridamole, pentoxifylline, systemic corticosteroids, nicotinamide, clofazimine, and ticlopidine.10 Topical benzoyl peroxide, seaweed-based alginate dressings, and hydrocolloid occlusive dressings may be helpful.3 Grafting of ulcerated areas may occasionally be necessary, although recurrences are common. Spontaneous resolution of necrobiosis occurs infrequently, but the lesions may fade in appearance. GRANULOMA ANNULARE Granuloma annulare is a dermatitis with histologic similarities to necrobiosis lipoidica diabeticorum. Clinically, these lesions may initially have a similar appearance; subsequently, as they enlarge, differences become apparent. Granuloma annulare lesions do not develop the yellow color and rarely ulcerate or form permanent scar tissue. Patients with the generalized type of granuloma annulare are more likely to demonstrate impaired glucose tolerance or overt diabetes.2 The lesions occur predominantly on the upper extremities and manifest as flesh-colored plaques with a raised border. Resolution in the patient with diabetes may occur spontaneously over several years.1 Treatment with photochemotherapy has been used successfully for this condition.11 BULLOSIS DIABETICORUM Tense bullae of the feet and hands are rare but distinctive manifestations of diabetes. The lesions are of uncertain cause, although trauma, microangiopathy, and vasculitis have been implicated.1,3 Differential diagnosis includes bullous impetigo, bullous pemphigoid, and bullous erythema mulitiforme secondary to drugs. The blisters may develop in patients during bed rest, indicating that factors other than trauma are involved. The plantar surfaces and margins of the feet are the most common sites of occurrence (Fig. 153-4). The lesions develop acutely, are asymptomatic, and manifest as tense, 0.5- to 3.0-cm bullae without surrounding

inflammation. Healing without scarring occurs over a period of 2 to 4 weeks. Rupture of the lesions should be avoided to prevent superinfection. Drainage and topical antibiotics may be required for large lesions to prevent secondary infection.1

FIGURE 153-4. Bullosis diabeticorum. The lesion was noted in a 48-year-old black man with a 28-year history of type 1 diabetes, complicated by severe neuropathy and end-stage renal disease. (Courtesy of Dr. Michael Pfeifer.)

ACANTHOSIS NIGRICANS Acanthosis nigricans is an important skin lesion most commonly associated with type 2 diabetes mellitus, which is considered a marker for hyperinsulinemia. There may be a genetic basis for this association.11a Benign acanthosis nigricans is also associated with obesity, lipodystrophic diabetes, and hyperandrogenic states (e.g., polycystic ovary syndrome); it is less frequently associated with acromegaly, Cushing syndrome, other endocrine disorders, and multiple nonfamilial congenital syndromes (e.g., Prader-Willi).12 In addition, hormonal therapy (e.g., the use of methyltestosterone in a hypogonadal patient) has been known to induce acanthosis nigricans. The prevalence of the benign form of acanthosis nigricans in the general population varies with race. One study of adolescents noted incidences of this lesion in 0.5% of whites, 5% of Hispanic Americans, and 13% of blacks.13 This same study noted a 27% incidence in moderately obese (120% of ideal body weight)adolescents and a 66% incidence in the same age group with severe obesity (200% of ideal body weight). The term acanthosis nigricans literally means blackish thorn, which is an apt description of the characteristic velvet-like epidermal thickening associated with hyperpigmentation. The lesions have a predilection for the back and sides of the neck, axillae, and other skinfolds (Fig. 153-5).

FIGURE 153-5. Acanthosis nigricans. Typical lesion on the back of the neck in a young obese black woman with type 2 diabetes.

The pathogenesis of acanthosis nigricans is theorized to result from excess insulin binding to insulin-like growth factor receptors in the epidermis.14 It is important to identify this skin lesion in younger patients who may have hyperinsulinemia and be at high risk to develop type 2 diabetes mellitus. Hyperinsulinemia, even without diabetes, is a known risk factor for cardiovascular disease. Weight loss has been shown to lead to complete regression of the skin lesion,12 presumably secondary to the reduction of hyperinsulinemia. Whether oral agents such as metformin and the thiazolidinediones, which reduce hyperinsulinemia, will reverse this skin lesion is yet to be determined. Patients are occasionally bothered by the acanthosis lesions and may unsuccessfully attempt to scrub them off. The use of keratolytic agents such as salicylic acid has been reported to improve the cosmetic appearance. In addition, the use of omega-3 fatty acid as found in dietary fish oil supplements has also been reported to improve the skin lesions.15 Clinicians should keep in mind that acanthosis nigricans may also be the sign of an occult malignancy, most commonly a gastrointestinal adenocarcinoma. In the elderly, nonobese patient with diabetes, new onset of acanthosis nigricans warrants a thorough evaluation to exclude a malignancy.12 ACROCHORDONS More than 20 years ago, acrochordons (skin tags) were found to be associated with diabetes.16 More recent studies have confirmed this association: It appears that 66% to 75% of all patients with skin tags have diabetes.17 These soft fibromas have a predilection for the eyelids, neck, and axillae and increase in frequency with age (Fig. 153-6). Skin tags may be associated with acanthosis nigricans and may also be related to an excess of insulin-like growth factors. These lesions, although only a cosmetic problem, may be yet another marker for type 2 diabetes mellitus.17

FIGURE 153-6. Acrochordons with acanthosis nigricans. Axillary lesions noted in a middle-aged white man with type 2 diabetes.

NECROLYTIC MIGRATORY ERYTHEMA Necrolytic migratory erythema is a unique skin rash associated with A-cell tumors of the pancreas, that is, glucagonoma. The glucagonoma syndrome includes the

clinical triad of diabetes mellitus, weight loss, and necrolytic migratory erythema. The diagnosis is confirmed by elevated glucagon levels and pancreatic islet cell neoplasm.18 These rare tumors are most frequently found in the tail of the pancreas, which has the highest concentration of glucagon-producing A cells. The rash of necrolytic migratory erythema is distinctive with erythema, vesicles, bullae, pustules, and erosions with a predilection for the face, groin, and other intertriginous areas (Fig. 153-7). This rash may be present for years before the correct diagnosis is made. It is often confused with Candida or misdiagnosed as psoriasis or chronic eczema.19 Clues to the diagnosis include associated glossitis, stomatitis, and angular cheilitis. These findings with a similar rash may also be found in acroder-matitis enteropathica, which is associated with zinc deficiency.

FIGURE 153-7. Necrolytic migratory erythema. A 72-year-old white man with metastatic islet cell tumor and a glucagon level of 11,496 pg/mL. Eruptions were noted on the scrotum, buttocks, and lower extremities. (Courtesy of Dr. George Lawrence.)

The patient with necrolytic migratory erythema usually presents with mild type 2 diabetes, anorexia, weight loss, anemia, and often gastrointestinal ulcerations. The rash is considered to be secondary to excessive catabolism of amino acids. The resultant amino-acid deficiency, which is similar to pellagra, has a toxic effect on keratinocytes. Patients typically present with glucagon levels >1000 pg/mL, although tumors may occur in patients whose glucagon levels are 500 patients who otherwise faced a major amputation. If successful revascularization is accomplished, more conservative reconstructive foot surgery, distal amputations, or revisions can be carried out to achieve healing and limb salvage.17 Gradual and progressive weight-bearing under careful surveillance is mandatory and often requires special orthotics and/or shoes to keep these high-risk areas healed and pressure free.

CONCLUSION The fear of gangrene or amputation is one of the overwhelming concerns of diabetic patients who experience the many complications of their disease. Neuropathy, ischemia, and an altered host defense mechanism make these patients particularly prone to developing foot ulcers, which often become infected. Occasionally, it is

only the complications of odor, hyperglycemia, or systemic symptoms that bring the patient to the hospital with a septic foot. Preventing limb-threatening ulcers or infections begins with patient education and understanding. Early recognition of any foot problem and its prompt treatment are essential. Treating serious limb-threatening conditions requires considerable experience. Diabetics generally have greater risk factors (usually cardiac), and diabetic arteries require the maximum skill and experience of the operating surgeon. A team approach is the most cost-effective method to salvage the diabetic foot.18 In the past decade, the amputation rate at all levels of the diabetic lower extremity has been reduced by utilizing the concepts outlined in this chapter.19 An aggressive approach to limb salvage is less expensive than resorting to major amputation, and the benefits to the patient and society are unquestionably superior.20 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

US Department of Health and Human Services. Healthy People 2000— national health promotion and disease prevention objectives. Washington, DC: US Government Printing Office, 1991:73. LoGerfo FW, Coffman JD. Vascular and microvascular disease in the diabetic foot: implications for foot care. N Engl J Med 1984; 311:1615. Berceli SA, Chan AK, Pomposelli FB Jr. Efficacy of dorsal pedal artery bypass in limb salvage for ischemic heel ulcers. J Vasc Surg 1999; 30:499. King TA, DePalma RG, Rhodes RS. Diabetes mellitus and atherosclerosis involvement of the profunda femoris artery. Surg Gynecol Obstet 1984; 159:553. Gibbons GW, Wheelock FC Jr. Problems in the noninvasive evaluation of the peripheral circulation in the diabetic. Prac Cardiol 1982; 8:115. Gibbons GW, Freeman DV. Diabetic foot infections. In: Howard RJ, Simmons RL, eds. Surgical infectious diseases. Norwalk, CT: Appleton & Lange, 1988:585. Gibbons GW. Diabetic foot sepsis. In: Brewster D, ed. Common problems in vascular surgery. Chicago: Year Book, 1989:412. Delbridge L, Appleberg M, Reeve TS. Factors associated with development of foot lesions in the diabetic. Surgery 1983; 93:78.

8a. Benotmane A, Mohammedi F, Ayed F, et al. Diabetic foot lesions: etiologic and prognostic factors. Diabetes Metab 2000; 26:113. 9. 10. 11. 12.

Gibbons GW, Wheelock FC Jr. Cutaneous ulcers of the diabetic foot. In: Ernest CB, Stanley JC, eds. Current therapy in vascular surgery. Philadelphia: BC Decker, 1987:233. Gibbons GW. Diabetic foot sepsis. Semin Vasc Surg 1992; 5(4):1. Sinha S, Frykberg RG, Kozak GP. Neuroarthropathy in the diabetic foot. In: Kozak G, ed. Clinical diabetes mellitus. Philadelphia: WB Saunders, 1983:415. Gibbons GW, Eliopoulos GM. Infection of the diabetic foot. In: Kozak GP, Hoar CS, Rowbotham JL, et al., eds. Management of the diabetic foot problem. Joslin Clinic and New England Deaconess Hospital. Philadelphia: WB Saunders, 1984:97. 13. Gibbons GW, Freeman DV. Vascular evaluation and treatment of the diabetic. Clin Podiatr Med Surg 1987; 4:337. 13a. Feinglass J, Kaushik S, Handel D, et al. Peripheral bypass surgery and amputation: northern Illinois demographics. Arch Surg 2000; 135:75. 14. Hoar CS, Campbell DR. Aortoiliac reconstruction. In: Kozak GP, Hoar CS, Row-botham JL, et al., eds. Management of the diabetic foot problem. Joslin Clinic and New England Deaconess Hospital. Philadelphia: WB Saunders, 1984:159. 15. Wheelock FC, Gibbons GW. Arterial reconstruction: femoral, popliteal, tibial. In: Kozak GP, Hoar CS, Rowbotham JL, et al., eds. Management of the diabetic foot problem. Joslin Clinic and New England Deaconess Hospital. Philadelphia: WB Saunders, 1984:173. 15a. Akbari CM, Pomposelli FB Jr, Gibbons GW, et al. Lower extremity revascularization in diabetes: late observations. Arch Surg 2000; 135:452. 16. 17. 18. 19. 20.

Pomposelli FB, Jepson SJ, Gibbons GW, et al. A flexible approach to infrapopliteal vein grafts in patients with diabetes mellitus. Arch Surg 1991; 126:724. LoGerfo FW, Gibbons GW, Pomposelli FB, et al. Trends in the care of the diabetic foot. Arch Surg 1992; 127:617. Caputo GM, Cavanagh PR, Ulbrecht JS, et al. Current concepts: assessment and management of foot disease in patients with diabetes. N Engl J Med 1994; 13:854. Gibbons GW, Maracaccio EJ, Burgess AM, et al. Improved quality of diabetic foot care, 1984 vs. 1990: reduced length of stay and costs, insufficient reimbursement. Arch Surg 1993; 1283:576. Gibbons GW, Burgess AM, Guadagnoli E, et al. Return to well-being and function after infrainguinal revascularization. J Vasc Surg 1995;21:35.

CHAPTER 155 DIABETIC ACIDOSIS, HYPEROSMOLAR COMA, AND LACTIC ACIDOSIS Principles and Practice of Endocrinology and Metabolism

CHAPTER 155 DIABETIC ACIDOSIS, HYPEROSMOLAR COMA, AND LACTIC ACIDOSIS K. GEORGE M. M. ALBERTI Diabetic Ketoacidosis Epidemiology Pathophysiology Precipitating Factors Signs and Symptoms Diagnosis Initial Laboratory Investigations Treatment Secondary Phase of Treatment Complications of Therapy Prevention Hyperosmolar Hyperglycemic Nonketotic Coma Pathophysiology Precipitating Factors Presentation Diagnosis Treatment Lactic Acidosis Pathophysiology Diagnosis and Presentation Treatment Alcoholic Ketoacidosis Appendix 1. Guidelines to the Management of Diabetic Ketoacidosis Initial Measures Treatment Appendix 2. Sick-Day Rules for Type 1 Patients Chapter References

The diabetic acidoses and comas remain a significant cause of mortality and morbidity, much of it unnecessary. Many of the problems encountered could be avoided by the education of patients, health care professionals, and physicians in appropriate preventive measures and by the use of systematic, logical therapy. Several reviews are recommended for further reading. 1,2,3,4,5,6,7,8 and 9

DIABETIC KETOACIDOSIS Diabetic ketoacidosis (DKA) may be defined as a state of uncontrolled diabetes mellitus in which there is hyperglycemia (usually >300 mg/dL or 16.7 mmol/L) with a significant lowering of arterial blood pH (5 mmol/L). The cutoff between DKA and hyperosmolar hyperglycemic nonketotic coma (HONK) is somewhat arbitrary, although hyperglycemia tends to be much more severe in the latter, with ketone body levels lower. EPIDEMIOLOGY There are few good data available on the incidence of DKA. In a survey in Rhode Island, DKA accounted for 1.6% of all admissions to the hospital, with previously undiagnosed diabetes accounting for 20% of these. The annual incidence was 14 per 100,000 people.10 In Denmark an annual incidence of 4.5% was reported, with highest risk in female adolescents.11 The best data have been compiled by the National Diabetes Group in the United States, who report an annual incidence of three to eight episodes per 1000 diabetic patients, with 20% to 30% occurring in new diabetics.12 Higher figures have been reported by the Centers for Disease Control and Prevention13 at 26 to 38 per 100,000 population and an annual incidence of 1.0% to 1.5% of all diabetics in the period 1980 to 1987. Rates were threefold higher in black men than in white men. There is increasing recognition of DKA in type 2 diabetic patients, particularly nonwhites such as American Indians,14 Japanese,15 blacks,16 and Asian Indians.17 Mortality for established DKA is relatively high, accounting for 10% of all diabetes-related deaths in the United States between 1970 and 1978.18 Rates were highest in nonwhites and in the older than 65-year-old age group (59% of all DKA deaths). In most published series, mortality lies between 4% and 10%, although 0% has been reported in one large series in the United States.19 A report from Birmingham, United Kingdom, showed a rate of 3.9% in 929 episodes over a 21-year period, with 50% of identified causes in nearly all series. This may occur in established type 1 diabetics or in previously undiagnosed patients, who form 20% to 30% of the total admissions. Infections may be minor, such as urinary tract infection, skin lesions, or bacterial throat infections, or more severe. Infection causes a marked increase in secretion of cortisol and glucagon. Established type 1 diabetics with infections often diminish their food intake and may mistakenly decrease their insulin doses, whereas these should be increased based on the results of home blood glucose monitoring. Urine ketones should also be checked routinely in ill type 1 diabetics with increased insulin doses if ketones are more than trace positive. Other precipitating factors include omission of insulin doses in known type 1 diabetics, a well-known phenomenon in many “brittle” diabetics and in any young type 1 diabetic with recurrent episodes of DKA. In a large study in Scotland, young people admitted in DKA were less likely to have taken their prescribed doses of insulin,26 and in New Zealand, 61% of DKA patients were found to have made errors in insulin self-administration.27 Children with DKA have a greater risk of associated psychopathology28 and are more likely to come from disadvantaged backgrounds.29 Children presenting for the first time in DKA tend to be those with the lowest C-peptide (i.e., insulin) reserves.30 Another cause of “pure” insulin deficiency is the malfunction of insulin infusion devices, with very high occurrence rates in some centers, in particular because of catheter failure, with a rate of 0.14 per patient-year in one large series.31 Other established precipitating factors include cerebrovascular accidents, acute myocardial infarction, and trauma. Each of these is accompanied by increased secretion of catecholamines, glucagon, and cortisol, with predictable metabolic consequences if insulin doses are not increased. Rarer causes include Fourier gangrene (sudden, severe gangrene of the scrotum), infusion of b-sympathomimetic agents, pheochromocytoma, and treatment with the antipsychotic drugs clozapine and olanzapine.32 SIGNS AND SYMPTOMS The classic clinical presentation of severe DKA includes Kussmaul respiration (i.e., deep, sighing hyperventilation), dehydration, hypotension, tachycardia, warm skin, normal or low temperature, and altered state of consciousness. Only ~10% are totally unconscious, and even this figure has diminished recently. Often, the breath smells of acetone, and there may be oliguria in the later stages. Preceding symptoms include nausea and vomiting, thirst, and polyuria in nearly all cases, and, less commonly, leg cramps and abdominal pain. There may also be no bowel sounds, with gastric stasis and pooling of fluid. Occasionally, the patient may present with an acute abdomen (pain and rigidity). If this occurs in young patients, virtually all cases resolve with conservative treatment; in older patients, there may be intraabdominal disease.33 It is sensible to treat the metabolic disturbance first. If there is no resolution or there is worsening of the abdominal state during the first 3 to 4 hours of treatment, diagnostic reevaluation should be considered. Nearly all the signs and symptoms of DKA can be ascribed to different aspects of the metabolic disturbance (Table 155-1) Thus, the dehydration, polyuria, and thirst are secondary to the osmotic diuresis. The hypotension and tachycardia are caused by the fluid loss and the acidemia. Acidemia also causes the hyperventilation. The

inappropriately low body temperature and warm skin are the result of the vasodilatation caused by the acidemia. The vomiting and nausea are probably the consequence of hyperketonemia, and the leg cramps and gastric stasis may be secondary to intracellular potassium depletion. The impaired consciousness correlates only with plasma osmolality, implying that intracellular fluid loss from cerebral cells is involved, although impaired cerebral circulation caused by hypotension may also contribute. Hypovolemia and hypotension cause prerenal failure with consequent oliguria.

TABLE 155-1. Signs and Symptoms of Diabetic Ketoacidosis

DIAGNOSIS In most cases, the rapid diagnosis of DKA should be possible at the bedside (Table 155-2). The clinical history usually is helpful. Clinical examination and bedside measurement of blood glucose and plasma ketones (using a test strip) should complete the diagnosis. Emergency room staff should be instructed in teststrip glucose measurement. The condition of “euglycemic” ketoacidosis should also be noted, in which blood glucose levels are not very elevated despite severe ketoacidosis,34 although if this is defined as a blood glucose level of £ 180 mg/dL, it is rare, occurring in only 1% of cases.35 This has been found in patients who have been on insulin pump therapy, in those who have fasted for long periods before admission, and in pregnancy.36 One study has confirmed that insulin deficiency in the fasted state is associated with a more rapid rise in hydrogen ion concentration but a slower rise in blood glucose than in the fed state, although glucose levels still were not strictly “euglycemic.”37

TABLE 155-2. Bedside Differential Diagnosis of Coma

Ketone bodies in plasma are tested for, using either test strips or tablets. For tablets, plasma should be diluted and a positive result at a dilution of 1 in 8 or above is significant. For the test strips, 1-plus or more is significant. These nitroprusside- based tests respond only to acetoacetate and, to a lesser extent, acetone. If the 3-hydroxybutyrate:acetoacetate ratio is greatly elevated, which occasionally occurs in severe acidemia or anoxia, a falsenegative finding may result. A sensitive whole blood b-hydroxybutyrate test strip has been developed, which should be helpful although it is not yet widely used.38 Using Table 155-2, it should be possible to distinguish clearly between the different diabetic comas. If there is doubt as to the presence of hypoglycemia, 20 mL of 50% glucose can be given intravenously (iv). The other diagnostic difficulty may be the distinction between severe DKA and lactic acidosis, because there is always some excess ketosis in the latter. Usually, blood glucose levels are much more elevated in DKA, but laboratory measurement may be necessary to distinguish between these two acidoses. INITIAL LABORATORY INVESTIGATIONS At the same time as the diagnosis is being made at the bedside, various tests should be ordered (Table 155-3). The most urgent of these is arterial blood pH, because the extent of acidemia will influence treatment. If there is a problem obtaining arterial blood, capillary and venous measurements give remarkably similar results.39 Generally, PCO2 and PO2 are estimated at the same time, the former helping if there is a mixed respiratory and metabolic acid-base disturbance.40 Usually, the PO2 is normal or high but, paradoxically, may be low in a significant number of older patients, presumably owing to pulmonary arteriovenous shunting.

TABLE 155-3. Initial Investigation for the Diagnosis and Management of Diabetic Ketoacidosis

Formal measurement of plasma glucose is required to confirm the diagnosis. Plasma urea, creatinine, and electrolyte levels also are essential. Creatinine gives an indication of dehydration and prerenal failure; plasma urea may indicate the extent of protein catabolism. The measurement of the four major electrolytes allows calculation of the anion gap. This usually is increased (>12), but patients can present with a normal anion gap and a hyperchloremic acidosis.40 There are other causes of an acidosis with an increased anion gap, including lactic acidosis, uremia, and ingestion of agents such as methanol and salicylates; therefore, plasma ketones must be checked as well.6 Plasma sodium levels usually are low despite the total body deficit of hypotonic saline. This is because the hyperglycemia draws fluid from the intracellular space, diluting the plasma. It can be calculated that for every 100-mg/dL elevation in plasma glucose, plasma sodium is decreased by 2.4 mmol/L; a plasma glucose of 750 mg/dL with a plasma sodium of 128 mmol/L implies a “true” sodium of 144 mmol/L.41 Pseudohyponatremia occasionally is found in DKA patients with severe hyperlipidemia.42 The laboratory often informs the clinician of the problem, because the blood sample may have caused blockage of the laboratory autoanalyzers.

Plasma potassium levels can be high, normal, or low. High levels usually are indicative of very acute onset of DKA, with urinary excretion not having kept pace with intracellular losses. They also may be found in the “sick-cell syndrome” or in acute anuria occurring simultaneously with DKA. Normal plasma levels generally are associated with a significant total body potassium deficit. Low plasma potassium levels are an indication of a very large total body deficit. They occur either in an insidious onset of DKA or in patients presenting with DKA who have been taking diuretics without adequate oral replacement. Serum bicarbonate levels are less helpful. In any significant metabolic acidosis, bicarbonate levels are very low, and the pH and PCO2 give more useful information. Chloride levels do not help, except in occasional cases of hyperchloremic acidosis. Plasma osmolality is a useful guide to the severity of the metabolic state. It can be calculated easily using the following equation:

In those centers using Systeme Internationale units, the equation becomes

This gives results that agree closely with measured osmolality, except in occasional situations, such as alcohol overdose. Other tests that should be carried out routinely include sending urine and blood samples for culture and sensitivity and a throat swab, if indicated. Because infection is a common precipitating factor, antibiotics frequently will be used. Samples for culture should be sent before antibiotic therapy begins. Blood for hemoglobin (Hb), hematocrit, and white blood cell count (WBC) tends to be sent routinely as well. These tests are less helpful. The hematocrit indicates hemoconcentration, but so does creatinine. The WBC is singularly unhelpful and can be misleading. There is almost always a leukocytosis in severe DKA, but correlates with blood ketone body levels and not with the presence of infection.43 The two main signs of infection—pyrexia and leukocytosis—are either absent or misleading in DKA. An electrocardiogram should be performed in the emergency room. This may give information on ischemic heart disease or myocardial infarction. More important, it provides a yardstick for subsequent acute changes in plasma potassium levels, which should be followed by electrocardiographic monitoring. The baseline electrocardiogram is necessary because acidemia itself can produce changes, some of which can mimic ischemia, and a pseudoinfarction pattern has been described.44,45 A chest radiograph is often done routinely. This is probably reasonable, in case of subsequent iatrogenic disorders, such as the adult respiratory distress syndrome or infection. TREATMENT There are five main elements of treatment: fluid, insulin, potassium, alkali, and “other” measures. These are discussed in turn (see Appendix 1). FLUID The first priority of treatment is fluid replacement. Insulin therapy is effective only if fluid is given rapidly in the early stages. The total water deficit ranges from 50 to 100 mL/kg body weight, with a sodium deficit of 7 to 10 mmol/kg. Several tailored replacement fluids that were relatively hypotonic were previously recommended. Now, it is agreed that isotonic saline (0.154 mol/L; 0.9%) is the appropriate initial replacement fluid. Concerns have been expressed about overzealous rates of fluid replacement. The author's current practice is to give 1 L in the first hour, then 1 L in 2 hours, and then 1 L every 4 hours until the patient is well hydrated (see Appendix 1). This routine should be varied according to the fluid status of the patient. In patients with cardiovascular disease or in elderly or shocked patients, who are at increased risk for development of heart failure,46 a central venous pressure (CVP) line should be inserted and the rate of fluid administration guided by CVP measurement. The exception to the use of isotonic saline is the presence or development of hypernatremia (measured plasma sodium >150–155 mmol/L). In this case, half-normal (0.45%) saline should be used, but infused more slowly. When saline is used, plasma sodium levels inevitably rise, partly because the infused fluid has a higher sodium content than the extracellular fluid in DKA patients, partly because water without sodium will be moving into cells, and partly because glucose levels will be falling. This rise in sodium levels is helpful, however, in that it prevents plasma osmolality from falling too quickly. This may be beneficial in preventing cerebral edema. Hyperchloremia almost invariably develops late in treatment, but there is little evidence to suggest that it is harmful. Most DKA patients who are hypotensive on presentation respond with a rise in BP to the first 1 to 2 L saline. If systolic BP remains below 90 mm Hg, 1 to 2 U blood or plasma expanders should be given. If this fails, 100 mg hydrocortisone sodium succinate may be given iv, with an appropriate increase in subsequent insulin therapy. Once blood glucose levels have fallen to 250 mg/dL (13.8 mmol/L), 10% glucose is substituted for saline. If this occurs before the patient is adequately rehydrated, saline should be continued simultaneously. The importance of adequate early rehydration cannot be over-emphasized. Simple rehydration alone lowers blood glucose levels by improving the renal excretion of glucose, with hemodilution accounting for as much as a 23% fall in blood glucose. Tissue perfusion is also improved, allowing the small amounts of insulin present to begin to act. Rehydration, even without insulin, decreases counterregulatory hormone secretion. INSULIN Before 1973, large doses of insulin (i.e., hundreds of units) were routinely used in DKA. Then it was shown that relatively small amounts of insulin given intramuscularly (im) or as a continuous iv infusion were just as effective in lowering blood glucose, and had several advantages.46,47 Among the advantages of the low-dose regimens are decreased problems with hypokalemia during therapy, a lower occurrence of late hypoglycemia, and a more predictable response to therapy. There is also an adequate but somewhat slower rate of fall of blood glucose levels, which is less likely to cause osmotic disequilibrium. Insulin Resistance. The major concern in the use of low doses of insulin has been insulin resistance. It has been shown that fractional glucose turnover and the rate of fall of blood glucose levels in DKA patients are decreased up to tenfold compared with nonketotic, well-controlled diabetics. Similarly, in animals and humans, insulin binding by adipocytes is decreased, as is postreceptor insulin action. This is accompanied, in ketoacidotic rats, by a marked diminution in total body insulin responsiveness. These changes correlate with the degree of acidemia, which is most severe when the pH is 150 µU/mL, well above those found in normal persons. Similarly, with standard im regimens, levels of >80 µU/mL are engendered. These levels are sufficient, even in the face of insulin resistance, to inhibit lipolysis, thereby cutting off the supply of substrate for ketogenesis, restraining hepatic gluconeogenesis and helping to decrease glucagon levels. There is little impact on peripheral glucose uptake initially, but there are adequate circulating fuels, and ketone body use steadily increases; the insulin influence on potassium transport into cells is also submaximal, which may be an advantage. Intravenous Insulin Regimens. Several insulin regimens have been proposed, with doses ranging from 2 to 10 U per hour as a continuous infusion. Good results have been reported with 1.4 to 1.6 U per hour after an initial small bolus.19 The author's routine is to give 6 U per hour in saline, using an infusion pump and a separate line (see Appendix 1). In children, a dosage of 0.1 U/kg per hour is used. Because insulin has a circulating half-life of 4 to 5 minutes and a biologic half-life of no more than 30 minutes, it is critical that the insulin be given continuously. If the infusion stops for any reason, the effects will rapidly disappear. Insulin adsorbs to plastic and glass; therefore, some physicians recommend making the insulin solution in polygeline or albumin or drawing back 1 mL of the patient's blood into the syringe. In practice, adsorption is not a problem. Blood glucose levels should be checked after the first 2 hours. If there has not been a significant fall (50–100 mg/dL or 2.8–5.7 mmol/L), then the infusion pump and

line, the rehydration scheme, and the BP should be checked, and the insulin infusion rate should be doubled if these are satisfactory. This should be repeated every 2 hours until blood glucose levels are falling satisfactorily. When blood glucose levels have fallen to 250 µg/dL and 10% dextrose has been substituted for saline (100 mL per hour), the insulin dose should be decreased to 4 U per hour, and subsequently modified according to hourly bed-side blood glucose readings. Intramuscular Regimen. Hourly im insulin provides an alternative to continuous iv insulin and is particularly useful in centers where reliable infusion pumps are not available or where nursing care is inadequate. In this case, a loading dose of insulin should be given as either 20 U im or 10 U im plus 10 U iv in hypotensive or very dehydrated patients. Thereafter, 5 to 6 U should be given hourly as deep im injections. In children, a loading dose of 0.25 U/kg is given, followed by 0.1 U/kg hourly. Adequate rehydration is critical for im insulin to be effective. If, after 2 hours, there is not a significant response, the im regimen should be substituted with continuous iv insulin, having first checked that rehydration is progressing satisfactorily. When blood glucose reaches 250 mg/dL and iv glucose is commenced, the insulin dose should be decreased to 5 to 6 U every 2 hours (see Appendix 1). POTASSIUM More iatrogenic deaths during the treatment of DKA have been caused by changes in plasma potassium than by any other factor. There is usually a large deficit of intracellular total body potassium (3–12 mmol/kg body weight). Despite this, as many as 33% of DKA patients may have elevated plasma K+ levels initially. This loss of intracellular potassium into the extracellular space, which occurs in all cases, has been attributed to the acidemia, intracellular volume depletion, and lack of insulin. The role of acidemia has been questioned, however. Additional factors include a direct effect of hyperglycemia and hyperglucagonemia. In a careful analysis, glucose, pH (negatively), and the anion gap were independent, significant determinants of plasma K+ on presentation.48 Once treatment commences, plasma K+ levels inevitably fall, except in those presenting with the sick-cell syndrome. The fall is the result of intracellular volume repletion, hemodilution, reversal of the acidemia, loss of K+ in the urine as urine flow is reestablished, and a direct effect of insulin on intracellular K+ transport. Thus, hypokalemia is inevitable unless potassium is replaced. There are arguments about when potassium replacement should begin. Some recommend waiting until plasma potassium levels are known, levels are low normal or low, and urine flow has been reestablished; probably, this is too late. The author's practice is to start cautious replacement at 20 mmol KCl per hour in the saline infusion from the time of the first dose of insulin, then to modify the amount infused according to subsequent plasma values (see Appendix 1). It has been suggested that potassium should be given as phosphate or half as phosphate and half as chloride. However, phosphate requirements are very different from those for potassium, so it is probably sensible to replace them separately, if at all, and to use KCl. Electrocardiographic monitoring is an invaluable guide to rapid changes in plasma potassium, and all patients should be monitored at least in the early stages of therapy. For as long as iv therapy is continued, iv potassium replacement should be continued. Thereafter, oral potassium replacement should be continued for several days, because much of the potassium administered iv will be lost in the urine, and the total body deficit will be only partly replenished. If alkali is given, additional potassium should be given (20 mmol/100 mmol sodium bicarbonate). OTHER ELECTROLYTES There is a deficiency of magnesium, calcium, and phosphate, as well as of sodium and potassium in DKA patients. It is arguable, however, whether these need to be replaced immediately. Most debate has concerned phosphate. During treatment of DKA, plasma phosphate levels fall, sometimes to undetectable levels. Red cell 2,3-diphosphoglycerate levels are also very low and take 4 to 48 hours to return to normal. This may cause impaired oxygen delivery to tissues when the acidemia is corrected. It has been argued that the low phosphate levels impede recovery of 2,3-diphosphoglycerate. Several trials of phosphate replacement have been carried out. None of the more recent trials has shown benefit, and in all cases, biochemical hypocalcemia was found in the treated group.49 It is possible that phosphate changes are less with the use of low-dose insulin than they were previously. It is not the author's practice to replace phosphate. Similarly, although magnesium levels are low during therapy, no convincing evidence shows that replacement is beneficial. ALKALI There is still no universal agreement about correcting the acidemia of severe DKA. The acidemia has certain pathophysiologic consequences, including negative inotropism, peripheral vasodilatation, central nervous system depression, and insulin resistance. On the other hand, vigorous alkalinization has deleterious consequences, including hypokalemia, a paradoxical fall in cerebrospinal fluid pH, impaired oxyhemoglobin dissociation, and rebound alkalosis.50 Human data suggest that bicarbonate either has no benefit or, indeed, may slow clearance of ketones and metabolic normalization.51,52 and 53 Despite this, it is usually considered advisable to give moderate amounts of bicarbonate when the pH is 90% infant mortality rate and a 30% maternal mortality rate.1 In the mid-1970s, physicians were still counseling diabetic women to avoid pregnancy.2 This viewpoint was justified because of the poor obstetric history in 30% to 50% of diabetic women. Infant mortality rates finally improved when treatment strategies stressed better control of maternal plasma glucose levels. As the pathophysiology of pregnancy complicated by diabetes has been elucidated and as management programs have achieved and maintained normoglycemia throughout pregnancy, perinatal mortality rates have decreased to levels seen in the general population (Fig. 156-1).3

FIGURE 156-1. The relationship between mean maternal glucose level and infant mortality over the years. (DKA, diabetic ketoacidosis.)

DIABETOGENIC FACTORS OF PREGNANCY PLASMA GLUCOSE DURING PREGNANCY The fetal demise associated with pregnancy complicated by diabetes seems to arise from glucose.4 Elevated maternal plasma glucose levels should always be avoided. To achieve normoglycemia, a clear understanding of “normal” carbohydrate metabolism in pregnancy is paramount.4a Glucose is transported to the fetus by facilitated diffusion, whereas amino acids are actively transported across the placenta. Moreover, alanine is siphoned selectively to the fetus.5 The maternal serum glucose concentration drops below nonpregnancy (between 55 and 65 mg/dL in the fasting state).6 Simultaneously, plasma ketone concentrations are several times higher and free fatty acids are elevated after an overnight fast.7,8 Thus, pregnancy simulates a state of “accelerated starvation” in which alternative fuels are used for maternal metabolism, while glucose is spared for fetal consumption (Fig. 156-2).9

FIGURE 156-2. Schematic representation of nutrient fluxes across the placenta.

DIABETOGENIC HORMONES OF THE PLACENTA The second half of pregnancy is characterized by a further lowering of plasma glucose levels. Although maternal glucose levels remain below levels during nonpregnancy, insulin levels increase markedly, partly because of increasing antiinsulin hormonal activity. The major diabetogenic hormones of the placenta are human somatomammotropin (hCS), previously referred to as human placental lactogen (hPL), estrogen, and progesterone. Also, serum maternal cortisol levels (both bound and free) are increased. At the elevated levels seen during gestation, prolactin has a diabetogenic effect.10 HUMAN CHORIONIC SOMATOMAMMOTROPIN The strongest insulin antagonist of pregnancy is hCS. This placental hormone appears in increasing concentration beginning at 10 weeks of gestation (see Chap. 108).

By 20 weeks of gestation, plasma hCS levels are increased 300-fold, and by term, the turnover rate is ~1000 mg per day.11 The mechanism of action whereby hCS raises plasma glucose levels is unclear but probably originates from its growth hormone–like properties. The hCS also promotes free fatty acid production by stimulating lipolysis, which promotes peripheral resistance to insulin. CORTISOL Most of the marked rise of serum cortisol during pregnancy can be attributed to the increase of cortisol-binding globulin induced by estrogen. However, free cortisol levels are also increased.12 Thus, the rising cortisol levels may unmask diabetes in a predisposed individual. PROGESTERONE When progesterone is administered to normal nonpregnant fasting women, the serum insulin concentration rises and the glucose concentration remains unchanged. In monkeys, progesterone increases both early and total insulin responses to glucose. PROLACTIN The rise in pituitary prolactin early in pregnancy is triggered by the rising estrogen levels. The structure of prolactin is similar to that of growth hormone, and at concentrations reached by the second trimester (>200 ng/mL) prolactin can affect glucose metabolism. Although no studies have examined the role of prolactin alone as an insulin antagonist, indirect evidence exists that suppressing prolactin in gestational diabetic women with large doses of pyridoxine improves glucose tolerance. INSULIN DEGRADATION DURING PREGNANCY Degradation of insulin is increased in pregnancy. This degradation is caused by placental enzymes comparable to liver insulinases. The placenta also has membrane-associated insulin-degrading activity.13 Concomitant with the hormonally induced insulin resistance and increase in insulin degradation, the rate of disposal of insulin slows. The normal pancreas can adapt to these factors by potentiation of insulin secretion. If the pancreas fails to respond adequately to these alterations, or if the clearance of glucose is defective, then gestational diabetes results. EFFECT OF HYPERGLYCEMIA ON THE PREGNANCY AND THE FETUS If the mother has hyperglycemia, the fetus will be exposed to either sustained hyperglycemia or intermittent pulses of hyperglycemia. Both situations prematurely stimulate fetal insulin secretion. The Pedersen hypothesis links maternal hyperglycemia-induced fetal hyperinsulinemia to morbidity of the infant.4 Fetal hyperinsulinemia may cause increased fetal body fat (macrosomia) and, therefore, a difficult delivery, or may cause inhibition of pulmonary maturation of surfactant and, therefore, respiratory distress of the neonate. The fetus may also have decreased serum potassium levels caused by the elevated insulin and glucose levels and may therefore have cardiac arrhythmias. Neonatal hypoglycemia may cause permanent neurologic damage. Hyperglycemia in the mother may lead to maternal complications, such as polyhydramnios, hypertension, urinary tract infections, candidal vaginitis, recurrent spontaneous abortions, and infertility. Thus, a vigorous effort should be made to diagnose diabetes early and to achieve and maintain normoglycemia throughout pregnancy. As mentioned earlier, improved treatment protocols have lowered maternal plasma glucose levels, and the infant mortality rate has dropped (see Fig. 156-1). Treatment protocols should be designed to establish normoglycemia before conception (Table 156-1).

TABLE 156-1. Flow Chart for Management of Diabetes (Type 1) and Pregnancy

GESTATIONAL DIABETES The literal meaning of gestational diabetes mellitus is pregnancy-related diabetes, which usually refers to diabetes that occurs during pregnancy and disappears after the pregnancy is over. This label may be applied to any pregnant woman who is diagnosed with a high blood glucose level during pregnancy, even if she has developed type 1 or type 2 diabetes during pregnancy. The accepted definition is glucose intolerance of variable degree with onset or first recognition during pregnancy.14 Usually, gestational diabetes occurs during the second half of pregnancy in those who are overweight, unfit, or aging, and who have a family history of diabetes. The report from the Fourth International Gestational Diabetes Workshop-Conference15 suggests that women who are at any risk of an elevated blood glucose should be screened for gestational diabetes at some time during their 24th to 28th week of gestation. The definition of any risk is the presence of one or more of the following: patient is older than 25 years; ethnic background is Asian, black, Hispanic, and/or Native American (up to 12% prevalence in these groups)14; patient has a family history of type 2 diabetes; patient has a body mass index of 27 (i.e., even mildly overweight). In programs in which large populations of pregnant women are screened, screening only women who have one or more of these risk factors is most cost effective; nonetheless, 10% of women found to have gestational diabetes have none of the risk factors. Women from a low-risk group still have been reported to have a 2.8% prevalence rate.16 These women had pregnancy outcomes similar to those of other women with gestational diabetes. The authors of this study concluded that the recommendation not to test women from low-risk groups requires further evaluation in different populations before it can be endorsed. An editorial based on this study17 suggested that “until more data are available to support the new American Diabetes Association recommendations, the health of children will best be served by making every effort to determine the presence and degree of glucose intolerance in every pregnancy.” The California Diabetes and Pregnancy Sweet Success Guidelines for care have continued to recommend screening of all pregnant women between 24 and 28 weeks of gestation.18 The diabetogenic stimuli during these weeks are sufficient to manifest diabetes in at least 75% of women with gestational diabetes. If the screening is delayed until the 32nd week of gestation, 100% of women with gestational diabetes will be detected, but by then, after 4 to 6 weeks of sustained hyperglycemia, 75% of the infants may already have developed fetopathy.19 If the suspicion exists that a pregnant woman in her first trimester may have preexisting, undiagnosed diabetes, then the screening should be performed at the first visit. If the testing is negative in the first trimester and at the usual examination between 24 and 28 weeks, the woman may be among the 25% of women with gestational diabetes who develop diabetes in the third trimester. These women tend to be older than 33 years of age and are >120% of ideal body weight.19 The screening test consists of oral administration of 50 g of glucose and a plasma glucose determination at 1 hour, and it may be given regardless of the time of the last meal. If the plasma glucose concentration 1 hour after the oral load is 140 mg/dL (>7.8 mmol/L), a glucose-tolerance test is indicated. In the summary of the Fourth International Gestational Diabetes Mellitus Workshop-Conference,15 the recommendations for the diagnostic oral glucose-tolerance test allow the clinician to choose between a 100-g glucose drink20 and a 75-g glucose drink.21 Table 156-2 describes these tests. Either glucose load should be administered after a minimum of 8 hours of fasting. A fasting plasma glucose blood sample should be drawn no later than 9:00 a.m. In the case of the 100-g load, plasma glucose levels should be obtained 1, 2, and 3 hours after the glucose load. In the case of the 75-g glucose load, testing is really necessary only at the 0-, 1-, and 2-hour time points.

TABLE 156-2. Diagnosis of Gestational Diabetes Mellitus

The glucose level should be determined on venous plasma using the hexokinase method, and not on capillary blood using glucose oxidase–impregnated test strips, which are less accurate for this purpose. Once the diagnosis of gestational diabetes is made, however, these strips become the mainstay of treatment strategies. DIAGNOSTIC STRATEGIES The diagnostic criteria for gestational diabetes are based on the oral glucose-tolerance test (see Table 156-2).20 These criteria correctly identify women at risk for a stillbirth. They do not identify women at risk of delivering a macrosomic infant. Other tests and cutoffs need to be used to identify the macrosomic fetus.22 Blood glucose values at 4 and 5 hours after glucose load have no diagnostic significance. Measurement of glycosylated hemoglobin levels, which is a test of long-term plasma glucose control in type 1 and type 2 diabetes mellitus, is not sensitive enough to diagnose gestational diabetes.19 Use of a single diagnostic test, rather than the two-step method used in most of the United States, greatly simplifies the procedure. A single 75-g glucose load21 is all that is necessary; gestational impaired glucose tolerance is defined as a 2-hour postload glucose concentration of >140 mg/dL (7.8 mmol/L). This constitutes a diagnosis of gestational diabetes and warrants treatment. TREATMENT OF GESTATIONAL DIABETES The goal of management of gestational diabetes is to maintain normoglycemia. Most pregnant women never exceed 120 mg/dL of plasma glucose, even at 1 hour after a meal, despite the ingestion of large quantities of carbohydrate. If the peak postprandial glucose value is >120 mg/dL, the risk of macrosomia rises exponentially.23 Because a nutritious meal for the mother and her unborn child necessitates at least a 40 mg/dL increase of plasma glucose, if the woman's fasting glucose level is much greater than 90 mg/dL (whole-blood capillary glucose), she will be unable to maintain her postprandial levels at 90% have been achieved.7,15

SCREENING FOR METASTATIC DISEASE Malignant islet cell tumors metastasize to peripancreatic lymph nodes and to the liver. Spread to local-regional nodes is not considered a contraindication to attempts at curative surgery.15,36 All of the cross-sectional imaging techniques (ultrasonography, CT, MRI) are effective for evaluation of the liver, but MRI is probably the best of these techniques.20 MRI has superseded angiography, which had been considered the most sensitive technique for the detection of hepatic metastases for these tumors. Newer MRI contrast agents may further enhance the sensitivity of this modality.37,38 In patients with gastrinomas, SRS is clearly the procedure of choice for the detection of liver metastases (see Fig. 159-4).3,20,23,39 ASVS has a high specificity for the detection of liver metastases in these patients, but its sensitivity is low.40 Bone metastases occur in 7% of all patients with gastrinoma and in 31% of gastrinoma patients with liver metastases.41 SRS is the best method for the detection of bone metastases in patients with gastrinomas (see Fig. 159-4).41,42 MRI is recommended for patients with malignant insulinoma, given the poor sensitivity of SRS in these patients and the effectiveness of MRI for the detection of bone metastases in gastrinoma patients.41

RADIOLOGIC APPROACH TO ISLET CELL TUMOR LOCALIZATION Because of the different biology and clinical characteristics of insulinomas and gastrinomas, the radiologic approach to localization differs for the two lesions. The algorithms presented in the following sections are not rigid. The choice of procedures is highly dependent on the skill and experience of the radiologist (and for EUS, the gastroenterologist) who performs them. The patient may best be served by referral to a center where larger numbers of these patients are seen.43 INSULINOMA Given the nearly 100% rate of intraoperative tumor localization using IOUS and surgical palpation,6,14,35 a reasonable question is whether any preoperative localizing procedures are warranted in patients with insulinomas. The surgeons at the Mayo Clinic use conventional abdominal ultrasonography as the only preoperative localizing procedure.6 Some also believe that preoperative localization is unnecessary.43a This minimalist approach to localization requires that radiologists and

surgeons be highly skilled and experienced. Most surgeons prefer a more extensive series of preoperative localization procedures.43,44 and 45 A reasonable first step would be either CT with dynamic spiral technique or MRI. These noninvasive procedures are used to search for a primary tumor as well as to evaluate the liver for possible metastatic disease. If noninvasive imaging is negative, EUS should be performed next, as the first invasive imaging study. If EUS is also negative, many surgeons proceed to arteriography with ASVS.44,45 ASVS may draw attention to tumors that would otherwise be missed by IOUS and palpation.4 Finally, IOUS must be available during surgical exploration. GASTRINOMA Gastrinomas are frequently extrapancreatic and multiple. In addition, duodenal gastrinomas may be extremely small.22,46 Because the sensitivities of all of the imaging studies used for islet cell tumor localization are proportional to tumor size, and because detection of extrapancreatic primary lesions is more difficult than detection of lesions within the pancreas, preoperative localization of gastrinomas is more difficult than localization of insulinomas, and frequently fails. SRS should be the first localization procedure. It is a useful technique for detection of both primary gastrinomas and metastases. Although SRS fails to identify 20% to 33% of gastrinomas,2,20 it is still the most sensitive noninvasive imaging method for these tumors. The addition of SRS information to the findings of other imaging procedures has been shown to change surgical treatment strategy and patient management.3,23 SRS has poor spatial resolution, which makes CT a useful additional noninvasive study. Some surgeons prefer CT as the initial localization study.5,15 Some authors recommend proceeding directly to surgery if no metastases have been demonstrated with these procedures, even if the primary tumor has not been located.2,15 Others suggest that, if previous localization studies were negative, EUS should be performed next.5 ASVS may occasionally be useful if all other imaging studies have been unrevealing.5 ASVS is usually not worthwhile in patients with gastrinomas, however, because it cannot distinguish among pancreatic, duodenal, and peripancreatic nodal disease, any or all of which may be present. As with surgery for insulinomas, IOUS must be available in the operating room and should be part of a thorough surgical exploration of the pancreas and gastrinoma triangle. MULTIPLE ENDOCRINE NEOPLASIA TYPE 1 Gastrinomas and insulinomas may occur sporadically or as part of the multiple endocrine neoplasia type 1 (MEN1) syndrome. The syndrome is a genetic disorder, inherited in an autosomal dominant fashion and characterized by mutations of the MEN1 tumor-suppressor gene on chromosome band 11q13.47 Gastrinomas are the most common functioning islet cell tumors in patients with MEN1, and MEN1 is present in 20% to 38% of all patients with gastrinomas.48 Patients with MEN1 are more likely to have multiple islet cell tumors—gastrinomas, insulinomas, and clinically nonfunctioning tumors (see Fig. 159-3). Gastrinomas in these patients are more likely to be multiple and to be located in the duodenum.48,49 (Also see Chap. 188.) In general, surgical exploration in MEN1 patients with gastrinomas is not recommended.49a MEN1 patients with hyperinsulinism and multiple pancreatic tumors may benefit from ASVS, which can aid in determining which lesion is the insulin source.50 CHAPTER REFERENCES 1. Miller DL. Islet cell tumors of the pancreas: diagnosis and localization. In: Freeny PC, Stevenson GW, eds. Margulis and Burhenne's alimentary tract radiology, 5th ed. St. Louis: CV Mosby, 1994:1167. 2. Alexander HR, Fraker DL, Norton JA, et al. Prospective study of somatostatin receptor scintigraphy and its effect on operative outcome in patients with Zollinger-Ellison syndrome. Ann Surg 1998; 228:228. 3. Lebtahi R, Cadiot G, Sarda L, et al. Clinical impact of somatostatin receptor scintigraphy in the management of patients with neuroendocrine gastroenteropancreatic tumors. J Nucl Med 1997; 38:853. 4. Brown CK, Bartlett DL, Doppman JL, et al. Intraarterial calcium stimulation and intraoperative ultrasonography in the localization and resection of insulinomas. Surgery 1997; 122:1189. 5. Prinz RA. Localization of gastrinomas. Int J Pancreatol 1996; 19:79. 6. Grant CS. Insulinoma. Baillières Clin Gastroenterol 1996; 10:645. 7. Jensen RT. Gastrinproducing tumors. Cancer Treat Res 1997; 89:293. 8. Passaro E Jr, Howard TJ, Sawicki MP, et al. The origin of sporadic gastrinomas within the gastrinoma triangle: a theory. Arch Surg 1998; 133:13. 9. Wu PC, Alexander HR, Bartlett DL, et al. A prospective analysis of the frequency, location, and curability of ectopic (nonpancreaticoduodenal, non-nodal) gastrinoma. Surgery 1997; 122:1176. 10. Gibril F, Curtis LT, Termanini B, et al. Primary cardiac gastrinoma causing Zollinger-Ellison syndrome. Gastroenterology 1997; 112:567. 11. Zimmer T, Stölzel U, Bäder M, et al. Endoscopic ultrasonography and somatostatin receptor scintigraphy in the preoperative localisation of insulinomas and gastrinomas. Gut 1996; 39:562. 12. Angeli E, Vanzulli A, Castrucci M, et al. Value of abdominal sonography and MR imaging at 0.5 T in preoperative detection of pancreatic insulinoma: a comparison with dynamic CT and angiography. Abdom Imaging 1997; 22:295. 13. Soga J, Yakuwa Y. The gastrinoma/Zollinger-Ellison syndrome: statistical evaluation of a Japanese series of 359 cases. J Hepatobiliary Pancreat Surg 1998; 5:77. 14. Kuzin NM, Egorov AV, Kondrashin SA, et al. Preoperative and intraoperative topographic diagnosis of insulinomas. World J Surg 1998; 22:593. 15. Kisker O, Bastian D, Bartsch D, et al. Localization, malignant potential, and surgical management of gastrinomas. World J Surg 1998; 22:651. 16. Hayashi Y, Nakazawa S, Kimoto E, et al. Clinicopathologic analysis of endoscopic ultrasonograms in pancreatic mass lesions. Endoscopy 1989; 21:121. 17. Chung MJ, Choi BI, Han JK, et al. Functioning islet cell tumor of the pancreas: localization with dynamic spiral CT. Acta Radiol 1997; 38:135. 18. King AD, Ko GTC, Yeung VTF, et al. Dual phase spiral CT in the detection of small insulinomas of the pancreas. Br J Radiol 1998; 71:20. 19. Mori M, Fukuda T, Nagayoshi K, et al. Insulinoma: correlation of short-TI inversion-recovery (STIR) imaging and histopathologic findings. Abdom Imaging 1996; 21:337. 20. Gibril F, Reynolds JC, Doppman JL, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas. A prospective study. Ann Intern Med 1996; 125:26. 20a.Thoeni RF, Mueller-Lisse UG, Chen R, et al. Detection of small, functional islet cell tumors in the pancreas: selection of MR imaging sequences for optimal sensitivity. Radiology 2000; 214:483. 20b.Semelka RC, Custodio CM, Cem Balci N, Woosley JT. Neuroendocrine tumors of the pancreas: spectrum of appearances on MRI. J Magn Reson Imaging 2000; 11:141. 21. 22. 23. 24. 25. 26.

Tang C, Biemond I, Verspaget HW, et al. Expression of somatostatin receptors in human pancreatic tumor. Pancreas 1998; 17:80. Cadiot G, Lebtahi R, Sarda L, et al. Preoperative detection of duodenal gastrinomas and peripancreatic lymph nodes by somatostatin receptor scintigraphy. Gastroenterology 1996; 111:845. Termanini B, Gibril F, Reynolds JC, et al. Value of somatostatin receptor scintigraphy: a prospective study in gastrinoma of its effect on clinical management. Gastroenterology 1997; 112:335. Lamberts SWJ, Krenning EP, Reubi J-C. The role of somatostatin and its analogs in the diagnosis and treatment of tumors. Endocr Rev 1991; 12:450. Van Eyck CHJ, Bruining HA, Reubi J-C, et al. Use of isotope-labeled somatostatin analogs for visualization of islet cell tumors. World J Surg 1993; 17:444. Kisker O, Bartsch D, Weinel RJ, et al. The value of somatostatin-receptor scintigraphy in newly diagnosed endocrine gastroenteropancreatic tumors. J Am Coll Surgeons 1997; 184:487.

26a.Ardengh JC, Rosenbaum P, Ganc AJ, et al. Role of EUS in the preoperative localization of insulinomas compared with spiral CT. Gastrointest Endosc 2000; 51:552. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Rüsch T, Lightdale CJ, Botet JF, et al. Localization of pancreatic endocrine tumors by endoscopic ultrasonography. N Engl J Med 1992; 326:1721. Schumacher B, Lübke HJ, Frieling T, et al. Prospective study on the detection of insulinomas by endoscopic ultrasonography. Endoscopy 1996; 28:273. Doppman JL, Miller DL, Chang R, et al. Insulinomas: localization with selective intraarterial injection of calcium. Radiology 1991; 178:237. Doppman JL, Miller DL, Chang R, et al. Gastrinomas: localization by means of selective intraarterial injection of secretin. Radiology 1990; 174:25. O'Shea D, Rohrer-Theurs AW, Lynn JA, et al. Localization of insulinomas by selective intraarterial calcium injection. J Clin Endocrinol Metab 1996; 81:1623. Pereira PL, Roche AJ, Maier GW, et al. Insulinoma and islet cell hyperplasia: value of the calcium intraarterial stimulation test when findings of other preoperative studies are negative. Radiology 1998; 206:703. Defreyne L, König K, Lerch MM, et al. Modified intra-arterial calcium stimulation with venous sampling test for preoperative localization of insulinomas. Abdom Imaging 1998; 23:322. Correnti S, Liverani A, Antonini G, et al. Intraoperative ultrasonography for pancreatic insulinomas. Hepatogastroenterology 1996; 43:207. Huai J-C, Zhang W, Niu H-O, et al. Localization and surgical treatment of pancreatic insulinomas guided by intraoperative ultrasound. Am J Surg 1998; 175:18. Thompson NW. Current concepts in the surgical management of multiple endocrine neoplasia type 1 pancreatic-duodenal disease. Results in the treatment of 40 patients with Zollinger-Ellison syndrome, hypoglycaemia or both. J Intern Med 1998; 243:495. Vandevenne JE, Deckers F, Mana F, et al. Insulinoma associated with liver lesions: value of MR imaging. Am J Gastroenterol 1998; 93:1559. Wang C, Ahlström H, Eriksson B, et al. Uptake of mangafodipir trisodium in liver metastases from endocrine tumors. J Magn Reson Imaging 1998; 8:682. Termanini B, Gibril F, Doppman JL, et al. Distinguishing small hepatic hemangiomas from vascular liver metastases in gastrinoma: use of a somatostatin-receptor scintigraphic agent. Radiology 1997; 202:151. Gibril F, Doppman JL, Chang R, et al. Metastatic gastrinomas: localization with selective arterial injection of secretin. Radiology 1996; 198:77. Gibril F, Doppman JL, Reynolds JC, et al. Bone metastases in patients with gastrinomas: a prospective study of bone scanning, somatostatin receptor scanning, and magnetic resonance image in their detection, frequency, location, and effect of their detection on management. J Clin Oncol 1998; 16:1040. Cadiot G, Bonnaud G, Lebtahi R, et al. Usefulness of somatostatin receptor scintigraphy in the management of patients with Zollinger-Ellison syndrome. Gut 1997; 41:107. Lo C-Y, Lam KYL, Kung AWC, et al. Pancreatic insulinomas: a 15-year experience. Arch Surg 1997; 132:926.

43a.Hashimoto, Walsh RM. Preoperative localization of insulinomas is not necessary. J Am Coll Surg 1999; 189:368. 44. 45. 46. 47.

Bliss RD, Carter PB, Lennard TWJ. Insulinoma: a review of current management. Surg Oncol 1997; 6:49. Wiersema MJ. Insulinoma: finding the needle in a haystack. Endoscopy 1996; 28:310. Schröder W, Hölscher AH, Beckurts T, et al. Duodenal microgastrinomas associated with Zollinger-Ellison syndrome. Hepatogastroenterology 1996; 43:1465. Zhuang Z, Vortmeyer AO, Pack S, et al. Somatic mutations of the MEN1 tumor suppressor gene in sporadic gastrinomas and insulinomas. Cancer Res 1997; 57:4682.

48. Jensen RT. Management of the Zollinger-Ellison syndrome in patients with multiple endocrine neoplasia type 1. J Intern Med 1998; 243:477. 49. Mignon M, Cadiot G. Diagnostic and therapeutic criteria in patients with Zollinger-Ellison syndrome and multiple endocrine neoplasia type 1. J Intern Med 1998; 243:489. 49a.Norton JA, Fraker DL, Alexander HR, et al. Surgery to cure the Zollinger-Ellison syndrome. N Engl J Med 1999; 341:635. 50. Doppman JL. Problems in endocrinologic imaging. Endocrinol Metab Clin North Am 1998; 26:973.

CHAPTER 160 SURGERY OF THE ENDOCRINE PANCREAS Principles and Practice of Endocrinology and Metabolism

CHAPTER 160 SURGERY OF THE ENDOCRINE PANCREAS JON C. WHITE Functional Islet Cell Tumors Insulinoma Gastrinoma Glucagonoma Vipoma Somatostatinoma Miscellaneous Functional Islet Cell Tumors Pancreatic Polypeptide Parathyroid Hormone–Related Protein Growth Hormone–Releasing Hormone Adrenocorticotropic Hormone and Corticotropin-Releasing Hormone Nonfunctional Islet Cell Tumors Surgical Aspects of Pancreatic Endocrine Tumors Exocrine Tumors Versus Endocrine Tumors Surgical Procedures Intraoperative Localization Chapter References

Surgery for tumors of the pancreas often is considered futile because of the predominance of exocrine tumors of this organ, which are usually malignant, often unresectable, and rarely curable.1 Indeed, pancreatic cancer is the fifth leading cause of cancer death in the United States,2 with a 2-year survival rate of 8%. 3 Five percent to 15% of all pancreatic neoplasms are nonductal, and 50% of these tumors are of endocrine origin.4 Although approximately one-half of pancreatic endocrine tumors are reported to be benign,5 this figure varies across series because there are no histologic features that distinguish benign from malignant lesions.6 Quantitative DNA analysis has demonstrated that the DNA content of endocrine tumors has an influence on long-term survival, with hyper-triploid tumors having a significantly worse prognosis than diploid, monoploid, or triploid tumors. Nevertheless, neither molecular nor cellular techniques can draw a distinct difference between benign and malignant neoplasms.7 Malignancy may be reliably determined only by the presence of distant metastases or local tumor invasion.8 When these tumors are determined to be benign, they can be resected and cured. Even malignant tumors may be resected, resulting in significant palliation or even cure. The first successful resection of an insulinoma took place in 1929, although the insulinoma syndrome was not described by Whipple until 1935. Since then, many different islet cell tumors and syndromes have been recognized, including gastrinomas, vasoactive intestinal polypeptide tumors (VIPomas), glucagonomas, somatostatinomas, and other lesions that produce a variety of humoral products, such as pancreatic polypeptide (PP), adrenocorticotropic hormone (ACTH), parathyroid hormone–related protein, and serotonin. Clinically, these tumors are differentiated by the syndromes they induce or by the biochemical profiles they produce. The surgical approach to each of these pancreatic endocrine neoplasms varies according to the end organs affected by its secreted products, its malignant potential, its solitary or multiple nature, and the ease with which it can be localized (Table 160-1 and Table 160-2). The surgical procedures used to cure or palliate pancreatic endocrine tumors are more varied than are those used for pancreatic exocrine tumors.

TABLE 160-1. Characteristics of Functional Pancreatic Endocrine Tumors

TABLE 160-2. Surgical Aspects of Pancreatic Endocrine Tumors

As many as 19 different gastroenteropancreatic neuroendocrine cells have been identified that elaborate up to 40 different humoral products.9 Some of these humoral products are associated with clinical endocrine manifestations (e.g., insulin, gastrin, glucagon, somatostatin, ACTH, corticotropin-releasing hormone [CRH], parathyroid hormone–related protein [PTHrP], growth hormone–releasing hormone [GHRH]). Others are not associated with any known syndromes but may have an important role as a tumor marker for diagnosing or following the clinical progress of the neoplasm (e.g., PP, neurotensin, calcitonin, a human chorionic gonadotropin10,11,12 and 13). The tumors associated with the most clinically relevant hormones are considered in the next section.

FUNCTIONAL ISLET CELL TUMORS INSULINOMA Insulinomas are the most common functional islet cell tumors, with an incidence of one case per million population per year.14 Most are solitary and benign (75%), but a few are malignant (10% to 15%) or multifocal (10% to 15%).15 These tumors may be extremely small or they may present as hyperplasia. Because of variability in their size and appearance, they can be difficult to localize. Despite the fact that most of these lesions are benign, the hyperinsulinemia they cause can be dangerous, so that complete resection or maximal cytoreduction is desirable and often necessary (see Chap. 158). Localization of insulinomas, which can be accomplished with computed tomography (CT), magnetic resonance imaging (MRI), mesenteric angiography, or portal venous sampling, should be attempted before surgery. CT and angiography usually are performed first and are most useful in detecting larger lesions (see Chap.

159).16 Although these techniques are relatively specific, they are not very sensitive. Their accuracy generally is reported to be ~80%. If a tumor cannot be identified with CT or angiography, either portal venous sampling or injection of the mesenteric artery with calcium to stimulate insulin secretion may allow localization of the neoplasm to a particular region. Portal venous sampling is the most sensitive preoperative localization technique, with a reported sensitivity of 70% to 100%.17 Radiolabeled octreotide18,19 and provocative angiographic studies using calcium stimulation20 also have been used to improve the accuracy of preoperative localization. Both endoscopic ultrasonography (EUS) and intraductal ultrasonography (IDUS)21 have been used to localize tumors as small as 5 mm. Some surgeons, however, believe that extensive preoperative localization is neither indicated nor cost effective. They contend that intraoperative localization by ultrasonography, endoscopic transillumination, and palpation are more sensitive than preoperative techniques.22,23 When all attempts to localize a tumor at surgery are unsuccessful, some surgeons perform a blind distal pancreatectomy in the hope of excising an occult tumor. If the hyperinsulinemia is responsive to diazoxide, another approach is to perform a biopsy of the pancreas to rule out hyperplasia and then to institute medical therapy.24 On occasion, patients can be maintained on diazoxide until their tumors grow to a detectable size and can be localized and resected. If diazoxide therapy is ineffective before surgery, an 80% pancreatectomy may be required to control the hyperinsulinemia.24 If the insulinoma has been localized and is small and benign, enucleation or local excision is the preferred approach. If the tumor is large or malignant, a formal pancreatic resection, including pancreaticoduodenectomy,25 or tumor debulking may be needed to resect the tumor completely and control the hyperinsulinemia. Although enucleation is appealing because of the greater conservation of pancreatic tissue, the incidence of pancreatic fistula is higher than after pancreatectomy.26 The results of surgery with all procedures tend to be excellent, with 70% to 90% of patients experiencing relief of symptoms or complete cure. During surgical intervention, a thorough exploration of the abdominal cavity should be performed to look for metastases, with focus on the liver and regional lymph nodes. Any extrapancreatic spread of tumor indicates malignancy, and the lesions should be removed at the time of the pancreatic resection. When the surgical localization or extirpation of an insulinoma is not possible, streptozocin, especially in combination with 5-fluorouracil, can be administered to destroy the B cells.15 Evidence also exists that the somatostatin analog octreotide may effectively suppress insulin release and reduce the morbidity and mortality of the persistent hyperinsulinemia (see Chap. 169).27 Although this drug has been less effective for treating insulinomas than for treating other pancreatic tumors, a therapeutic trial is warranted.28 Most cases of insulinoma are nonfamilial and sporadic, but 10% are associated with the multiple endocrine neoplasia type 1 (MEN1) syndrome (see Chap. 188). Patients with insulinoma and MEN1 often have multiple pancreatic tumors. Consequently, surgical extirpation is not as effective in these patients as it is in patients with sporadic tumors. The principal objective of surgery in patients with hyperinsulinemia and MEN1 is to resect the primary tumor, which is likely to be secreting most of the insulin. When a single tumor has been identified and enucleated, a blind distal pancreatectomy is recommended to remove occult tumors, which are inevitably present.29 Although hyper-insulinemia can be controlled using this approach, these patients must be observed carefully for recurrent disease. GASTRINOMA Gastrinomas are a type of ectopic islet cell tumor30 and are the most common malignant functional endocrine tumors of the pancreas (see Chap. 220). Although 60% of gastrinomas are malignant, long-term survival can be achieved by resection or by appropriate treatment of the end-organ symptomatology. Islet cell tumors are present in 65% to 80% of patients with the MEN1 syndrome, and gastrinoma is the most common type.31 When gastrinomas occur in MEN1, the pancreatic tumors always are multiple and are not amenable to surgical resection. No case of MEN1 has ever been reported in which the local resection of a tumor has normalized serum gastrin levels, although aggressive subtotal resection has been reported to provide symptomatic relief.32 Gastrinoma is most noted for its endocrine effects. In 1955, Zollinger and Ellison33 described two patients with ulcers in the jejunum and outlined a syndrome in which pancreatic tumors were associated with aggressive peptic ulcer disease. Because Zollinger-Ellison syndrome is being recognized and diagnosed earlier, the associated ulcers often are less virulent and tend to be routine ulcers of the duodenum. The presence of gastrinoma is suggested by a serum gastrin level of >500 pg/mL and is confirmed with a secretin test. Both gastrinomas and insulinomas differ from other pancreatic endocrine tumors in that they often are small and difficult to localize. Insulinomas are distributed equally throughout the pancreas, but most gastrinomas are found in the gastrinoma triangle, which is bordered by the cystic and common bile ducts superiorly, the junction of the second and third portions of the duodenum inferiorly, and the junction of the neck and body of the pancreas medially.34 Most intraabdominal gastrinomas also are found on the right side of the abdomen.35 A smaller number are found to the left of the superior mesenteric artery. Tumors in different locations do not exhibit the same biologic behavior, suggesting that they may have different causes.36 CT, MRI, sonography, arteriography, and percutaneous transhepatic venous sampling are used for preoperative localization of gastrinoma in the same manner as for insulinoma. Secretin is the agent used to stimulate gastrin secretion during provocative angiography.37 In vivo imaging with a gamma camera after injection of radiolabeled octreotide37a has been reported to localize nearly 100% of gastrinomas.38 Although most gastrinomas can be localized before surgery using these techniques, intraoperative exploration may be the most direct and sensitive means of detecting occult tumors.5 These neoplasms are being found increasingly often in the duodenum, and occasionally can be located by preoperative duodenal endoscopy, intraoperative endoscopy with transillumination, or intraoperative duodenotomy with exploration.39 Duodenal tumors are located most often in the proximal duodenum just distal to the pylorus. Small duodenal tumors also can be located with the help of intraarterial methylene blue injection during surgical exploration. This is usually done through the same catheter used for preoperative angiographic secretin stimulation.40 Despite the variety of preoperative and intraoperative techniques available for localizing these tumors, many are never located and successfully resected. Often, metastases are discovered in lymph nodes but the primary tumor cannot be found. Zollinger and Ellison performed total gastrectomies on their initial two patients to treat their fulminant ulcer disease. Now the view is that tumor localization should be attempted in all patients. Patients other than those with widespread liver metastases should undergo surgical exploration and extirpation of the tumor. If preoperative localization is successful and resection is feasible, as much of the gastrinoma should be removed as possible.41,42 This can involve enucleation or a formal resection, which achieves cure rates approaching 35%.43 If a curative resection is not possible, tumor debulking may offer significant palliation. Pancreaticoduodenectomy is seldom necessary14 but may be used for both duodenal and pancreatic tumors if they are large or have multiple foci.44,45 and 46 If the tumor cannot be localized or resected, symptoms may be relieved by treating the ulcer disease medically with H2-receptor blockers or H+-K+ proton pump inhibitors. In fact, the introduction of proton pump inhibitors has almost eliminated the need for gastrectomy in these patients.45 Vagotomy has been used to reduce medication requirements, with mixed results.47 These procedures have played a smaller role since effective medications to inhibit gastric acid secretion have become available. Patients with malignant tumors treated either medically or with ulcer operations ultimately succumb to metastatic disease. The complete resection of all grossly evident tumor seems to provide the greatest long-term survival48 and should be the primary surgical goal in all patients without metastatic disease or the MEN1 syndrome. Surgical exploration not only provides the only chance for cure but also is helpful in identifying prognostic features associated with long-term survival, such as small tumor size, an extrapan-creatic primary tumor, and the absence of tumor metastases.49 Patients with MEN1 and gastrinomas represent a special case, because their tumors are multifocal, difficult to localize, and may have less malignant behavior than sporadic tumors.50 Some surgeons advocate an aggressive approach, similar to that pursued with other gastrinomas, whereas others recommend that patients with MEN1 not undergo exploration at all. Formerly, the removal of metastatic lymph nodes in addition to gastrectomy was believed to result in regression of the primary tumor in some patients with gastrinoma. It now seems more likely that most of these apparent “regressions” resulted from the serendipitous removal of small duodenal tumors during gastrectomy,51 and little support exists for the concept of gastrinoma regression. All patients should receive long-term postoperative follow-up, because recurrent disease can present many years after apparently curative resection. Calcium or secretin provocative tests may be helpful for detecting recurrent disease early or for predicting whether a patient will remain free of symptoms.52 Even after apparently curative resection, some patients with gastrinoma retain a slightly hypersecretory state and require low doses of antisecretory drugs.53 When resection of a gastrinoma is not possible or the tumor cannot be found at surgical exploration, streptozocin and 5-fluorouracil can be used for palliation, as they are in insulinoma. These chemotherapeutics have no positive effect on survival. H2 receptors or H+-K+ proton pump blockers can be used to control the ulcer disease. Octreotide may reduce symptoms and occasionally controls tumor growth.54 Lipidol-transcatheter arterial embolization has been used for unresectable hepatic metastases.54a GLUCAGONOMA

Although glucagonomas are the third most frequently encountered functional tumor, only 400 cases have been reported in the world literature.55 The first glucagon-producing tumor associated with a cutaneous rash was reported in 1942,56 but the “glucagonoma syndrome” was not fully described until 1966.57 Although 60% of these tumors are malignant, most are solitary; hence, surgical resection can be curative.58 The most characteristic and debilitating clinical finding in patients with glucagonoma is a necrolytic migratory rash. This skin lesion results from an amino-acid deficiency59 and, when seen in association with diabetes and weight loss, should suggest the diagnosis. Elevated serum immunoreactive glucagon levels should instigate a search for the tumor (see Chap. 220). These tumors, which tend to be large and located in the body and tail of the pancreas, are usually easy to detect. If a glucagonoma cannot be seen on routine imaging studies, mesenteric arteriography or selective venous sampling with radioimmunoassay for glucagon may be required.60 If the diagnosis is made early and the tumor is localized to the body or tail of the pancreas, a distal pancreatectomy can be curative. This operation, which does not require concomitant duodenectomy with reanastomosis of the pancreatic duct, common bile duct, and gastrointestinal tract, has minimal morbidity. Tumors in the head of the pancreas require pancreatico-duodenectomy for complete excision. Aggressive cytoreduction of large, unresectable tumors is indicated before the use of chemotherapy. When resection is complete (no metastases), 10-year survival is better than 60%.29 Chemotherapy for glucagonoma has been disappointing. Streptozocin has been used in combination with 5-fluorouracil and octreotide to alleviate symptoms in patients with unresectable disease,29a but has met with limited success. Because the tumor grows slowly, patients who are not candidates for curative resection still can enjoy prolonged symptom-free survival.58 VIPOMA VIPomas are extremely rare tumors that elaborate vasoactive intestinal peptide (VIP), producing a syndrome characterized by watery diarrhea, hypokalemia, and achlorhydria (WDHA syndrome) or pancreatic cholera (see Chap. 220). VIPomas typically secrete a range of peptides, of which VIP is the most clinically relevant. Diagnosing these neoplasms on the basis of basal serum VIP levels can be difficult,61 but the use of pentagastrin as a provocative agent has been helpful.11 Similar to glucagonomas, VIPomas usually are large and easy to locate. Although 80% are located in the pancreas, others can be found in the retroperitoneum, the adrenal gland, or the lung,62 making preoperative localization helpful. These tumors are located predominantly in the body and tail of the pancreas. CT and angiography are the primary radiographic techniques used when localization is required. VIPomas have also been localized by technetium-99m sestamibi scanning.63 Because the VIPoma syndrome can be caused by a malignant tumor (50%) or by hyperplasia (20%), the surgical approach varies widely, depending on the results of the preoperative evaluation. As with all tumors of the endocrine pancreas, complete resection is desirable, but aggressive cytoreduction helps to control the clinical syndrome when total extirpation is not possible. Tumors located in the body and tail of the pancreas usually can be resected by distal pancreatectomy. When tumors are not found in the pancreas, an extensive exploration may be necessary. When the lesions still cannot be located after careful exploration of the abdomen, including the retroperitoneum, some surgeons advocate distal pancreatectomy. Patients whose disease is not controlled with distal pancreatectomy may require total pancreatectomy to alleviate their symptoms.62 Resolution of symptoms has been reported even with resection of recurrent metastatic disease.64 When surgery fails to control the WDHA syndrome, octreotide administration has been shown to inhibit the release of VIP from human VIPoma cells and, rarely, to shrink metastases.65 SOMATOSTATINOMA Somatostatinomas are among the more recently described pancreatic endocrine tumors. They are usually found in the pancreas or duodenum, with the pancreatic tumors carrying a worse prognosis.66 Most pancreatic or duodenal tumors reported have been solitary, malignant, and virulent.62,67,68 and 69 Overproduction of somatostatin, an inhibitory hormone acting on the endocrine system and digestive organs, leads to vague clinical symptoms (see Chap. 220). The true incidence of somatostatinoma may not be known, because the symptom complex of diabetes, cholelithiasis, and steatorrhea is nonspecific and usually not very severe.70 Somatostatin inhibits the release of glucagon as well as insulin, so the diabetes produced is mild and often is not insulin-dependent. A finding of increased serum levels of somatostatin and depressed plasma concentrations of both immunoreactive insulin and glucagon can be used to establish the diagnosis of somatostatinoma.30 These tumors have been discovered in the head and tail of the pancreas and in the duodenum.71 Most somatostatinomas located in the duodenum cause only symptoms related to local disease, such as obstructive jaundice.72 Because humoral manifestations of the somatostatinoma syndrome are vague, these tumors often present with local symptoms. Most are detected as incidental findings on surgical exploration or abdominal radiography, so the need to localize occult somatostatinomas is unusual. When localization is required, CT and arteriography are useful. When the results of hormonal studies raise suspicion for somatostatinoma, endoscopy or contrast radiography are helpful in localizing the tumor because of the frequent occurrence of duodenal neoplasms. Somatostatin analog (octreotide) scintigraphy has been used to localize lesions, and, subsequently, the drug has been used to treat symptoms of patients with somatostatinoma.73 As with other functional pancreatic endocrine tumors, surgical excision of somatostatinomas should be attempted when possible. Somatostatinomas in the head of the pancreas usually require pancreaticoduodenectomy. In addition, distal pancreatectomy and enucleation have been reported for treatment of pancreatic tumors.30 Masses in the periampullary region of the duodenum cause jaundice and may require pancreaticoduodenectomy, whereas other duodenal tumors can be excised by segmental duodenal resection. When metastatic disease is encountered, an aggressive approach to the primary disease as well as tumor debulking is warranted. MISCELLANEOUS FUNCTIONAL ISLET CELL TUMORS PANCREATIC POLYPEPTIDE Pancreatic polypeptide (PP) is a pancreatic hormone whose function is not completely understood. Because it is secreted by many pancreatic endocrine tumors, >50% of the cells of a neoplasm must secrete PP for it to be considered a PPoma.9 A fasting PP level of >300 pmol/mL is the best diagnostic parameter because clinical symptomatology is not well established.74 These tumors are usually large, making wide resection necessary.75 In the case of an unresectable tumor, debulking is appropriate. PARATHYROID HORMONE–RELATED PROTEIN PTHrP-producing tumors of the pancreas are of particular interest because this hormone can cause hypercalcemia (see Chap. 52 and Chap. 219). Although not all PTHrP-secreting tumors produce elevated levels of calcium,76 some are associated with life-threatening hypercalcemia.77 These lesions can be mistaken for a MEN1 syndrome with associated parathyroid hyperplasia, but the two entities can be differentiated by measuring PTH and PTHrP levels.78 Tumor resection results in resolution of the hypercalcemia. GROWTH HORMONE–RELEASING HORMONE GHRH, which usually is secreted by the hypothalamus, can be produced and secreted by a pancreatic islet cell tumor. The resulting acromegaly can be treated by resection of the pancreatic tumor, which corrects the biochemical abnormalities (see Chap. 219).79 Recurrent tumor causes a relapse of the endocrine disorder, demonstrating that the optimal therapy is complete extirpation of the neoplasm. ADRENOCORTICOTROPIC HORMONE AND CORTICOTROPIN-RELEASING HORMONE ACTH or CRH can be secreted by pancreatic tumors.80,81 CRH is commonly produced by pancreatic endocrine tumors but is only rarely associated with elevated levels of ACTH or manifestations of Cushing syndrome.81 When large pancreatic tumors secrete ACTH, however, they can cause severe hypercortisolism, with asthenia, muscle weakness, hypertension, hypokalemic alkalosis, and carbohydrate intolerance. When tumors producing ACTH or CRH can be located, they should be resected (see Chap. 75 and Chap. 219). When Cushing syndrome is produced, bilateral adrenalectomy may be helpful.82

NONFUNCTIONAL ISLET CELL TUMORS Once thought to be rare, nonfunctioning islet cell tumors account for ~50% the endocrine tumors of the pancreas reported in some series.5,83 This increased incidence is due to the widespread use of abdominal CT and MRI, which can detect small and previously occult lesions. Although some of these masses may represent functional tumors that are discovered before they become clinically significant, many nonfunctional tumors grow large, calcific, cystic, or necrotic and do not produce any recognized hormonal syndromes.84 They are often said to present at a later stage when resectability is less likely, although this concept has been challenged.85 These

clinically silent tumors usually demonstrate neurosecretory granules that contain immunoreactive peptides on immunohistochemical staining.86 This suggests that these nonfunctional tumors secrete peptide products that are not bioactive or are secreted in amounts too small to be clinically relevant. The presence in the serum of clinically silent secretory products can be useful to detect recurrence of these nonfunctional tumors after resection. PP, neurotensin, calcitonin, and a human chorionic gonadotropin can be used as markers for some nonfunctional pancreatic endocrine tumors.10,11,12 and 13 Because nonfunctional tumors do not induce any characteristic clinical syndromes, their presentation is similar to that of adenocarcinoma of the pancreas. Nonfunctional islet cell and ductal tumors usually cause symptoms from their mass effects, which lead to biliary and gastrointestinal obstruction, back pain, and weight loss. Because small tumors cause no symptoms, they are not discovered except as an incidental finding during abdominal exploration or imaging. Localization of occult nonfunctional tumors usually is not required. In one large series investigating nonfunctioning pancreatic masses, nonductal neoplasms represented >8% of all the pancreatic tumors evaluated.87 Because the prognosis and treatment of nonductal tumors is different from that of the more commonly encountered ductal neoplasms, a histopathologic diagnosis should be established for all pancreatic masses before a treatment plan is formulated. Nonfunctional tumors are predominantly malignant (90%)88 and are found most often in the head of the pancreas. Despite their size and location, 40% are resectable at the time of discovery.15 Because they do not produce debilitating syndromes related to the elaboration of humoral products, the risks and benefits of resective surgery should be weighed carefully. If a formal pancreatic resection can extirpate the entire tumor, most surgeons would agree that this is the preferred approach.89 Although some surgeons advocate tumor debulking,90 others question the advisability of any resection short of curative extirpation in patients without humorally related disease. Biliary bypass, gastrointestinal bypass, and chemical splanchnicectomy are used to relieve symptoms created by the mass effects of the tumor in patients with adenocarcinoma of the pancreas. These also are appropriate operations in patients with nonfunctional endocrine tumors of the pancreas, who often have symptoms related to local disease. Nonoperative percutaneous or endoscopic biliary bypass also can be helpful in selected cases when the risks of surgery are prohibitive. Unlike patients with adenocarcinoma of the pancreas, patients with nonfunctional endocrine tumors can have long-term survival, and this should be taken into account when considering palliative procedures. Specifically, if an operation is performed for biliary obstruction, a concomitant gastrointestinal bypass should be considered, because enteric obstruction becomes more likely with prolonged survival. Long-term survival after gastroje-junostomy also makes peristomal jejunal ulceration more likely. For this reason, vagotomy should be performed or appropriate H2-receptor–blocker prophylaxis initiated. Elevated plasma levels of a-fetoprotein have been measured in metastatic nonfunctioning endocrine tumors of the pancreas.91 Alpha-fetoprotein is, most likely, a tumor marker for all metastatic islet cell malignancies and may be used to track the progress of metastatic disease. This feature would make it particularly helpful in following nonfunctioning tumors that do not elaborate other measurable hormones or peptide markers.

SURGICAL ASPECTS OF PANCREATIC ENDOCRINE TUMORS EXOCRINE TUMORS VERSUS ENDOCRINE TUMORS Adenocarcinoma of the pancreas is the fifth leading cause of cancer death in the United States. Approximately 20,000 new cases are diagnosed each year, and 5-year survival is ~2% regardless of therapy.92 By comparison, 200 to 1000 endocrine tumors of the pancreas are found in the United States each year,29,93 and 5-year survival after surgery is nearly 100% for benign tumors and >40% for malignant tumors.94 Another advantage of operating on endocrine tumors is the significant symptomatic relief from hormonal syndromes that can be obtained by curative resection or tumor cytoreduction. SURGICAL PROCEDURES Like ductal tumors, islet cell tumors of the pancreas may be solid or cystic.95 The treatment of cystic endocrine tumors, whether benign, malignant, functional, or nonfunctional, is similar to that of solid endocrine tumors. Operations performed for endocrine tumors of the pancreas include enucleation, segmental resection, distal pancreatectomy, pancreaticoduodenectomy, total and near-total pancreatectomy, tumor cytoreduction, bypass procedures, and surgery on other involved organs such as the stomach, duodenum, and liver (see Table 160-2). Multiple cases of laparoscopic (minimally invasive) resections of islet cell tumors of the distal pancreas have been reported.96 Some centers have performed total hepatectomy with orthotopic liver transplantation for patients with metastases confined to the liver.97 One unusual feature of endocrine tumors of the pancreas not shared by nonendocrine tumors is the response to tumor debulking. With some of these tumors, surgical reduction of the size of the lesion alone can significantly improve long-term survival.98 In contrast, surgery for adenocarcinoma of the pancreas is limited mainly to total resection by pancreaticoduodenectomy or palliation with bypass. INTRAOPERATIVE LOCALIZATION Tumor localization is important and sometimes difficult for endocrine neoplasms, which can be small but clinically symptomatic. On occasion, tumors cannot be localized before surgery and must be found at surgery. Surgeons performing these operations should be familiar with intraoperative ultrasonography, duodenotomy, intraarterial methylene blue administration, and other intraoperative localizing techniques, as well as with the indications for biopsy or blind resection. Any physician evaluating patients with pancreatic masses must understand the possibility and significance of finding an endocrine tumor. Also, any surgeon operating on patients with pancreatic endocrine tumors must be thoroughly familiar with the evaluation and localization of these lesions, with the intra-operative decision-making process, and with the wide range of ablative procedures used for these unusual neoplasms. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Warshaw AL, Swanson RS. What's new in general surgery: pancreatic cancer in 1988. Ann Surg 1988; 208:541. Gold EB. Epidemiology of and risk factors for pancreatic cancer. Surg Clin North Am 1995; 75:819. National Institutes of Health, National Cancer Institute. SEER Cancer Statistics Review, 1973–1990. Bethesda, MD: National Institutes of Health, 1993. NIH publication 93-2789. De Jong SA, Pickleman J, Rainsford K. Nocturnal tumors of the pancreas. Arch Surg 1993; 128:730. Yeo CJ, Wang BH, Anthone GJ, Cameron JL. Surgical experience with pancreatic islet-cell tumors. Arch Surg 1993; 128:1143. Capella C, Heitz PH, Hofler H, et al. Revised classification of neuroendocrine tumors of the lung, pancreas and gut. Virchows Arch 1995; 425:547. Bottger TH, Seidl C, Seifert JK, et al. Value of quantitative DNA analysis in endocrine tumors of the pancreas. Oncology 1997; 54:318. Grant CS. Surgical management of malignant islet cell tumors. World J Surg 1993; 17:498. Delcore R, Friesen SR. Gastrointestinal neuroendocrine tumors. J Am Coll Surg 1994; 178:187. Strodel WE, Vinik AI, Lloyd RV, et al. Pancreatic polypeptide-producing tumors. Silent lesions of the pancreas? Arch Surg 1984; 119:508. Brunt LM, Mazoujian G, O'Dorisio TM, Wells SA. Stimulation of vasoactive intestinal peptide and neurotensin secretion by pentagastrin in a patient with VIPoma syndrome. Surgery 1994; 115:362. McCleod MK, Vinik AI. Calcitonin immunoreactivity and hypercalcitoninemia in two patients with sporadic, nonfamilial, gastroenteropancreatic neuroendocrine tumors. Surgery 1992; 111:484. Perkins PL, McLeod MK, Jin L, et al. Analysis of gastrinomas by immuno-histochemistry and in situ hybridization histochemistry. Diagn Mol Pathol 1992; 1:155. Bieligk S, Jaffe BM. Islet cell tumors of the pancreas. Surg Clin North Am 1995; 75:1025. Modlin IM, Lewis JJ, Ahlman H, et al. Management of unresectable malignant endocrine tumors of the pancreas. Surg Gynecol Obstet 1993; 176:507. Hammond PJ, Jackson JA, Bloom SR. Localization of pancreatic endocrine tumors. Clin Endocrinol (Oxf) 1994; 40:3. Vinik AI, Delbridge L, Moattari R, et al. Transhepatic portal vein catheterization for localization of insulinomas: a ten year experience. Surgery 1991; 109:1. Van Eijck CH, Brunning HA, Reubi JC, et al. Use of isotope-labeled somatostatin analogs for visualization of islet-cell tumors. World J Surg 1993; 17:444. Van Eijck CH, Lamberts SW, Lemaire LC, et al. The use of somatostatin receptor scintigraphy in the differential diagnosis of pancreatic duct cancers and islet cell tumors. Ann Surg 1996; 224:119. Doppman JL, Miller DL, Chang R, et al. Intraarterial calcium stimulation test for detection of insulinomas. World J Surg 1993; 17:439. Ariyama J, Suyama M, Satoh K, Wakabayashi K. Endoscopic ultrasound and intraductal ultrasound in the diagnosis of small pancreatic tumors. Abdom Imaging 1998; 23:380. Bottger TC, Junginger T. Is preoperative radiographic localization of islet cell tumors in patients with insulinomas necessary? World J Surg 1993; 17:427. Norton JA, Cromack DT, Shawker TH, et al. Intraoperative ultrasonographic localization of islet cell tumors. Ann Surg 1988; 207:160. Joffe SN. Pancreatic islet cell tumor. In: Cameron JL, ed. Current surgical therapy, 2nd ed. St. Louis: CV Mosby, 1986:285. Udelsman R, Yeo CJ, Hruban RH, et al. Pancreaticoduodenectomy for selected pancreatic endocrine tumors. Surg Gynecol Obstet 1993; 177:269. Menegaux F, Schmitt G, Mercadier M, Chigot JP. Pancreatic insulinomas. Am J Surg 1993; 165:243. Von Eyben FE, Grodum E, Gjessing HJ, et al. Metabolic remission with octreotide in patients with insulinoma. J Intern Med 1994; 235:245. Buchanan KD. Effects of somatostatin on neuroendocrine tumors of the gastrointestinal system. Recent Results Cancer Res 1993; 129:45. Lo CY, Lam KY, Fan ST. Surgical strategy for insulinomas in multiple endocrine neoplasia type I. Am J Surg 1998; 175:305.

29a.Tomassetti P, Migliori M, Corinaldesi R, Gullo L. Treatment of gastroentero-pancreatic neuroendocrine tumors with octreotide LAR. Aliment Pharmacol Ther 2000; 14:557. 30. Friesen SR. Tumors of the endocrine pancreas. N Engl J Med 1982; 306:580. 31. Shepard JJ, Challis DR, Davies PF, et al. Multiple endocrine neoplasia, type 1. Arch Surg 1993; 128:1133.

32. 33. 34. 35. 36. 37.

Cherner JA, Sawyers JL. Benefit of resection of metastatic gastrinoma in multiple endocrine neoplasia type 1. Gastroenterology 1992; 102:109. Zollinger RM, Ellison EH. Primary peptic ulcerations of the jejunum associated with islet cell tumors of the pancreas. Ann Surg 1955; 142:709. Stabile BE, Morrow DJ, Passaro E. The gastrinoma triangle: operative implications. Am J Surg 1984; 147:25. Sawicki MP, Howard TJ, Dalton M, et al. The dichotomous distribution of gastrinomas. Arch Surg 1990; 125:1584. Howard TJ, Sawicki MP, Stabile BE, et al. Biologic behavior of sporadic gastrinoma located to the right and left of the superior mesenteric artery. Am J Surg 1993; 165:101. Imamura M, Takahashi K. Use of selective arterial secretin injection test to guide surgery in patients with Zollinger Ellison syndrome. World J Surg 1993; 17:433.

37a.Jensen RT, Gibril F. Somatostatin receptor scintography in gastrinomas. Ital J Gastroenterol Hepatol 1999; 31(Suppl 2):S179. 38. Kvols L, Brown M, O'Connor L, et al. Evaluation of radiolabeled somatostatin analog (e.g., 123 I-octreotide) in the detection and localization of carcinoid and islet cell tumors. Radiology 1993; 187:129. 39. Thompson NW, Pasieka J, Fukuuchi A. Duodenal gastrinomas, duodenotomy, and duodenal exploration in the surgical management of Zollinger-Ellison syndrome. World J Surg 1993; 17:455. 40. Ko TC, Flisak M, Prinz RA. Selective intra-arterial methylene blue injection: a novel method of localizing gastrinoma. Gastroenterology 1992; 102:1062. 41. Harmon JW, Norton JA, Collin MJ, et al. Removal of gastrinomas for control of Zollinger-Ellison syndrome. Ann Surg 1984; 200:396. 42. Norton JA, Doppman JL, Jensen RT. Curative resection in Zollinger-Ellison syndrome: results of a 10-year prospective study. Ann Surg 1992; 215:8. 43. Howard T, Zinner M, Stabile B, et al. Gastrinoma excision for cure. Ann Surg 1990; 211:9. 44. Delcore R, Friesen SR. Role of pancreatoduodenectomy in the management of primary duodenal wall gastrinomas in patients with Zollinger-Ellison syndrome. Surgery 1992; 112:1016. 45. Orloff SL, Debas HT. Advances in the management of patients with Zollinger-Ellison syndrome. Surg Clin North Am 1995; 75:511. 46. Phan GQ, Yeo CJ, Cameron JL, et al. Pancreaticoduodenectomy for selected periampullary neuroendocrine tumors: fifty patients. Surgery 1997; 122:989. 47. Richardson CT, Feldman M, McClelland RN, et al. Effect of vagotomy in Zollinger-Ellison syndrome. Gastroenterology 1979; 77:682. 48. Zollinger RM, Ellison EC, Fabri PJ, et al. Primary peptic ulcerations of the jejunum associated with islet cell tumors: twenty five year evaluation. Ann Surg 1980; 192:422. 49. Farley DR, van Heerden JA, Grant CS. The Zollinger-Ellison syndrome: a collective surgical experience. Ann Surg 1992; 215:561. 50. Sheppard B, Norton J, Doppman J, et al. Management of islet cell tumors in patients with multiple endocrine neoplasia: a prospective study. Surgery 1989; 106:1108. 51. Delcore R, Friesen SR. Zollinger-Ellison syndrome. Arch Surg 1991; 126:556. 52. Fishbeyn VA, Norton JA, Benya RV, et al. Assessment and prediction of long-term cure in patients with the Zollinger-Ellison syndrome: the best approach. Ann Intern Med 1993; 119:199. 53. Pisegna JR, Norton JA, Slimak GG, et al. Effects of curative gastrinoma resection on gastric secretory function and antisecretory drug requirement in the Zollinger-Ellison syndrome. Gastroenterology 1992; 102:767. 54. Arnold R, Neuhaus C, Benning R, et al. Somatostatin analog Sandostatin and inhibition of tumor growth in patients with metastatic endocrine gastroenteropancreatic tumors. World J Surg 1993; 17:511. 54a.Sato T, Konishi K, Kimura H, et al. Strategy for pancreatic endocrine tumors. Hepatogastroenterology 2000; 47:537. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100.

Soga J, Yakuwa Y. Glucagonomas/diabetico-dermatogenic syndrome (DDS): a statistical evaluation of 407 reported cases. J Hepatobiliary Pancreat Surg 1998; 5:312. Becker SW, Kahn D, Rothman S. Cutaneous manifestations of internal malignant tumors. Arch Dermatol Syph 1942; 45:1069. McGavran MH, Unger RH, Recant L, et al. A glucagon-secreting alpha-cell carcinoma of the pancreas. N Engl J Med 1966; 274:1408. Higgins GA, Recant L, Fischman AB. The glucagonoma syndrome: surgically curable diabetes. Am J Surg 1979; 137:142. Norton JA, Kahn CR, Shiebinger R, et al. Amino acid deficiency and the skin rash associated with glucagonoma. Ann Intern Med 1979; 91:213. Ingemansson S, Holst J, Larsson LI, Lunderquist A. Localization of glucagonomas by catheterization of the pancreatic veins and with glucagon assay. Surg Gynecol Obstet 1977; 145:509. Park S, O'Dorisio M, O'Dorisio T. Vasoactive intestinal polypeptide-secreting tumours: biology and therapy. Baillières Clin Gastroenterol 1996; 10:673. Jaffe BM. Surgery for gut hormone-producing tumors. Am J Med 1987; 82:68. Cesani F, Ernst R, Walser E, Villanueva-Meyer J. Tc-99m sestamibi imaging of a pancreatic VIPoma and parathyroid adenoma in a patient with multiple type I endocrine neoplasia. Clin Nucl Med 1994; 19:532. Nagorney DM, Bloom SR, Polak JM, Blumgart LH. Resolution of recurrent Verner-Morrison syndrome by resection of metastatic vipoma. Surgery 1983; 93:348. Kraenzlin ME, Ch'ng JLC, Wood SM, et al. Long-term treatment of a VIPoma with somatostatin analogue resulting in remission of symptoms and possible shrinkage of metastases. Gastroenterology 1985; 88:185. Mathoulin-Portier MP, Payan MJ, Monges G, et al. Pancreatic and duodenal somatostatinoma. Two clinico-pathologic entities. Ann Pathol 1996; 16:299. Larsson LI, Holst JJ, Kuhl C, et al. Pancreatic somatostatinoma: clinical features and physiologic implications. Lancet 1977; 1:666. Sakazaki S, Umeyama K, Nakagawa H, et al. Pancreatic somatostatinoma. Am J Surg 1983; 146:674. Kelly TR. Pancreatic somatostatinoma. Am J Surg 1983; 146:671. Krejs GJ, Orci L, Conlon JM, et al. Somatostatinoma syndrome: biochemical, morphologic and clinical features. N Engl J Med 1979; 301:285. Kaneko H, Yanaihara N, Ito S, et al. Somatostatinoma of the duodenum. Cancer 1979; 44:2273. O'Brien TD, Chejfec G, Prinz RA. Clinical features of duodenal somatostatinomas. Surgery 1993; 114:1144. Angeletti S, Corleto VD, Schillaci O, et al. Use of the somatostatin analogue octreotide to localise and manage somatostatin-producing tumours. Gut 1998; 42:792. Adrian T, Uttenthal L, Williams S, et al. Secretion of pancreatic polypeptide in patients with pancreatic endocrine tumors. N Engl J Med 1986; 315:287. Mozell E, Stenzell P, Woltering E, et al. Functional endocrine tumors of the pancreas: clinical presentation, diagnosis, and treatment. Curr Probl Surg 1990; 27:303. Miraliakbari BA, Asa L, Boudreau SF. Parathyroid hormone–like peptide in pancreatic endocrine carcinoma and adenocarcinoma associated with hypercalcemia. Hum Pathol 1992; 23:884. Tarver DS, Birch SJ. Case report: life-threatening hypercalcemia secondary to pancreatic tumor secreting parathyroid hormone–related protein—successful control by hepatic arterial embolization. Clin Radiol 1992; 46:204. Mitlak BH, Hutchinson JS, Kaufman SD, Nussbaum SR. Parathyroid hormone–related peptide mediates hypercalcemia in an islet cell tumor of the pancreas. Horm Metab Res 1991; 23:344. Price DE, Absalom SR, Davidson K, et al. A case of multiple endocrine neoplasia: hyperparathyroidism, insulinoma, GRF-oma, hypercalcitoninemia and intractable peptic ulceration. Clin Endocrinol (Oxf) 1992; 37:187. Gullo L, De Giorgio R, D'Errico A, et al. Pancreatic exocrine carcinoma producing adrenocorticotropic hormone. Pancreas 1992; 7:172. Tsuchihashi T, Yamaguchi K, Abe K, et al. Production of immunoreactive corticotropin-releasing hormone in various neuroendocrine tumors. Jpn J Clin Oncol 1992; 22:232. Amikura K, Alexander HR, Norton JA, et al. Role of surgery in management of adrenocorticotropic hormone–producing islet cell tumors of the pancreas. Surgery 1995; 118:1125. Venkatesh S, Ordonez NG, Ajani J, et al. Islet cell carcinoma of the pancreas. Cancer 1990; 65:354. Buetow PC, Miller DL, Parrino TV, Buck JL. Islet cell tumors of the pancreas: clinical, radiologic, and pathologic correlation in diagnosis and localization. Radiographics 1997; 17:453. White TJ, Edney JA, Thompson JS, et al. Is there a prognostic difference between functional and nonfunctional islet cell tumors? Am J Surg 1994: 168:627. Heitz PU, Kasper M, Polak JM, et al. Pancreatic endocrine tumors. Hum Pathol 1982; 13:263. De Jong SA, Pickleman J, Rainsford K. Nonductal tumors of the pancreas. The importance of laparotomy. Arch Surg 1993; 128:730. Kent RB, van Heerden JA, Weiland LH. Nonfunctioning islet cell tumors. Ann Surg 1981; 193:185. Evans DB, Skibber JM, Lee JE, et al. Nonfunctioning islet cell carcinoma of the pancreas. Surgery 1993; 114:1175. Eckhauser FE, Cheung PS, Vinik AI, et al. Nonfunctioning malignant neuroendocrine tumors of the pancreas. Surgery 1986; 100:978. Lesur G, Bergemer AM, Turner L, et al. Increases in alpha-fetoprotein in pancreatic endocrine tumors with hepatic metastases. Gastroenterol Clin Biol 1996; 20:204. Gordis L, Gold EB. Epidemiology of pancreatic cancer. World J Surg 1984; 8:808. Brennan MF, MacDonald JS. The endocrine pancreas. In: DeVita V, Hellman S, Rosenberg SA, eds. Principles and practice of oncology, 2nd ed. Philadelphia: JB Lippincott, 1985:1206. Thompson GB, van Heerden JA, Grant CS, et al. Islet cell carcinoma of the pancreas: a twenty-year experience. Surgery 1988; 104:1011. Schwartz RW, Munfakh NA, Zweng T, et al. Nonfunctioning cystic neuroendocrine neoplasms of the pancreas. Surgery 1994; 115:645. Vezakis A, Davides D, Larvin M, McMahan MJ. Laparoscopic surgery combined with preservation of the spleen for distal pancreatic tumors. Surg Endosc 1999; 13:26. Dousset B, Houssin D, Soubrane O, et al. Metastatic endocrine tumors: is there a place for liver transplantation? Liver Transpl Surg 1995; 1:111. Danforth DN, Gorden P, Brennan MF. Metastatic insulin-secreting carcinoma of the pancreas: clinical course and the role of surgery. Surgery 1984; 96:1027. Debas HT, Mulvihill SJ. Neuroendocrine gut neoplasms: important lessons from uncommon tumors. Arch Surg 1994; 129:965. Buchanan KD, Johnston CF, O'Hare MMT, et al. Neuroendocrine tumors: a European view. Am J Surg 1986; 81:14.

CHAPTER 161 HYPOGLYCEMIA OF INFANCY AND CHILDHOOD Principles and Practice of Endocrinology and Metabolism

CHAPTER 161 HYPOGLYCEMIA OF INFANCY AND CHILDHOOD JOSEPH I. WOLFSDORF AND MARK KORSON Increased Susceptibility of the Infant and Child to Hypoglycemia An Overview of Fuel Metabolism Definition of Hypoglycemia Clinical Manifestations of Hypoglycemia Neonatal Hypoglycemia Causes of Hypoglycemia in Infancy and Childhood Accelerated Starvation (Ketotic Hypoglycemia; Transient Intolerance of Fasting) Diagnosis Treatment Hyperinsulinism Diagnosis Treatment Hormone Deficiency Adrenocorticotropic Hormone/Cortisol Deficiency Hypopituitarism Disorders of Glycogen Synthesis, Glycogen Degradation, and Gluconeogenesis Glycogen Synthetase Deficiency Glucose-6-Phosphatase Deficiency Amylo-1,6-Glucosidase Deficiency (Type III Glycogen Storage Disease; Glycogen Debrancher Enzyme Deficiency) Hepatic Glycogen Phosphorylase Deficiency Phosphorylase Kinase Deficiency Defects in Gluconeogenesis Galactosemia Hereditary Fructose Intolerance Defects of Carnitine Metabolism and Fatty Acid b-oxidation Defects in Amino-Acid Metabolism Determining the Cause Evaluation of Response to Therapy and Laboratory Evaluation Chapter References

Glucose is the predominant metabolic fuel utilized by the brain. Because the brain cannot synthesize glucose or store more than a few minutes' supply as glycogen, survival of the brain requires a continuous supply of glucose.1 Recurrent hypoglycemia during the period of rapid brain growth and differentiation in infancy can result in long-term neurologic sequelae and psychomotor retardation. Therefore, prevention of hypoglycemia and expeditious diagnosis and vigorous treatment when it occurs are essential to prevent the potentially devastating consequences of hypoglycemia on the brain.

INCREASED SUSCEPTIBILITY OF THE INFANT AND CHILD TO HYPOGLYCEMIA Hypoglycemia is most common in the newborn period. During infancy and childhood, it occurs most frequently when nighttime feeding is discontinued and when intercurrent illness interrupts the normal feeding pattern, causing periods of relative starvation. Basal energy needs during infancy are high. A full-term newborn baby, for example, has a ratio of surface area to body mass that is more than twice that of an average adult, necessitating a high rate of energy expenditure to maintain body temperature. Also, the infant brain is large relative to body mass and its energy requirement is mainly derived from the oxidation of circulating glucose. To meet the high demand for glucose, the rate of glucose production in infants and young children is two to three times that of older children and mature adults.2 Although the demand for glucose is high, the activity of several liver enzymes involved in energy production is low in the newborn compared to that of older children and adults. Consequently, until feeding is well established, maintenance of glucose homeostasis in the newborn period is more precarious than it is later in childhood. In the postabsorptive state, the rate of glucose turnover in adults is ~2 mg/kg per minute (8–10 g per hour), whereas the average basal (4–6 hours after feeding) rate of glucose turnover is 6 mg/kg per minute in newborns, approximately three times the adult rate. During prolonged fasting, infants and children cannot sustain this high rate of glucose production. Normal children, 18 months to 9 years of age, fasted for 24 hours, have a mean blood glucose concentration of 52 ± 14 (standard deviation [SD]) mg/dL. Indeed, 22% have blood glucose concentrations 100 µU/mL) and is confirmed by finding concomitantly low or suppressed serum C-peptide levels.

TREATMENT The goal of therapy of the hyperinsulinemic hypoglycemia of infancy is to prevent hypoglycemia to protect the developing brain from possible damage, using a regimen that can be safely and effectively implemented at home. Successful treatment is defined by the following criteria: maintaining plasma glucose concentrations >60 mg/dL (>3.3 mmol/L) without intravenous infusions of glucose or glucagon, a feeding schedule appropriate for the age of the infant, and a fasting tolerance of 6 to 8 hours in the newborn infant, 12 hours in infants up to 1 year of age, or 16 hours or more in older children. Prompt effective treatment is necessary to minimize the risk of long-term adverse neurologic sequelae.42 Initially, this requires a glucose infusion at two- to four-fold (average 14.5 ± 1.7 mg/kg per minute43) the basal rate of glucose production, and occasionally reaches 25 mg/kg per minute. Placement of a central venous line is usually necessary to be able to infuse hypertonic glucose solutions. Treatment with oral diazoxide, which opens normal KATP channels and thereby suppresses insulin secretion, should be given a trial (15–25 mg/kg per day in 3 doses at 8-hour intervals). Its effect may be potentiated by the addition of a thiazide diuretic. Diazoxide is ineffective in infants whose hyperinsulinism is caused by mutations of the KATP channel. A long-acting somatostatin analog (octreotide) may be successful in maintaining normoglycemia in up to 50% of cases of congenital hyperinsulinism. Octreotide inhibits insulin secretion by decreasing the influx of calcium ions into B cells and through a direct effect on secretory granules. The starting dose is 5 µg/kg every 6 to 8 hours. If glucose is not maintained (³60 mg/dL), the dosage of octreotide is increased up to a maximum of 40 to 60 µg/kg per day, divided into three to six doses. Because of the marked variability of response to octreotide, the therapeutic regimen has to be adapted for each individual patient, and its effects closely monitored.44 Many infants fail to respond to medical therapy and require a 95% subtotal pancreatectomy to restore normoglycemia.45,45a HORMONE DEFICIENCY ADRENOCORTICOTROPIC HORMONE/CORTISOL DEFICIENCY Cortisol limits glucose utilization in several tissues, including skeletal muscle, by directly opposing the action of insulin and, secondarily, by promoting lipolysis. It stimulates protein breakdown and increases release of gluconeogenic precursors from muscle and fat. Cortisol stimulates hepatic gluconeogenesis and glycogen synthesis and exerts permissive influences on the gluconeogenic and glycogenolytic effects of glucagon and epinephrine. By all these effects, cortisol tends to raise plasma glucose concentrations. Adrenocortical insufficiency should be considered in the differential diagnosis of patients who present with hypoglycemia and ketosis. In infancy, adrenocortical insufficiency may be secondary to congenital adrenal hyperplasia or congenital adrenal hypoplasia. In older children, adrenocortical insufficiency is more likely to be caused by Addison disease. Adrenocorticotropic hormone (ACTH) deficiency or panhypopituitarism can present with hypoglycemia in infancy or in later childhood. Diagnosis. A serum cortisol concentration 10 mg/dL confers a 100% increased risk in women, and a substantially increased risk in men.70,71,72,73 and 74 Similarly, an Lp(a) cholesterol >10 mg/dL confers a 100% increased CHD risk, especially in men.98 An Lp(a) cholesterol and remnant lipoprotein cholesterol should be part of CHD risk assessment.74,98,112,113,114 and 115 Measurement of other parameters cannot be recommended at this time. NATIONAL CHOLESTEROL EDUCATION PROGRAM GUIDELINES The NCEP Adult Treatment Panel has developed guidelines for the diagnosis and treatment of individuals older than 20 years of age with elevated blood cholesterol levels associated with an increase in LDL cholesterol levels.3,4 The goals of therapy and the particular level of LDL cholesterol requiring the initiation of diet and drug therapy depend on the presence or absence of CHD or two or more CHD risk factors (Table 163-2). The presence of secondary causes of elevated LDL cholesterol levels (>160 mg/dL or 4.1 mmol/L) must be ruled out. These include hypothyroidism, obstructive liver disease, and nephrotic syndrome. LDL cholesterol decision points for initiating diet and drug therapy are given in Table 163-3. The NCEP guidelines have been accepted by all major U.S. medical organizations, including the American College of Physicians, the American Heart Association, and the American Medical Association.4 Guidelines for the general population and children and adolescents have also been developed.116,117

TABLE 163-2. National Cholesterol Education Program Major Coronary Heart Disease Risk Factors in Addition to Low-Density Lipoprotein Cholesterol

TABLE 163-3. National Cholesterol Education Program Adult Treatment Panel II Treatment Guidelines

The recommendation that LDL cholesterol values be used as the primary criterion for treatment decisions for patients with elevated cholesterol levels makes accurate measurement a national public health imperative as reviewed by the NCEP Laboratory Standardization Panel.118 If a patient has an LDL cholesterol level of 160 mg/dL (4.1 mmol/L), it represents approximately the 75th percentile for middle-aged Americans (see Table 163-1). It is important to confirm any abnormalities by repeat determinations. Hospitalization or acute illness can affect lipid values, and lipid determinations should generally be carried out in the free-living state.119 An elevated or borderline-high triglyceride level (200 mg/dL or 2.3 mmol/L) has not clearly been shown to be an independent risk factor for premature heart disease. However, an elevated triglyceride level is inversely associated with a low level of HDL cholesterol, which has been shown to be a significant risk factor for CHD. Common secondary causes of elevated LDL cholesterol and triglyceride values and of decreased HDL cholesterol include diabetes, hypothyroidism, obstructive liver disease, kidney disease, excess alcohol intake, and the use of corticosteroids, anabolic steroids, estrogens, b-blocking agents, and thiazide diuretics.4 If possible, these factors should be screened for and treated before diet or drug therapy for lipid disorders is initiated. Screening should include an evaluation of glucose, albumin, liver transaminases, alkaline phosphatase, creatinine, and thyroid-stimulating hormone, and the patient should be asked about alcohol intake and the use of b-blockers, estrogens, corticosteroids, anabolic steroids, and thiazides. LOW-DENSITY LIPOPROTEIN CHOLESTEROL MEASUREMENT Unlike total cholesterol quantitation, there is no consensus-approved and validated reference method for the direct measurement of LDL cholesterol. The accurate measurement of LDL cholesterol depends on the separation of LDL particles in serum from other lipoproteins: chylomicrons, VLDL, and HDL. Traditionally, LDL has been defined as all lipoproteins within the density range of 1.019 to 1.063 g/mL. However, in common practice, the definition has been broadened to include intermediate-density lipoprotein (IDL; 1.006–1.019 g/mL). Using this definition, LDL is composed of LDL + IDL + Lp(a). This definition serves as the basis for the cut-points defined by the NCEP Adult Treatment Panel. The options for measuring LDL cholesterol include ultracentrifugation, the Friedewald calculation for estimating LDL cholesterol levels, and a direct method for measuring LDL cholesterol that uses immunoseparation of lipoproteins by their respective apolipoprotein content. Ultracentrifugation involves the separation of lipoproteins based on their density differences after an 18-hour spin at 109,000 × g. The VLDL and chylomicrons float to the top and are separated using a tube slicing technique from the 1.006-g/mL infranatant (i.e., “1.006 bottom”). This infranatant fraction contains LDL and HDL. A heparin-manganese precipitation reagent is added to the 1.006 bottom to precipitate LDL, leaving HDL in the supernatant. The cholesterol concentrations of the 1.006 g/mL of infranatant and the HDL cholesterol supernatant are measured using the Abell-Kendall cholesterol reference method: LDL cholesterol = infranatant cholesterol – HDL cholesterol. This procedure has been adopted by the Centers for Disease Control and Prevention and the Reference Network Laboratories for Standardizations as a means of directly measuring LDL cholesterol in the research setting and serves as the standard.118 However, ultracentrifugation is poorly suited to the routine, clinical laboratory for several reasons. It requires cumbersome procedures; it is extremely labor intensive and technique dependent; it requires expensive instrumentation; and although it is the accepted reference method, it is an indirect measurement. Most clinical laboratories use the equation known as the Friedewald formula119 to estimate a patient's LDL cholesterol concentration: estimation of LDL cholesterol = total cholesterol – HDL cholesterol – VLDL cholesterol. The estimation of VLDL cholesterol equals the triglyceride level divided by five.119 The Friedewald formula estimates the LDL cholesterol concentration by subtracting the cholesterol associated with the other classes of lipoproteins from total cholesterol. This involves three independent lipid analyses, each contributing a potential source of error. It also involves a potentially inaccurate estimate of VLDL cholesterol. Because no direct VLDL cholesterol assay is available, it is calculated from the triglyceride value divided by a factor of five. This divisor can also add error to all LDL cholesterol estimates, but it is especially inappropriate for individuals with elevated triglyceride levels. Clinical laboratories use automated enzymatic analyses for cholesterol and triglyceride within serum or plasma, and HDL cholesterol is measured after precipitation of other lipoproteins in serum or plasma with heparin manganese chloride, dextran magnesium sulfate, or phosphotungstic acid.118 On-line direct assays of HDL cholesterol are now available. The drawbacks of using the Friedewald formula for determining levels of LDL cholesterol are that it is estimated by calculation; it requires multiple assays and multiple steps, each adding a potential source of error; it is increasingly inaccurate as triglyceride levels increase; it requires that patients fast for 12 to 14 hours before specimen collection to avoid a triglyceride bias; and it is not standardized.118,119 Moreover, LDL cholesterol concentrations cannot be reported for individuals with elevated triglyceride levels (>400 mg/dL or 4.5 mmol/L) or in the nonfasting state.119,120 It has been reported that the formula becomes increasingly inaccurate in calculating true LDL cholesterol levels at borderline triglyceride levels (200–400 mg/dL or 2.3–4.5 mmol/L).112,113,114 and 115 The inadequacies of the methods for measuring LDL cholesterol necessitated the development of a direct method by which clinical laboratories may accurately and practically assess LDL cholesterol concentrations in patient samples. In 1990, the Laboratory Standardization Panel of the NCEP recommended the development of a direct LDL cholesterol measurement method.118 The direct method for measuring serum or plasma LDL cholesterol concentration that was introduced was suitable for routine use in the clinical laboratory. This immunoseparation technology uses affinity-purified goat polyclonal antisera to human apo A-I and apo E, which are coated on latex particles; this facilitates the removal of chylomicrons, VLDL, and HDL in nonfasting or fasting specimens. After incubation and centrifugation, LDL cholesterol remains in the filtrate solution. The LDL cholesterol concentration is obtained by performing an enzymatic cholesterol assay on the filtrate solution. The direct LDL cholesterol immunoseparation method allows for the direct quantitation of LDL cholesterol from one measurement, the use of fasting and nonfasting samples, and an LDL cholesterol measurement regardless of elevated triglyceride levels. When the direct LDL cholesterol assay was carried out on serum obtained from 115 subjects, who were fasting or nonfasting and were normal or hyperlipidemic, and was compared with those obtained by ultracentrifugation analysis, the correlation was 0.97, with a small negative bias of 2.9%. Subjects with LDL cholesterol levels ³160 mg/dL, as obtained by ultracentrifugation, were correctly classified 93.8% of the time. In a similar study carried out on serum obtained from 177 subjects with normal or elevated lipid levels, the correlation between the direct LDL cholesterol and the value obtained by ultracentrifugation was 0.98, with between-run and within-run coefficients of variation 250 mg/dL, but triglyceride and HDL cholesterol levels are generally normal.50 Clinically, these patients usually develop arcus senilis and tendinous xanthomas. The clinical diagnosis, in the author's view, is established by an LDL cholesterol level above the 90th percentile in two or more family members and the presence of tendinous xanthomas within the kindred (Fig. 163-1). The average age of onset of CHD is ~45 years for men and 55 years for women with untreated heterozygous familial hypercholesterolemia.50 These patients may also have a higher than normal prevalence of calcific aortic stenosis.

FIGURE 163-1. A and B, Tuberous xanthomas in a patient with homozygous familial hypercholesterolemia (i.e., low-density lipoprotein receptor negative) and tendinous xanthomas (arrows). C and D, Circus cornease (arrow) and severe coronary atherosclerosis in patients with heterozygous familial hypercholesterolemia.

Treatment consists of an NCEP step 2 diet low in saturated fat (300 mg/dL and total cholesterol >400 mg/dL. POLYGENIC FAMILIAL HYPERCHOLESTEROLEMIA WITHOUT XANTHOMAS Among the 3% of CHD patients who have familial hypercholesterolemia, only one-third have tendinous xanthomas and are truly heterozygous for familial hypercholesterolemia with potential LDL receptor defects.147,148 and 149 Other CHD kindreds with familial hypercholesterolemia have more modest LDL cholesterol elevations (>190 mg/dL) without xanthomas. These kindreds have been classified as polygenic familial hypercholesterolemia; no clear defect has been found. Having the apo E-4 allele is known to be associated with elevations in LDL cholesterol levels, and these patients may be more likely to be heterozygous or homozygous for apo E-4.115,163,164 The clinical diagnosis of this disorder is established by the presence of LDL cholesterol values greater than the 90th percentile in two or more family members and a lack of xanthomas in the family. The treatment of these patients includes implementation of an NCEP step 2 diet and the use of cholesterol-lowering medications, such as HMG-CoA reductase inhibitors, resins, or niacin. FAMILIAL COMBINED HYPERLIPIDEMIA Familial combined hyperlipidemia was initially characterized by the finding of hypercholesterolemia and hypertriglyceridemia within the same kindred and by relatives having one or both of these abnormalities.147,148 and 149,165 This disorder was found in ~10% of myocardial infarction survivors younger than 60 years of age. Using 95th percentile criteria for serum cholesterol and triglyceride levels, affected subjects were shown to have elevations in VLDL, LDL, or both on phenotyping analysis.148,149 In the author's series, ~14% of kindreds with premature CHD had familial combined hyperlipidemia.

The clinical diagnosis of familial combined hyperlipidemia is established by the finding of serum or plasma LDL cholesterol or triglyceride levels above the 90th percentile (usually LDL cholesterol >190 mg/dL or triglyceride >50 mg/dL) within the family and in at least two family members, with both abnormalities occurring within the kindred.150 Most patients with familial combined hyperlipidemia also have HDL cholesterol values below the tenth percentile.150 It has been reported that patients with familial combined hyperlipidemia have overproduction of apo B-100, but the precise defect is unknown.166,167 and 168 Data indicate that hepatic apo B-100 secretion is largely substrate driven.169 Patients with familial combined hyperlipidemia often are overweight and hypertensive, and they may also be diabetic and have gout. Treatment with diet, an exercise program, and, if necessary, the use of an HMG-CoA reductase inhibitor is important for CHD prevention. Niacin is an alternative form of therapy in this condition. FAMILIAL HYPERAPOBETALIPOPROTEINEMIA Familial hyperapobetalipoproteinemia is characterized by apo B values above the 90th percentile in the absence of other lipid abnormalities in the kindred with at least two affected family members, using age- and gender-adjusted norms.169,170,171 and 172 This disorder occurred in 5% of CHD kindreds in the author's series; it is thought to be a variant of familial combined hyperlipidemia. It also is associated with overproduction of apo B-100.150 FAMILIAL DYSLIPIDEMIA Familial hypertriglyceridemia is a common familial lipid disorder in which at least two kindred members have fasting triglyceride levels greater than the 90th percentile of normal. Approximately 5% of myocardial infarction survivors younger than 60 years of age were found to have this disorder, using the 95th percentile as the standard.147,148 and 149 In the author's studies, using the 90th percentile, ~15% of CHD kindreds had this disorder.150 In the author's series, all kindreds except one had HDL cholesterol deficiency within the family as well. The author and colleagues have named this disorder familial dyslipidemia. Hypertriglyceridemia and HDL cholesterol levels must be less than the tenth percentile in the kindred for the family to have this condition.150 These patients are frequently overweight and may have male-pattern obesity, insulin resistance, type II diabetes, and hypertension. The precise defect is unknown, but the patients have increased hepatic triglyceride secretion and enhanced HDL apo A-I fractional catabolism.166,167 and 168,173,174,175 and 176 No clear therapeutic guidelines have been formulated, other than treatment of other CHD risk factors, a diet and exercise program, and optimization of LDL cholesterol levels. However, in CHD patients with this disorder and LDL cholesterol 10 mg/dL) and decreased HDL C (500:1, the primary target of released dynorphin probably is within the pituitary itself. A paracrine role for secretory regulation of the neurohypophysial peptides seems likely, because opioid-induced modulation of vasopressin and oxytocin secretion from isolated neural lobes of rat pituitary has been reported. Also k-selective agonists induce a profound diuresis,44 whereas the opiate antagonist naloxone potentiates the stimulated release of vasopressin and oxytocin. Thus, evidence exists for opioid regulation of neurohypophysial peptide secretion. ENDOGENOUS OPIOIDS IN THE ADRENAL MEDULLA Proenkephalin products are present in relatively high concentrations in the adrenal medulla of most species. In some species, including humans, opioids derived from prodynorphin and POMC are also present. The enkephalins are stored together with catecholamines in secretory vesicles of adrenal chromaffin cells and are secreted into adrenal venous blood after treatments that stimulate catecholamine secretion.45 Studies have also revealed the presence of opioid receptors in adrenal medulla, a finding that suggests a local mechanism for controlling the release of adrenal catecholamines. Enkephalin-containing peptides are secreted into the circulation after splanchnic nerve stimulation to exert effects in other tissues. Although pentapeptide enkephalins are hydrolyzed very rapidly in circulating blood, a role for enkephalins secreted from the adrenal medulla in modulating the function of circulating cells of the immune system is likely. Finally, the colocalization of enkephalins with catecholamines, as well as the nature of some of the potential target sites for adrenal enkephalins, suggest that they may have a role that is complementary to that of the catecholamines in the physiologic response to stress. ENDOGENOUS OPIOIDS AND THE GASTROINTESTINAL TRACT

Both enkephalins and dynorphins are present in the neurons that innervate the gastrointestinal tract.46,47 Enkephalins and dynorphins are present in many myenteric plexus neurons and a few neurons of the submucous plexus throughout the gastrointestinal tract. These neurons innervate sphincters, circular muscle, and longitudinal muscle of the stomach and intestine. The release of dynorphin-like material into the vascular perfusate of guinea pig intestine during peristalsis has been reported. The presence of µ, d, and k opioid receptors throughout the digestive tract48 suggests that endogenous opioid peptides act locally in this system. Gastrointestinal motility is substantially modified by opioids, but the effects often vary widely among species. In humans, opioids increase the tone of sphincters and the frequency of segmenting, nonpropulsive contractions of circular muscle. This action has been ascribed to the increased local release of serotonin by opioids. A reduction in nitric oxide–dependent neurotransmission may also be involved.49 Concomitantly, the peristaltic reflex is depressed, and the contractions of longitudinal muscle are decreased as a result of a reduction in acetylcholine secretion. Enkephalins reduce secretions, both in the stomach and in the intestine. Thus, the propulsion of gastrointestinal contents is decreased, and water content is reduced. Opioids, therefore, provide symptomatic relief of diarrhea. These actions are all reversed by naloxone. Hydrolysis of the milk protein b-casein yields a peptide–b-casomorphin that has some activity at µ receptors.50 In breast-fed neonates, this peptide may be produced locally in sufficient concentrations to reduce gastrointestinal activity.51 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.

Hughes J, Smith TW, Kosterlitz HW, et al. Identification of two related pentapeptides from the brain with potent opiate agonist activity. Nature 1975; 258:577. Cox BM. Endogenous opioid peptides: a guide to structures and terminology. Life Sci 1982; 31:1645. Mollereau C, Parmentier M, Mailleux P, et al. ORL-1, a novel member of the opioid receptor family—cloning, functional expression and localization. FEBS Lett 1994; 341:33. Meunier JC, Mollereau C, Toll L, et al. Isolation and structure of the endogenous agonist of opioid receptor-like ORL-1 receptor. Nature 1995; 377:532. Reinscheid RK, Nothacker H-P, Bourson A, et al. Orphanin FQ: a neuropeptide that activates an opioidlike G protein-coupled receptor. Science 1995; 270:792. Calo G, Guerrini R, Rizzi A, et al. Pharmacology of nociceptin and its receptor: a novel therapeutic target. Br J Pharmacol 2000; 129:1261. Douglass J, Cox BM, Quinn B, et al. Expression of the prodynorphin gene in male and female mammalian reproductive tissues. Endocrinology 1987; 120:707. Herbert E, Seasholtz A, Comb M, et al. Study of the regulation of expression of neuropeptide genes by gene transfer methods. In: Meltzer HY, ed. Psychopharmacology: the third generation of progress. New York: Raven Press 1987:373. Leslie FM, Chen Y, Winzer-Serhan UH. Opioid receptor and peptide mRNA expression in proliferative zones of fetal rat central nervous system. Can J Physiol Pharmacol 1998; 76:284. Mollereau C, Simons MJ, Soularue P, et al. Structure, tissue distribution, and chromosomal localization of the prenociceptin gene. Proc Natl Acad Sci U S A 1996; 93:8666. Steiner DF. The proprotein convertases. Curr Opin Chem Biol 1998; 2:31. Jackson RS, Creemers JW, Ohagi S, et al. Obesity and impaired prohormone processing associated with mutations in the human prohormone convertase 1 gene. Nat Genet 1997; 16:218. Asteria C. Genetic alterations of enzymes involved in prohormone processing as common molecular mechanisms of human and rodent obesity. Eur J Endocrinol 1998; 139:150. Prigge ST, Kolhekar AS, Eipper BA, et al. Amidation of bioactive peptides: the structure of peptidylglycine alpha-hydroxylating monooxygenase. Science 1997; 278:1300. Bennett HPJ, Bradbury AF, Huttner WB, Smyth DG. Processing of propeptides: glycosylation, phosphorylation, sulfation, acetylation and amidation. In: Loh YP, ed. Mechanisms of intracellular trafficking and processing of proproteins. Boca Raton, FL: CRC Press, 1993:251. Dickerson IM, Noül G. Tissue-specific peptide processing. In: Fricker LD, ed. Peptide biosynthesis and processing. Boca Raton, FL: CRC Press 1991:71. Eipper BA, Bloomquist BT, Husten, et al. Peptidylglycine a-amidating monooxygenase and other processing enzymes in the neurointermediate pituitary. Ann N Y Acad Sci 1993; 689:147. Low MJ, Liu B, Hammer GD, et al. Post-translational processing of proopi-omelanocortin (POMC) in mouse pituitary melanotrope tumors induced by a POMC-simian virus 40 large T antigen transgene. J Biol Chem 1993; 268:24967. Jenck G, Moreau JL, Martin JR, et al. Orphanin FQ acts as an anxiolytic to attenuate behavioral responses to stress. Proc Natl Acad Sci U S A 1997; 94:14854. Manabe T, Noda Y, Mamiya T, et al. Facilitation of long-term potentiation and memory in mice lacking nociceptin receptors. Nature 1998; 394:577. Evans CF, Keith DE Jr, Morrison H, et al. Cloning of a delta opioid receptor by functional expression. Science 1992; 258:1952. Blake AD, Bot G, Reisine T. Structure-function analysis of the cloned opiate receptors: peptide and small molecule interactions. Chem Biol 1996; 3:967. Jiang Q, Takemori AE, Sultana M, et al. Differential antagonism of opioid delta antinociception by [D-Ala 2, Leu5, Cys6]enkephalin and naltrindole 5'-isothiocyanate: evidence for delta receptor subtypes. J Pharmacol Exp Ther 1991; 257:1069. Kim K-W, Cox BM. Inhibition of norepinephrine release from rat cortex slices by opioids: differences among agonists in sensitivities to antagonists suggest receptor heterogeneity. J Pharmacol Exp Ther 1993; 267:1153. Schulz R, Wüster M, Herz A. Pharmacological characterization of the e-opiate receptors. J Pharmacol Exp Ther 1981; 216:604. Cox BM. Opioid receptor–G protein interactions: acute and chronic effects of opioids. In: Herz A, ed. Handbook of experimental pharmacology, vol 104/I. Opioids I. Berlin: Springer-Verlag, 1993:145. Nestler EJ, Hope BT, Widnell KL. Drug addiction: a model for the molecular basis of neural plasticity. Neuron 1993; 11:995. Ignatova EG, Belcheva MM, Bohn LM, et al. Requirement of receptor internalization for opioid stimulation of mitogen-activated kinase: biochemical and immunofluorescence confocal microscopic evidence. Neuroscience 1999; 19:56. Fields HL. Brainstem mechanisms of pain modulation: anatomy and physiology. In: Hertz A, ed. Handbook of experimental pharmacology, Vol 104/II. Opioids II. Berlin: Springer-Verlag, 1993:3. Aicher SA, Punnoose A, and Goldberg A. Mu-opioid receptors often colocalize with the substance P receptor (NK1) in the trigeminal dorsal horn. J Neurosci 2000; 20:4325. Ferin M. Neuropeptides, the stress response, and the hypothalamic–pituitary– gonadal axis in the female rhesus monkey. Ann N Y Acad Sci 1993; 697:106. Faletti AG, Mastronardi CA, Lomnuczi A, et al. b-Endorphin blocks luteinizing hormone releasing hormone release by inhibiting the nitricoxidergic pathway controlling its release. Proc Natl Acad Sci U S A 1999; 722. Rivest S, Plotsky PM, Rivier C. CRF alters the infundibular LHRH secretory system from the medial preoptic area of female rats: possible involvement of opioid receptors. Neuroendocrinology 1993; 57:236. Pechnick RN. Effects of opioids on the hypothalamic-pituitary-adrenal axis. Annu Rev Pharmacol Toxicol 1993; 32:353. Vale W, Rivier J, Guillemin R, Rivier C. Effects of purified CRF and other substances on the secretion of beta-endorphin-like immunoreactivities by cultured anterior or neurointermediate pituitary cells. In: Collu R, Barbeau A, Ducharme JR, Rochefort J-G, eds. Central nervous system effects of hypothalamic hormones and other peptides. New York: Raven Press, 1979:163. Schäfer M, Carter L, Stein C. Interleukin 1b and corticotrophin-releasing factor inhibit pain by releasing opioids from immune cells in inflamed tissue. Proc Natl Acad Sci U S A 1994; 91:4219. Levine JD, Fields HL, Basbaum AI. Peptides and the primary afferent noci-ceptor. J Neurosci 1993; 13:2273. Peterson PK, Molitor TW, Chao CC. The opioid-cytokine connection. J Neuroimmunol 1998: 83:63. Sapolsky R, Rivier C, Yamamoto G, et al. Interleukin-1 stimulates the secretion of hypothalamic corticotropin-releasing factor. Science 1987; 238:522. Blalock JE. Proopiomelamoncortin and the immune-neuroendocrine connection. Ann NY Acad Sci 1999: 885:161. Yin D, Mufson RA, Wang R, Shi Y. Fas-mediated cell death promoted by opioids. Nature 1999; 397:218. Martin R, Geis R, Holl R, et al. Co-existence of unrelated peptides in oxytocin and vasopressin terminals of rat neurohypophyses: immunoreactive methionine 5-enkephalin-, leucine 5-enkephalin-, and cholecystokinin-like substances. Neuroscience 1983; 8:213. Eriksson M, Ceccatelli S, Uvnas-Moberg K, et al. Expression of Fos-related antigens, oxytocin, dynorphin and galanin in the paraventricular and supraoptic nuclei of lactating rats. Neuroendocrinology 1996; 63:356. Leander JD. A kappa opioid effect: increased urination in the rat. J Pharmacol Exp Ther 1983; 224:89. Chaminade M, Foutz AS, Rossier J. Co-release of enkephalins and precursors with catecholamines from perfused cat adrenal gland in situ. J Physiol 1984; 353:157. Yuferov VP, Culpepper-Morgan JA, La Forge KS, et al. Regional quantitation of preprodynorphin mRNA in guinea pig gastrointestinal tract. Neurochem Res 1998; 23:505. Kromer W. Endogenous and exogenous opioids in the control of gastrointestinal motility and secretion. Pharmacol Rev 1989; 40:121. Wittert G, Hope P, Pyle D. Tissue distribution of opioid receptor gene expression in the rat. Biochem Biophys Res Commun 1996; 218:877. Bayguinov O, Sanders KM. Regulation of neural responses in the canine pyloric sphincter by opioids. Br J Pharmacol 1993; 108:1024. Brantl V, Teschemacher H, Blasig J, et al. Opioid activities of beta-casomor-phins. Life Sci 1981; 28:1903. Froetschel MA. Bioactive peptides in digesta that regulate gastrointestinal function and intake. J Anim Sci 1996; 74:2500. Avidor-Reiss T, Nevo I, Saya D, et al. Opiate-induced adenylyl cyclase superactivation is isozyme-specific. J Biol Chem 1997; 272:5040. Smart D, Hirst RA, Hirota K, et al. The effects of recombinant rat µ-opioid receptor activation in CHO cells on phospholipase C, [Ca 2+]i and adenylyl cyclase. Br J Pharmacol 1997; 120:1165. Matthes HWD, Maldonaldo R, Simonin F, et al. Loss of morphine-induced analgesia, reward effect and withdrawal symptoms in mice lacking the µ-opioid-receptor gene. Nature 1996; 383:819. König M, Zimmer AM, Steiner H, et al. Pain responses, anxiety and aggression in mice deficient in pre-proenkephalin. Nature 1996; 383:535.

CHAPTER 169 SOMATOSTATIN Principles and Practice of Endocrinology and Metabolism

CHAPTER 169 SOMATOSTATIN YOGESH C. PATEL Somatostatin Genes and Gene Products Anatomic Distribution of Somatostatin Cells Somatostatin in the Plasma and Other Body Fluids Regulation of Somatostatin Secretion and Gene Expression Actions of Somatostatin Somatostatin Receptors, Receptor Subtypes, and Signal Transduction Somatostatin Receptor–Mediated Inhibition of Secretion and Cell Proliferation Physiologic Significance of Somatostatin Subtype-Selective Bioactions of Somatostatin Receptors Somatostatin in Disease Clinical Applications of Somatostatin and its Analogs Treatment of Neuroendocrine and Nonneuroendocrine Tumors Gastroenteropancreatic Tumors Treatment of Pituitary Tumors Possible Oncologic Applications Gastrointestinal Applications Orthostatic Hypotension Side Effects Somatostatin-Receptor Scans Somatostatin Receptor–Targeted Radioablation of Tumors Chapter References

Somatostatin (SST) was first found in the mammalian hypothalamus as a tetradecapeptide (SST-14), which inhibited the release of growth hormone (GH). It has since come to be known as a multifunctional hormone that is also produced throughout the central nervous system and in most peripheral organs. Somatostatin acts on a diverse array of endocrine, exocrine, neuronal, and immune cell targets to inhibit secretion, to modulate neurotransmission, and to regulate cell growth. These actions are mediated by a family of G protein–coupled receptors with five distinct subtypes (termed SSTR1 through SSTR5). SST is best regarded as an endogenous inhibitory regulator of the secretory and proliferative responses of many different target cells. In addition, the peptide may be of importance in the pathophysiology of several diseases such as neoplasia, inflammation, Alzheimer and Huntington diseases, and acquired immunodeficiency syndrome (AIDS), and has found a number of clinical applications in the diagnosis and treatment of neuroendocrine tumors and various gastrointestinal disorders.1,2,3,4,5,6,7 and 8

SOMATOSTATIN GENES AND GENE PRODUCTS Like other protein hormones, somatostatin is synthesized as part of a large precursor protein (proSST) that is processed to generate two bioactive forms, SST-14 and SST-28 (Fig. 169-1). In humans only one somatostatin gene is found, located on the long arm of chromosome 3, which encodes for both SST-14 and SST-28, whereas lower vertebrates (e.g., fish) have two somatostatin genes that separately encode for SST-14 or SST-28.2,4,8 The stranscriptional unit of the rat SST gene consists of exons of 238 and 367 base pairs (bp) separated by an intron of 621 bp (Fig. 169-2). The 5'-upstream region contains a number of regulatory elements for tissue-specific and extracellular signals, including a cyclic adenosine monophosphate (cAMP) response element (CRE) and two nonconsensus glucocorticoid response elements (GREs).2,4 The SST-14 sequence has been totally conserved throughout vertebrate evolution, whereas the aminoacid structure of SST-28 has changed ~30% during evolution from fish to humans.2,4,8 A novel second SST-like gene, cortistatin (CST), which has been described in humans, yields two cleavage products, CST-17 and CST-29, which are comparable to SST-14 and SST-28 (see Fig. 169-1).9 The CST peptides interact with all five SSTRs, but unlike somatostatin, expression of cortistatin is restricted to the cerebral cortex and its biofunction(s) remains unknown.3,9

FIGURE 169-1. Structure of naturally occurring somatostatin peptides. (SST-28, somatostatin-28; SST-14, somatostatin-14; CST-17, human cortistatin-17.) aminoacid residues necessary for bioactivity are shown in bold. Octreotide and lanreotide are synthetic somatostatin analogs.

FIGURE 169-2. Schematic depiction of the rat somatostatin gene and its regulatory domains. The mRNA coding region consists of two exons of 238 and 367 base pairs (bp) separated by an intron of 621 bp. Located upstream (i.e., 5' end) from the start site of mRNA transcription (arrow) are the regulatory elements TATA, cyclic adenosine monophosphate response element (CRE), atypical glucocorticoid response element (aGRE), and somatostatin promoter silence element (SMS-PS). Tissuespecific elements (TSE) consisting of TATA motifs that operate in concert with CRE to provide high-level constitutive activity are shown. (3'UT, 3' untranslated region; SST-14, somatostatin-14; SST-28, somatostatin-28.)

ANATOMIC DISTRIBUTION OF SOMATOSTATIN CELLS Somatostatin-producing cells occur in high densities throughout the central and peripheral nervous systems, and in the endocrine pancreas and gut. They occur in smaller numbers in the thyroid, adrenal medulla, testes, prostate, submandibular gland, kidneys, and placenta1,8,10 (Table 169-1). The typical morphologic appearance of an SST cell is that of a neuron with multiple branching processes, or of a secretory cell, often with short cytoplasmic extensions (D cells). In the brain, the highest concentrations of SST are found in the hypothalamus, neocortex, and basal ganglia, throughout the limbic system, and at all levels of the major sensory systems.1,8,10,11 The approximate relative amounts of SST in the major regions of the brain are as follows: cerebral cortex, 49%; spinal cord, 30%; brainstem, 12%; hypothalamus, 7%; olfactory lobe, 1%; and cerebellum, 1%.11

TABLE 169-1. Localization of Somatostatin

SST cells in the pancreas are almost exclusively islet D cells and account for 2% to 3% of the total adult islet cell population.12 Gut SST cells are of two types: D cells in the mucosa and neurons that are intrinsic to the submucous and myenteric plexuses.13 In the thyroid, SST coexists with calcitonin in a subpopulation of C cells.1 In addition to these typical SST-producing neuroendocrine cells, which secrete large amounts of the peptide from storage pools, inflammatory and immune cells also produce SST, usually in small amounts on activation.14,15 In the rat, the gut accounts for ~65% of total body SST, whereas lesser amounts occur in the brain (25%), the pancreas (5%), and the remaining organs (5%). The relative proportions of SST-14 and SST-28 synthesized and secreted vary considerably in different tissues.4 SST-14 is the predominant form in the brain, pancreas, upper gut, and enteric neurons, whereas SST-28 is an important constituent of brain and is the predominant molecular form in the intestinal mucosa.

SOMATOSTATIN IN THE PLASMA AND OTHER BODY FLUIDS Both SST-14 and SST-28 are released readily from tissues and are detected in blood.1,4,8,16 The main source of circulating SST is the gastrointestinal tract.17 Circulating SST is inactivated rapidly by the liver and kidneys. The plasma half-life of SST-14 is 2 to 3 minutes, whereas that of SST-28 is slightly longer.4 Fasting plasma concentrations of SST-like immunoreactivity (SST-LI) range from 5 to 18 pmol/L. These levels double in response to the ingestion of a mixed meal.4 The bioactive circulating forms consist of SST-14, desAla1 SST-14 (a postsecretory conversion product of SST-14), and SST-28.16 With few exceptions, fluctuations in peripheral plasma levels of SST-LI are small. The main clinical utility of plasma measurements is in the diagnosis of SST-producing tumors, which are associated with marked hypersomatostatinemia. SST is secreted into the cerebrospinal fluid (CSF), probably from all parts of the brain.18,19 It is stable in this medium and attains a concentration that is approximately twice that in the general circulation. Significant amounts are also excreted in the urine (4–6 pmol/L). Semen contains high levels of SST-LI, 200-fold greater than those in plasma. Amniotic fluid is rich in SST-LI originating from the fetus.

REGULATION OF SOMATOSTATIN SECRETION AND GENE EXPRESSION Because SST cells are so widely distributed and interact with many different body systems, the fact that the secretion of SST can be influenced by a broad array of secretagogues, ranging from ions and nutrients to neuropeptides, neurotransmitters, classic hormones, and growth factors, is not surprising.1,2,4,8 Glucagon, GH-releasing hormone (GHRH), neurotensin, corticotropin-releasing hormone (CRH), calcitonin gene–related peptide (CGRP), and bombesin are potent stimulators of SST release, whereas opioids and g-aminobutyric acid (GABA) generally inhibit SST secretion.1,4,8 Of the various hormones studied, thyroid hormones enhance SST secretion from the hypothalamus; their effect on secretion from other tissues has not been adequately investigated.4,8 Glucocorticoids exert a dose-dependent biphasic effect on SST secretion; low doses are stimulatory and high doses are inhibitory.4 Insulin stimulates hypothalamic SST release but has an inhibitory effect on the release of SST from islet and gut.4 Finally, members of the growth factor–cytokine family such as GH, insulin-like growth factor-I (IGF-I), interleukin-1 (IL-1), tumor necrosis factor-a (TNF-a), and interleukin-6 (IL-6) are capable of stimulating SST secretion from brain cells.2 Many of the agents that influence SST secretion also regulate SST gene expression.2,4,8 Steady-state SST mRNA levels are stimulated by growth factors and cytokines (e.g., GH, IGF-I, IGF-II, IL-1, TNF-a, IL-6, interferon-g, and interleukin-10), glucocorticoids, testosterone, estradiol, and N-methyl-D-aspartate–receptor agonists; and are inhibited by insulin, leptin, and transforming growth factor-b (TGF-a). Among the intracellular mediators known to modulate SST gene expression are Ca2+, cAMP, cyclic guanosine monophosphate (cGMP), and nitric oxide (NO).2,4,8 Activation of the adenylate cyclase–cAMP pathway plays an important role in the stimulation of SST secretion and gene transcription.20 Cyclic AMP–dependent transcriptional enhancement is mediated by the nuclear protein cAMP response element–binding protein (CREB), which binds to the cAMP response element on the SST gene.20 Ca2+-dependent induction of the SST gene occurs through phosphorylation of CREB by the Ca2+-dependent protein kinase I and protein kinase II. GH, IGF-I, IGF-II, and glucocorticoids have all been shown to induce the SST gene by direct interaction with its promoter.4 The molecular mechanisms underlying the effects of estrogens, testosterone, cGMP, and NO on SST mRNA levels remain to be determined.4

ACTIONS OF SOMATOSTATIN SST not only has wide anatomic distribution, but also acts on multiple targets, including the brain, pituitary, endocrine and exocrine pancreas, gut, kidney, adrenal, thyroid, and immune cells1,2,7,8 (Fig. 169-3). Its actions include inhibition of virtually every known endocrine and exocrine secretion, and of various neurotransmitters; behavioral and autonomic effects if centrally administered; and effects on gastrointestinal and biliary motility, vascular smooth muscle tone, and intestinal absorption of nutrients and ions. SST also blocks the release of growth factors (e.g., IGF-I, epidermal growth factor [EGF], and platelet-derived growth factor [PDGF]) and cytokines (e.g., IL-6, interferon-g), and inhibits the proliferation of lymphocytes and of inflammatory, intestinal mucosal, and cartilage and bone precursor cells.2 All of these diverse effects of SST can be explained by its inhibition of two key cellular processes, secretion and cell proliferation.

FIGURE 169-3. Principal actions of somatostatin. Somatostatin inhibits the release of dopamine from the midbrain and of norepinephrine, thyrotropin-releasing hormone (TRH), corticotropin-releasing hormone (CRH), and endogenous somatostatin from the hypothalamus. It also inhibits both the basal and the stimulated secretion of growth hormone (GH), thyroid-stimulating hormone (TSH), and islet hormones. It has no effect on luteinizing hormone (LH), follicle-stimulating hormone (FSH), prolactin, or adrenocorticotropic hormone (ACTH) in normal subjects. It does, however, suppress elevated ACTH levels in Addison disease and in ACTH-producing tumors. In addition, it inhibits the basal and the TRH-stimulated release of prolactin in vitro and diminishes elevated prolactin levels in acromegaly. In the gastrointestinal tract, somatostatin inhibits the release of virtually every gut hormone that has been tested. It has a generalized inhibitory effect on gut exocrine secretion (gastric acid, pepsin, bile, colonic fluid) and suppresses motor activity generally, through inhibition of gastric emptying, gallbladder contraction, and small intestine segmentation. Somatostatin, however, stimulates migrating motor complex activity. The effects of somatostatin on the thyroid include inhibition of the TSH-stimulated release of thyroxine (T4) and triiodothyronine (T4). The adrenal effects consist of the inhibition of angiotensin II–stimulated aldosterone secretion and the inhibition of acetylcholine-stimulated medullary catecholamine secretion. In the kidneys, somatostatin inhibits the release of renin stimulated by hypovolemia and inhibits antidiuretic hormone (ADH)–mediated water absorption. (CCK, cholecystokinin; VIP, vasoactive intestinal peptide.) (Modified from Patel YC. General aspects of the biology and function of somatostatin. In: Weil C, Muller EE, Thorner MO, eds. Somatostatin. Basic and clinical aspects of neuroscience series, vol 4. Berlin:

Springer-Verlag, 1992:1.)

SOMATOSTATIN RECEPTORS, RECEPTOR SUBTYPES, AND SIGNAL TRANSDUCTION Somatostatin acts through high-affinity plasma membrane receptors that are pharmacologically heterogeneous and feature several different isoforms.2,3,5,6,21 Molecular cloning has revealed a family of five structurally related SSTR subtype genes that encode for seven transmembrane domain, G protein–coupled receptor proteins that display distinct agonist-binding profiles for natural and synthetic SST peptides2,3,21 (Table 169-2). receptor types 1 through 4 bind SST-14 and SST-28 approximately equally, whereas the type 5 receptor displays relative selectivity for binding of SST-28.2,3,21 Four of the genes (the exception is SSTR2) appear to have no introns. Each of the receptor genes is located on a separate chromosome. The mRNA for individual human SSTR subtypes is widely expressed in brain, pituitary, pancreatic islets, stomach, jejunum, colon, lung, kidney, and liver, with a characteristic tissue-specific pattern for each receptor.2,3,21,22 Typically, more than one subtype occurs in a given target tissue (e.g., SSTR1 through SSTR5 in the brain, stomach, pancreatic islets, and aorta; SSTR1, SSTR2, SSTR3, and SSTR5 in the pituitary). SSTR2 is the most abundantly expressed subtype, in terms of both the number of tissues that express this receptor and the level of expression. It is preferentially expressed by islet A cells and immune cells. SSTR1 and SSTR5 are the main subtypes expressed by islet B cells.23 SSTR2 and SSTR5 are the principal subtypes found in somatotropes.

TABLE 169-2. Characteristics of Cloned Human Somatostatin Receptor (hSSTR), Types

SST receptors elicit their cellular responses through G protein–linked modulation of multiple second messenger systems (Fig. 169-4), including (a) receptor coupling to adenylate cyclase, (b) receptor coupling to K+ channels, (c) receptor coupling to Ca+ channels, (d) receptor coupling to exocytotic vesicles, (e) receptor coupling to phosphotyrosyl protein phosphatase (PTP), and (f) receptor coupling to the mitogen-activated protein kinase (MAPK) pathway.2,3,21 The five receptors share common signaling pathways, such as the inhibition of adenylate cyclase, activation of PTP, and modulation of MAPK2 (see Table 169-2). SSTR2, SSTR3, SSTR4, and SSTR5 are coupled to K+ channels; SSTR1 and SSTR2, to voltage-dependent Ca+ channels; SSTR2 and SSTR5, to phospholipase C; and SSTR1, to a Na+ /H+ exchanger.2

FIGURE 169-4. Schematic depiction of somatostatin-receptor (SSTR)–signaling pathways leading to inhibition of secretion (A) and to cell proliferation and induction of apoptosis (B). Receptor activation leads to a fall in intracellular cyclic adenosine monophosphate (cAMP) (due to inhibition of adenylate cyclase), a fall in Ca2+ influx (due to activation of K+ and Ca2+ ion channels), and stimulation of phosphatases such as calcineurin, which inhibits exocytosis, and serine threonine (Ser/Thr) phosphatases, which dephosphorylate and activate Ca2+ and K+ channel proteins. Blockade of secretion by somatostatin (SST) is in part mediated through inhibition of Ca2+ and cAMP (proximal effect) and through a more potent distant effect involving direct inhibition of exocytosis via SST-dependent activation of calcineurin. Induction of protein tyrosine phosphatase by SST plays a key role in mediating cell growth arrest (via SSTR1, SSTR2, SSTR4, SSTR5), or apoptosis (via SSTR3). Cell growth arrest is dependent on activation of the mitogen-activated protein kinase (MAPK) pathway and induction of Rb (retinoblastoma tumor-suppressor protein), and p21 (cyclin-dependent kinase inhibitor). C-src, which associates with both the activated receptor and phosphotyrosyl protein phosphatase (PTP), may provide the link between the receptor, PTP, and the mitogenic signaling complex. Induction of apoptosis is associated with dephosphorylation-dependent activation of the tumor-suppressor protein p53 and the proapoptotic protein Bax. (From Patel YC. Somatostatin and its receptor family. Frontiers Neuroendocrinol 1999; 20:157.)

SOMATOSTATIN RECEPTOR–MEDIATED INHIBITION OF SECRETION AND CELL PROLIFERATION The pronounced ability of SST to block regulated secretion from many different cells is due in part to receptor-induced inhibition of two key intracellular mediators, cAMP and Ca+ , because of receptor-linked effects on adenylate cyclase and on K+ and Ca+ ion channels2 (see Fig. 169-4). In addition, SST inhibits secretion stimulated by cAMP, Ca+ ions, or any other known second messenger through a distal effect, which is targeted directly to secretory granules and is dependent on activation of the protein phosphatase calcineurin. 2,3,7 The antiproliferative effects of SST were recognized through the use of long-acting analogs (i.e., octreotide) for the treatment of hormone hypersecretion from pancreatic, intestinal, and pituitary tumors. SST not only blocked hormone hypersecretion from these tumors but also caused variable tumor shrinkage through an additional antiproliferative effect. The antiproliferative effects of SST have since been demonstrated in normal dividing cells (e.g., intestinal mucosal cells), in activated lymphocytes, andin inflammatory cells as well as in vivo in solid tumors and various cultured tumor cell lines.2 These effects involve cytostatic (growth arrest) and cytotoxic (apoptotic) actions. SST acts directly (via SSTRs present on tumor cells) and indirectly (via SSTRs present on nontumor cell targets) to inhibit the secretion of hormones and growth factors that support angiogenesis and promote tumor growth. Several SSTR subtypes and signal-transduction pathways have been implicated2,3,21 (see Fig. 169-4). Most interest is focused on protein phosphatases that dephosphorylate receptor tyrosine kinases or modulate the MAPK-signaling cascade, thereby attenuating mitogenic signal transduction.24,25 Four of the receptors (SSTR1, SSTR2, SSTR4, SSTR5) induce cell cycle arrest via PTP-dependent modulation of MAPK associated with induction of the retinoblastoma tumor-suppressor protein and p21.2 In contrast, SSTR3 uniquely triggers PTP-dependent apoptosis accompanied by activation of p53 and the proapoptotic protein Bax.26

PHYSIOLOGIC SIGNIFICANCE OF SOMATOSTATIN Evidence is growing, both direct and indirect, that SST modulates the physiologic function of various target cells.1,2,4,5,8 It subserves mainly local regulatory functions, acting as either a neurotransmitter or neuromodulator; a neurosecretory substance (i.e., one released directly from nerve axons into the bloodstream as in the median eminence); or a paracrine-autocrine regulator (local cell-to-cell interaction or self-regulation). In addition, SST may act via the circulation as a true endocrine substance to influence distant targets. Direct evidence exists of a physiologic role for hypothalamic SST in the regulation of GH and thyroid-stimulating hormone (TSH) secretion by the pituitary.1,8,27 A variety of physiologic GH responses are orchestrated by SST, acting either alone or in concert with GHRH.1,8,27 SST participates in the genesis of the normal pulsatile pattern of GH secretion and in GH regulatory responses to physiologic stimuli such as stress, glucose administration, or food deprivation. GHRH and SST neurons in the

hypothalamus are anatomically coupled and influence each other reciprocally.8,27 SST inhibits GH secretion both by a direct action on the pituitary, and indirectly through suppression of GHRH release (Fig. 169-5). The secretion of SST in turn is stimulated by GHRH and is subject to positive feedback regulation by GH (short loop) and by IGF-I produced by GH action on the liver (long loop). Because of the extensive extrahypothalamic brain distribution of SST, its effects on the spontaneous electrical activity of neurons, its release from nerve endings in response to depolarization, and its behavioral effects, this peptide has been postulated to serve as a central neurotransmitter or neuromodulator. 1,5 Given the high concentration of both SST neuronal elements and SSTRs in limbic, neocortical, striatal, and sensory areas, SST appears to be particularly important in modulating functions in these regions. Within the pancreatic islets, the close anatomic proximity of SST cells to the A and B cells, the demonstration that insulin and glucagon are exquisitely sensitive to inhibition by low concentrations of SST, and the finding that inactivation of islet D-cell function augments insulin and Glucagon output, all provide evidence for the possible modulation of pancreatic islet A-and B-cell function by SST4,8 (Fig. 169-6). The suggestion has been made that islet D cells, which produce predominantly SST-14 and only negligible amounts of SST-28, regulate A cells by local action, whereas SST-28 released in the circulation from the gut in response to food ingestion modulates nutrient-stimulated insulin release by a hemocrine mechanism.28 The diffuse distribution of SST throughout the gut, together with its pleiotropic effects and the complex regulation of its secretion, suggests that SST exerts control over many discrete cell systems involved in gastrointestinal, pancreatic, and biliary functions. SST regulates acid secretion both directly through the circulation to inhibit parietal cells and through a paracrine mechanism to suppress gastrin release.29 Circulating SST is also a physiologic regulator of pancreatic exocrine secretion.30 Elsewhere in the gut, evidence exists to suggest that SST controls the rate of absorption of nutrients and participates in the regulation of gut hormone secretion, gastrointestinal motor tone, blood flow, and mucosal cell proliferation.31 Although acute changes in tissue or circulating levels of SST are accompanied by alterations in the function of target organs (e.g., the pituitary or gut), an interesting finding is that SST deficiency from birth (as in the SST knock-out mouse) does not produce any developmental defect or growth abnormality in young animals.32 This means either that the SST gene is redundant or, more likely, that adaptive responses occur from other genes (e.g., cortistatin), which compensate for the loss of the SST gene.

FIGURE 169-5. Schematic representation of the interaction between somatostatin (SST), growth hormone–releasing hormone (GHRH), growth hormone (GH), and insulin-like growth factor-I (IGF-I) in regulating GH secretion. GH release is stimulated by GHRH (produced by GHRH neurons in the arcuate nucleus) and inhibited by SST (produced by somatostatinergic neurons in the anterior hypothalamic periventricular nucleus [PVN]). SST inhibits GH secretion both by direct action at the pituitary level and indirectly through suppression of GHRH release. GHRH, in turn, stimulates SST secretion. GH exerts negative feedback on its own secretion by inhibiting GHRH release, stimulating SST release, and potentiating the release of IGF-I from the liver. IGF-I, in turn, stimulates SST release and inhibits GH secretion by a direct action on the pituitary. (Modified from Patel YC. General aspects of the biology and function of somatostatin. In: Weil C, Muller EE, Thorner MO, eds. Somatostatin. Basic and clinical aspects of neuroscience series, vol 4. Berlin: Springer-Verlag, 1992:1.)

FIGURE 169-6. Effects of endogenous or exogenous somatostatin (SST), insulin, and glucagon on the function of pancreatic islet cells. SST inhibits insulin and glucagon release, glucagon stimulates insulin and SST release, and insulin inhibits the release of glucagon and possibly of SST. In addition, all three islet hormones inhibit their own secretion by an autocrine mechanism. Physiologically, intraislet insulin and glucagon regulate the secretion of SST, and intraislet insulin regulates glucagon release. The precise physiologic role of intraislet SST remains unclear (see text for details). (Modified from Patel YC. General aspects of the biology and function of somatostatin. In: Weil C, Muller EE, Thorner MO, eds. Somatostatin. Basic and clinical aspects of neuroscience series, vol 4. Berlin: Springer-Verlag, 1992:1.)

SUBTYPE-SELECTIVE BIOACTIONS OF SOMATOSTATIN RECEPTORS Because SST exerts its numerous bioeffects through five receptors, the question arises whether a given response is selective for one subtype or whether multiple subtypes are involved. The marked overlap in the cellular pattern of expression of the different SSTR pathways, coupled with the finding that individual target cells typically express multiple SSTR subtypes and often all five isoforms in the same cell, suggests that SSTRs may operate in concert rather than as individual members. Nonetheless, evidence exists for relative subtype selectivity for some SSTR effects. At the level of cell secretion and cell proliferation, the two general cellular effects modulated by SST, four of the subtypes (SSTR1, SSTR2, SSTR4, SSTR5) are capable of arresting cell growth, whereas SSTR3 is uniquely cytotoxic. In contrast to the case for cell proliferation, surprisingly little is known about subtype selectivity, if any, for cell secretion. Immune, inflammatory, and neoplastic cells are important targets for SST action.2,14,15 Unlike the classic SST-producing neuroendocrine cells (e.g., in the hypothalamus or islets), which release large quantities of the peptide acutely from storage pools, SST and SST receptors in inflammatory cells (e.g., macrophages and lymphocytes) are coinduced, probably by growth factors and cytokines, as part of a general mechanism for activating the endogenous SST system for paracrine-autocrine modulation of the proliferative and hormonal responses associated with inflammatory and immune reactions. SSTR2, the main isotype expressed in lymphocytes and inflammatory cells, appears to be the functional SSTR responsible for modulating the proliferative and secretory responses of these cells.15 Based on the pattern of SSTR subtype expression as well as the effects of selective nonpeptide agonists, SSTR2 and SSTR5 are the subtypes involved in regulating GH and TSH secretion from the human pituitary.2,33 Similar studies in the case of islet hormones suggest a preferential effect of SSTR2 on glucagon release and of SSTR5 and possibly SSTR1 on insulin secretion.23,34 Despite the pharmacologic evidence for the involvement of SSTR2 in pituitary GH secretion, an SSTR2-deficient mouse shows only subtle abnormalities of neuroendocrine GH feedback control. The animals grow normally both in utero and postnatally, a finding that suggests maintenance of overall GH secretion.35 These animals display high basal gastric acid secretion in the face of normal gastrin levels, indicating that SSTR2 is the subtype responsible for SST suppression of endogenous gastric acid.

SOMATOSTATIN IN DISEASE Given the wide distribution of SST cells in the body, the fact that only a single disease—the somatostatinoma syndrome—has been attributed directly to SST dysfunction is surprising (see Chap. 220). No functional mutations of either SST or its receptor genes have been identified. Even in this syndrome, the associated profound hypersomatostatinemia is accompanied by relatively minor symptoms (cholelithiasis, steatorrhea, and mild diabetes).36 This probably is due to tachyphylaxis and to the fact that many of the target cells on which SST normally acts locally are not accessible to circulating SST. Most somatostatinomas are actively secreting, malignant islet cell tumors that are associated with high plasma levels of SST-LI (600–15,000 pmol/L).36 Lesser degrees of hypersomatostatinemia have been observed in nonpancreatic SST-producing tumors, such as duodenal somatostatinoma, medullary thyroid carcinoma, pheochromocytoma, extraadrenal paraganglioma, and small cell cancer of the lung. The serial measurement of plasma SST-LI values has proved useful as a tumor marker in the follow-up of these patients. In several diseases, disordered SST function occurs probably as a secondary feature. Foremost among these is Alzheimer disease, in which a decrease in the levels of SST in the cerebral cortex and CSF is seen.5,19 The reduction in cortical SST correlates with the number of senile plaques and neurofibrillary tangles, and although its pathophysiologic significance is unclear, it has become an important biochemical marker for the disease. Cere-brocortical SST, CSF SST, or both are also decreased

in other neuropsychiatric disorders such as depression, Parkinson disease, and multiple sclerosis.19 Whereas the reduction in brain or CSF SST in Alzheimer disease appears to be secondary to neurodegeneration, the SST changes that occur in depression and multiple sclerosis fluctuate with disease activity and reflect functional alterations in peptide production. In contrast to the loss of SST in Alzheimer disease, SST neurons in the striatum in Huntington disease are selectively resistant to neurodegeneration and show up-regulated function. Selective survival and up-regulation of SST gene expression in response to neuronal injury has also been demonstrated in the striatum in animal models of hypoxiaischemia and in the cortex of monkeys and human patients with AIDS encephalopathy. Plasma SST levels are elevated significantly in hepatic cirrhosis and in chronic renal failure, an elevation that reflects impaired metabolism. In experimental hyperinsulinemic diabetes and in human type 2 diabetes, release of gut SST in response to meal ingestion is impaired. Patients with duodenal ulcers have reduced antral SST-LI levels, implicating local deficiency of SST in the pathogenesis of this disorder. Despite the large amounts of SST present in the gut, primary gastrointestinal disease generally is not associated with alterations in circulating SST concentrations.

CLINICAL APPLICATIONS OF SOMATOSTATIN AND ITS ANALOGS The potent pharmacologic properties of SST have attracted much interest in the use of this substance as a therapeutic agent for the treatment of various diseases. The naturally occurring forms proved unsuitable, however, because their short halflives made continuous intravenous administration necessary. The specificity of endogenous SST derives from the fact that it is produced mainly at local sites of action and is rapidly inactivated after release by peptidases in tissue and blood, so that unwanted systemic effects are minimized. Injections of synthetic SST, on the other hand, produce a wide array of effects due to simultaneous activation of multiple target sites. Thus, for effective pharmacotherapy with SST, analogs must be designed that have more selective actions and greater metabolic stability than the naturally occurring peptides. Early observations that the SST-14 molecule was amenable to a wide range of chemical modifications allowed the synthesis of structural analogs with enhanced metabolic stability. One such analog is the long-acting cyclic octapeptide, octreotide acetate6,37 (Sandostatin; see Fig. 169-1). Virtually 100% of this drug is absorbed after subcutaneous administration, and it is eliminated from plasma with a half-life of 70 to 113 min.37 The bioresponse is maintained for 8 or more hours after a single injection. Octreotide thus has therapeutic efficacy with two to three daily subcutaneous (sc) injections and has emerged as the first SST analog suitable for long-term treatment.38 Long-acting slow-release formulations of both octreotide (Sandostatin LAR [long-acting release]) and a second octapeptide analog lanreotide (BIM23014, Somatuline) have become available.37,39,40,41 and 42 Sandostatin LAR incorporates octreotide into microspheres of a biodegradable polymer (polyDL-lactide-co-glycolide glucose), allowing slow release of the drug by cleavage of the polymer ester linkage through hydrolysis in tissues.37,39 It is administered as a once-a-month intramuscular (im) depot injection and, after a lag period of 7 to 10 days, produces stable blood octreotide concentrations comparable to that of continuous infusion.39 Lanreotide is available only as the slow-release formulation, and because of its comparatively shorter duration of action (10–14 days), it must be injected two to three times per month.40,41 Both analogs are at least as effective as sc octreotide in blocking the excessive production of hormones from neuroendocrine tumors of the gastrointestinal tract, pancreatic islets, and pituitary.37,38,39,40,41 and 42 Furthermore, they have a safety profile similar to that of the sc injections, and, because of better patient compliance, are likely to become the drugs of choice for SST pharmacotherapy. Both the subcutaneous and LAR forms of octreotide are approved by the Food and Drug Administration for the treatment of carcinoid tumors, VIPomas, and acromegaly. Lanreotide should shortly be available in North America. Octreotide and lanreotide bind to only three of the five SSTRs (types 2, 3, and 5) (see Table 169-2), displaying high affinity for subtypes 2 and 5, moderate affinity for subtype 3, and no binding to types 1 and 4.2,3,6,21 The binding affinity of these analogs for subtypes 2 and 5 is comparable to that of SST-14, indicating that they are neither selective for these subtypes nor more potent than endogenous SST. A series of high-affinity nonpeptide agonists have been identified for several of the human SSTRs. These should facilitate the development of orally active subtype-selective therapeutic compounds.34 TREATMENT OF NEUROENDOCRINE AND NONNEUROENDOCRINE TUMORS Octreotide provides potent palliative therapy for hormone-producing neuroendocrine tumors (Table 169-3). It acts in two ways to combat the effects of hormone hypersecretion: (a) directly on tumor cells to inhibit secretion, and (b) indirectly to block the action of the hypersecreted hormone at its target site. In addition, octreotide may cause tumor shrinkage or stabilization of tumor growth in some instances by shrinkage of tumor-cell volume through long-term inhibition of secretion (comparable to the shrinkage of prolactinomas induced by dopamine agonists) and inhibition of tumor growth via cytostatic and cytotoxic (apoptotic) effects (see Fig. 169-4).

TABLE 169-3. Proven and Potential Clinical Applications of Somatostatin Analogs

Binding studies have shown that neuroendocrine as well as common solid nonneuroendocrine tumors are rich in SSTRs (Table 169-4).3,38 Currently available binding analyses, however, cannot distinguish the individual SSTR subtypes because of the lack of subtype-selective radioligands. Accordingly, investigators have resorted to mRNA analysis and receptor immunocytochemical analysis with subtype-selective antibodies to characterize the pattern of expression of the five SSTR subtypes. Investigation of mRNA expression for the five SSTRs in >100 pituitary tumors, both secretory (producing GH, TSH, prolactin, or ACTH) and nonsecretory (i.e., chromophobe adenomas), and in 32 gastroenteropancreatic tumors (i.e., carcinoid, insulinoma, glucagonoma) has revealed multiple SSTR genes in most tumors.3,43,44 and 45 SSTR1 and SSTR2 appear to be the predominant forms in all tumors. Pituitary adenomas are also rich in the expression of SSTR5. Many nonneuroendocrine tumors (e.g., breast and renal carcinomas) are also SSTR positive.43,45 The extent of malignant disease as well as differential tumor expression of SSTRs may account for the variable clinical response observed with different tumor types. The presence of octreotide-sensitive subtypes, such as SSTR2 and SSTR5, in the majority of neuroendocrine tumors correlates with the responsiveness of these tumors to treatment with the analog.

TABLE 169-4. Expression of Somatostatin Receptors (SSTRs) in Tumors In Vitro and In Vivo

Treatment is indicated in patients with severe symptoms and resultant metabolic derangements that require control before surgery or other therapy; in patients with residual tumor or metastatic endstage disease who have had relapse after standard therapeutic modalities (e.g., surgery, chemotherapy, and hepatic artery embolization); and as prophylaxis against acute crises resulting from sudden discharge of tumor products, such as in carcinoid crisis. In these instances, octreotide administration has produced extended and useful symptomatic remissions. The initial dosage may be 50 to 100 µg of sc octreotide twice daily or 10 mg im of octreotide LAR per month. The dosage may be increased gradually over 6 to 12 months (or as required) to 500 µg three times daily of sc octreotide and 30 mg per month of the

LAR preparation. Generally, clinical improvement parallels a reduction in the plasma level of the hormone being hypersecreted by the tumor. GASTROENTEROPANCREATIC TUMORS CARCINOID TUMORS SST analogs play a central role in the management of symptoms of metastatic carcinoid disease. More than 200 cases of metastatic carcinoid tumors and carcinoid syndrome treated with octreotide have now been reported.6,38,42,46,47,48 and 49 The drug is highly effective in this condition and produces a marked clinical and biochemical improvement. Flushing and diarrhea, the two most prominent symptoms, are rapidly relieved in >90% of patients. The clinical improvement is paralleled by a reduction in urinary 5-hydroxyindoleacetic acid (5-HIAA) levels, which drop by >50% in 70% of treated patients without, however, being completely normalized in any patient. A small proportion of patients (~14%) show measurable regression of hepatic and lymph node metastases.46,49 In 80 patients treated with octreotide, median survival from the diagnosis of metastatic disease was 8.8 years compared with 1.8 years in historical controls. The average dose required for symptomatic and biochemical control ranges from 100 to 150 µg sc every 8 hours. Because of the dual actions of SST in blocking both the secretion and the action of hormones over-produced by tumors, the dose required to control symptoms such as diarrhea may be lower than that necessary for reducing urine 5-HIAA levels. Resistance to therapy eventually occurs with an increase in tumor bulk. Such resistance may be due to downregulation of SST receptors or, more likely, to the emergence of receptor-negative or more virulent clones. Recurrence of disease on therapy may be controlled in some but not all patients by increasing the dose. Doses as high as 1 to 2 mg sc every 8 hours have been administered in this disease and appear to be well tolerated. A study comparing sc and long-acting formulations of octreotide found that monthly injections of octreotide LAR were as effective as sc octreotide in controlling the symptoms and in suppressing urinary 5-HIAA levels in 79 patients.42 The recommended average dose is 20 mg octreotide LAR, increasing to 30 to 60 mg in some patients with resistant or advanced disease. Lanreotide 30 mg im every 10 to 14 days is also effective.41 In addition to its application in long-term treatment of metastatic carcinoid disease, octreotide is very useful in preventing carcinoid crisis when administered intravenously in susceptible patients immediately before and for 7 to 10 days after surgery, chemotherapy, or hepatic artery embolization. VIPOMAS Like carcinoid tumors, VIPomas tend to be malignant, with extensive metastatic disease from the outset. Octreotide has become the first-line drug for the symptomatic control of this disease. Of 25 patients with VIPoma treated with octreotide for 15 months, 85% showed improvement in or resolution of diarrhea, which was associated with a reduction in plasma levels of vasoactive intestinal peptide in 60%.47,48 and 49 The effect of octreotide is usually rapid and occurs with relatively low doses, 50 to 100 µg every 8 to 12 hours. With time, efficacy is lost in some patients and high-dose regimens are required. Reduction in tumor size has been reported in as many as 40% of patients treated with octreotide for 2 months or longer.49 GLUCAGONOMAS Approximately 10 patients with glucagonomas who have been treated with octreotide have been described.47,48 and 49 Glucagonomas are slowly growing, malignant tumors associated with severe constitutional symptoms as well as a characteristic rash (see Chap. 220). Therapy with octreotide resolves the rash in a few days and improves other symptoms such as weight loss, anorexia, abdominal pain, and diarrhea. Because no other form of therapy exists for the systemic manifestations of glucagonoma, especially its dermatologic complication, octreotide has assumed a primary role in the medical treatment of this disorder. Usually little effect is seen on the size of the tumor. INSULINOMAS Unlike non–B-cell tumors, 90% of insulinomas are benign and can be cured by surgical resection. Octreotide suppresses insulin levels by >50%, elevates blood glucose levels, and prevents hypoglycemic attacks in many patients.47,48 and 49 Thus, it is effective in the preoperative treatment of patients with benign insulinoma. Diazoxide, however, also lowers insulin levels effectively in this setting. It remains to be determined whether octreotide offers any specific advantage. In most patients with malignant insulinoma, octreotide is effective in preventing hypoglycemic attacks. Loss of efficacy is common, however, and it may be overcome by increasing the dosage. Of greater concern is the worsening of hypoglycemia in some patients as a result of the greater suppression by octreotide of the counterregulatory hormones (GH and glucagon) than of insulin. In view of this possibility, plasma glucose levels should be monitored carefully at the initiation of octreotide therapy. GASTRINOMAS Gastrinomas are commonly malignant islet cell tumors that have metastasized at the time of diagnosis. In the >50 reported cases treated with octreotide, the response has been variable, with only modest reductions in gastrin levels.47,48 and 49 Furthermore, tumor growth may progress during octreotide treatment. Because the clinical manifestations of gastrinoma result from the hypersecretion of gastric acid, which can be controlled effectively by oral medications such as H2-receptor antagonists or the Na+ /H+ adenosine triphosphatase inhibitor omeprazole, octreotide serves only an adjunctive role in treatment of this disorder. TUMORS PRODUCING GROWTH HORMONE–RELEASING HORMONE Several patients with GHRH-producing tumors (bronchial carcinoid tumors or metastatic islet cell tumors) who received treatment with octreotide showed good symptomatic and biochemical response.49 Octreotide reduced plasma concentrations of GHRH and GH, probably through a dual effect on tumor cells and pituitary somatotropes. TREATMENT OF PITUITARY TUMORS ACROMEGALY SST analogs significantly reduce GH secretion and IGF-I levels in patients with acromegaly and have become effective agents in the treatment of this disorder.6,37,38,39,40,47,49,50,51 and 52 Their use is indicated in the treatment of patients with a macroadenoma who have persistent disease after transsphenoidal surgery, as interim treatment in patients awaiting the full effects of external irradiation, and as preoperative treatment for 2 to 3 months to improve the medical condition of patients with severe disease. They can also be offered as primary therapy to patients who refuse surgery or those with severe medical problems that preclude surgery. Octreotide has been proposed as primary therapy for patients with invasive macroadenomas not causing chiasmatic compression, who are unlikely to be cured surgically.51 This is an attractive idea, especially with the availability of slow-release preparations; it needs to be further assessed in prospective studies comparing the relative efficacy and cost effectiveness of primary SST analog therapy versus surgery for treatment of acromegaly.51 With sc injections of 100 µg octreotide three times daily (mean optimal efficacious dosage), a prompt reduction in plasma GH levels occurs, followed by a decline in circulating IGF-I levels during the first week.50,52 This is accompanied by immediate relief of many of the clinical symptoms of acromegaly, including fatigue, headache, excessive perspiration, arthralgia, and soft-tissue swelling.50,52 Relief of headache is often seen immediately after the injections are started and before serum GH levels have declined, perhaps as a result of a vascular or analgesic action of the drug. Improvement in facial coarsening and acral enlargement occurs after longer periods of treatment. Sustained improvement with minimal side effects has been noted with treatment for 10 or more years. Drug resistance may develop during long-term treatment, necessitating increased dosages. In a double-blind, randomized, multicenter study of 115 patients with acromegaly,52 administration of octreotide at a dosage of 100 µg three times a day for 6 months was effective in reducing GH levels in 71% of patients and IGF-I levels in 93%. Mean integrated GH levels were reduced 70% from 40 ± 13 µg/L to 12 ± 4 µg/L, and mean integrated IGF-I levels were suppressed 60% from 4800 ± 354 U/L to 1900 ± 200 U/L. Normalization of GH secretion, defined as a reduction of integrated mean GH levels to 3000 patients. These currently ongoing studies will look at the effect of high-dose octreotide given alone as a monthly injection for 2 years or in combination with tamoxifen to women with estrogen receptor–positive stage I and stage II breast cancer. GASTROINTESTINAL APPLICATIONS Short- or long-term octreotide administration is useful in treating numerous gastrointestinal disorders (see Table 169-3). These conditions are more common than neuroendocrine tumors and account for much of the hospital-based usage of octreotide. SST is a potent constrictor of the splanchnic circulation and is effective in controlling variceal bleeding. In a randomized, double-blind, placebo-controlled trial of 120 patients, infusion of SST-14 at a dosage of 250 µg per hour for 5 days controlled bleeding and reduced transfusion requirements.58 In a study of 100 patients, octreotide given intravenously for 48 hours was as effective as emergency sclerotherapy in the treatment of acute variceal hemorrhage.6 The antisecretory effects of SST on the pancreas have led to the successful use of octreotide in the treatment of pancreatic fistulas. More than 60 patients so treated have been described.59 Most reports indicate a closure rate of pancreatic fistulas approximating 70% within a week. In addition to the diarrhea that occurs with hormone-secreting tumors, other forms of secretory diarrhea such as that associated with high-output ileostomy, diabetic diarrhea, AIDS diarrhea, chemotherapy-induced diarrhea, and diarrhea associated with amyloidosis all have been reported to be variably controlled with octreotide.47 In these diarrheal states, octreotide acts by blocking the normal production of gastric and pancreatic exocrine secretion destined for reabsorption in the distal bowel, and by directly inhibiting intestinal fluid and electrolyte secretion. In a randomized double-blind trial involving 10 patients with severe postgastrectomy dumping syndrome, the administration of octreotide 100 µg 30 minutes before eating prevented the development of vasomotor symptoms (early dumping) as well as the hypoglycemia and diarrhea characteristic of late dumping.60 Finally, in patients with scleroderma and intestinal dysmotility, the short-term administration of octreotide improved intestinal motility by stimulating the frequency of intestinal migrating motor complexes and reducing bacterial overgrowth, leading to an improvement in abdominal symptoms such as nausea, bloating, and pain.61 ORTHOSTATIC HYPOTENSION Postprandial and orthostatic hypotension in patients with autonomic neuropathy is abolished by low dosages of octreotide (0.2–0.4 µg/kg). In these cases, the drug acts as a potent pressor agent and has emerged as a new form of therapy for this condition as well as for other types of postprandial hypotension, such as that which occurs in the elderly.62

SIDE EFFECTS Although the acute administration of SST produces a large number of inhibitory effects, continued exposure results in an escape from the acute effects of the peptide.2,6 This may be due to up-regulation of some of the receptor subtypes, which compensates for the desensitized responses of others and thereby maintains normal overall SST responsiveness. Side effects associated with the long-term administration of octreotide for up to 2 years have been remarkably few.6,37,39,40,41,42,49,50,51 and 52 Most patients complain of pain at the injection site, and abdominal cramps and mild steatorrhea to which they become tolerant after 10 to 14 days. No nutritional deficiency has been reported with long-term octreotide therapy. Patients also adapt rapidly to some of the other effects of SST (e.g., inhibition of insulin and TSH secretion). A mild impairment of glucose tolerance may occur, consisting of a minimal increase in postprandial glucose levels only, with maintenance of normal fasting glucose levels. normal thyroid function is maintained and hypothyroidism is rare. Some effects, however, do persist; for example, inhibition of gallbladder emptying causes a significant increase in the incidence of biliary sludge and cholesterol gallstones in 20% to 30% of patients after 1 to 2 years of treatment.6 Those receiving long-term treatment warrant initial ultrasonographic evaluation, and subsequently evaluation as necessary, depending on symptoms. The incidence of adverse effects with the slow-release preparations of SST is comparable to that with the subcutaneous injections.37,39,40,41 and 42 Interestingly, the persistent steatorrhea and mild diabetes mellitus that are associated with the chronic hyper-somatostatinemia in patients with SST-producing tumors36 are not observed in patients during long-term treatment with the slow-release forms of the SST analogs, despite the sustained high blood levels of the drugs. This no doubt reflects the narrower binding specificity of the synthetic SST analogs, which bind to only three of the SSTR subtypes, compared with SST-14 and SST-28, which are typically produced by somatostatinomas and bind to all five SSTRs (see Table 169-2). Although normal tissues adapt to the long-term effects of octreotide, hormone-producing tumors continue to respond to octreotide injections with persistent suppression of hormone secretion, frequently for several years. This suggests a differential regulation of SSTRs in normal tissues and in tumors. Tumors express a higher density of SSTRs than do surrounding normal tissues. Conceivably, SSTRs in tumors behave differently due to a loss of normal receptor regulatory function, due to an alteration in the pattern and composition of the various subtypes expressed, or due to abnormal receptor signaling.

SOMATOSTATIN-RECEPTOR SCANS A method for the in vivo imaging of SSTR-positive tumors and their metastases has been developed using as radionuclides [123I]Tyr3 octreotide or an indium-labeled octreotide preparation that is now available commercially (111In-diethylenetriamine pentaacetic acid[DTPA]-D-Phe-1-octreotide).37,38,63,64 The rationale for the method lies in the demonstration that most endocrine tumors that respond to octreotide therapy do so because they are rich in SSTRs. Direct binding studies of surgically removed specimens have revealed a higher density of receptors in tumor tissue than in the surrounding healthy tissue.38 After intravenous administration, the radioligand binds to SSTRs on primary and metastatic tumor cells, which then can be visualized by computerized g-scintigraphy. Unbound radioactive material is rapidly degraded in the liver and excreted in the bile, or eliminated through the kidneys. The technique is relatively simple and effective. Among 59 patients with metastatic carcinoid disease or pancreatic endocrine tumors, primary tumors and their metastases were successfully localized in all but five cases (see Table 169-4). Particularly impressive is the ability of the receptor-scanning technique to visualize many unsuspected metastases that escape detection by clinical examination and conventional imaging techniques such as computed tomographic scanning and magnetic resonance imaging.38,63 Overall, the specificity of the method is high as judged by the close correlation among the in vivo tumor receptor scans, the inhibitory effect of octreotide on the secretion of hormone by the tumor, and the presence of SSTRs as demonstrated by direct in vitro binding studies of surgically removed tumor samples.38,63 SSTR scanning offers several distinct advantages over the currently used standard imaging methods. The first is its high power of resolution for tumors as small as 1 cm in diameter, which are difficult to visualize with conventional scanning techniques.63 Second, because the whole body is imaged, abnormalities in areas not under clinical suspicion can be detected, and the full extent of metastatic disease accurately mapped. Third, this method provides a functional index of the SSTR status of tumor cells that could be useful in tumor management. For instance, positivity for receptors can predict tumor responsiveness to SST therapy and may serve as a prognostic index of a favorable outcome for some tumors. The technique has some limitations. The first is that secretory tumors that do not express SSTRs are not amenable to analysis with this method. Second, because octreotide does not bind to two of the SSTR subtypes (types 1 and 4), those tumors that uniquely express these receptor isoforms will escape detection. SST-receptor scans are now available at most major centers and are widely used in the assessment of carcinoid and islet cell tumors, paragangliomas, pheochromocytomas, and medullary thyroid carcinoma.38 They are of little value in the imaging of pituitary tumors, which are better defined by conventional techniques. Because of the rich expression of the type 2 SSTR in immune and inflammatory cells, lesions of several granulomatous, inflammatory, and immune disorders have been successfully imaged by SSTR scintigraphy, for example, rheumatoid joints, sarcoid lesions, Hodgkin and non-Hodgkin lymphomas, and thyroid eye disease. The value of

SST-receptor scan in the routine assessment of these disorders, however, remains to be determined.

SOMATOSTATIN RECEPTOR–TARGETED RADIOABLATION OF TUMORS The property of rich SSTR expression by tumors which allows their visualization by the receptor-imaging approach can be further applied to achieve receptor-targeted ablation of the tumors. Binding of radioligand to SSTRs on the cell surface triggers their internalization.2 Specifically, SSTR subtypes 2, 3, and 5, which bind octreotide, all undergo agonist-dependent internalization, suggesting that the use of octreotide-based ligands coupled to a-or b-emitting radioisotopes could provide a method for delivering targeted radiation to the interior of the cell. Preliminary results with 111In-DTPA-D-Phe-1-octreotide in several patients with inoperable metastatic islet and carcinoid tumors suggest partial tumor responses as determined by radiographic tumor shrinkage.65 A second analog of octreotide labeled with yttrium-90 (90Y-DOTA-D-Phe1 -Tyr3 octreotide), in which the DTPA molecule is replaced by another chelator, DOTA (tetraazacyclododecane tetraacetic acid), which is a more potent b-emitting isotope that has been reported to be highly effective in inducing radionecrosis of human small cell lung tumor transplanted into nude mice, is undergoing clinical trials.37 Because SSTRs are expressed in many normal cells that would undoubtedly take up the radioligands, the question of whether significant damage to organs (e.g., the pituitary, islets, and gut) occurs with this type of treatment remains to be determined. Nonetheless, this is a promising new approach that could be applied not just to neuroendocrine tumors but also to all SSTR-rich tumors, and it awaits further study. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.

Reichlin S. Somatostatin. N Engl J Med 1983; 309:1495. Patel YC. Somatostatin and its receptor family. Frontiers Neuroendocrinol 1999; 20:157. Patel YC. Molecular pharmacology of somatostatin receptor subtypes. J Endocrinol Invest 1997; 20:348. Patel YC, Liu JL, Galanopoulou AS, Papachristou DN. Production, action, and degradation of somatostatin. In: Jefferson LS, Cherrington AD, eds. The handbook of physiology: the endocrine pancreas and regulation of metabolism. New York: Oxford University Press, 2000. Epelbaum J, Dournaud P, Fodor M, Viollet C. The neurobiology of somatostatin. Crit Rev Neurobiol 1994; 8:25. Lamberts SWJ, Van Der Lely A-J, de Herder WW. Drug therapy: octreotide. N Engl J Med 1996; 334:246. Patel YC, Tannenbaum GS, eds. Somatostatin: basic and clinical aspects. Metabolism 1990; 39(Suppl 2). Patel YC. General aspects of the biology and function of somatostatin. In: Weil C, Muller EE, Thorner MO, eds. Somatostatin. Basic and clinical aspects of neuroscience series, vol 4. Berlin: Springer-Verlag, 1992:1. De Lecea L, Criado JR, Prospero-Garcia O, et al. A cortical neuropeptide with neuronal depressant and sleep-modulating properties. Nature 1996; 381:242. Hokfelt T, Efendic S, Hellerstrom C, et al. Cellular localization of somatostatin in endocrine-like cells and neurons of the rat with special reference to the A 1-cells of the pancreatic islets and to the hypothalamus. Acta Endocrinol (Copenh) 1975; 80 (Suppl 200):5. Patel YC, Reichlin S. Somatostatin in hypothalamus, extrahypothalamic brain and peripheral tissues of the rat. Endocrinology 1978; 102:523. Baetens D, Malaisse-Lagae F, Perrelet A, et al. Endocrine pancreas: three dimensional reconstruction shows two types of islets of Langerhans. Science 1979; 206:1323. Larsson LI. Distribution and morphology of somatostatin cells. Adv Exp Biol Med 1985; 188:383. Aguila MC, Dees WL, Haensly WE, McCann SM. Evidence that somatostatin is localized and synthesized in lymphoid organs. Proc Natl Acad Sci U S A 1991; 88:11485. Elliott DE, Blum AM, Li J, et al. Preprosomatostatin messenger RNA is expressed by inflammatory cells and induced by inflammatory mediators and cytokines. J Immunol 1998; 160:3997. Shoelson SE, Polonsky KS, Nakabayashi T, et al. Circulating forms of somatostatin-like immunoreactivity in human plasma. Am J Physiol 1986; 250:E428. Taborsky GT Jr, Ensinck JW. Contribution of the pancreas to circulating somatostatin-like immunoreactivity in the normal dog. J Clin Invest 1984; 73:216. Patel YC, Rao K, Reichlin S. Somatostatin in human cerebrospinal fluid. N Engl J Med 1977; 296:529. Rubinow DR, Davis CL, Post RM. Somatostatin in neuropsychiatric disorders. In: Weil C, Muller EE, Thorner MO, eds. Somatostatin. Basic and clinical aspects of neuroscience series, vol 4. Berlin: Springer-Verlag, 1992:29. Montminy M, Brindle P, Arias J, et al. Regulation of somatostatin gene transcription by cAMP. In: Somatostatin and its receptors. Ciba Foundation Symposium 190. West Sussex: John Wiley and Sons, 1995:7. Reisine T, Bell GI. Molecular biology of somatostatin receptors. Endocr Rev 1995; 16:427. Bruno JF, Yun XU, Song J, et al. Tissue distribution of somatostatin receptor subtype messenger ribonucleic acid in the rat. Endocrinology 1993; 133:2561. Kumar U, Sasi R, Suresh S, et al. Subtypeselective expression of the five somatostatin receptors (hSSTR1-5) in human pancreatic islet cells: a quantitative double-label immunohistochemical analysis. Diabetes 1999; 48:77. Pan MG, Florio T, Stork PJC. G protein activation of a hormone-stimulated phosphatase in human tumor cells. Science 1992; 256:1215. Reardon DB, Dent P, Wood SL, et al. Activation in vitro of somatostatin receptor subtypes 2, 3, or 4 stimulates protein tyrosine phosphatase activity in membranes from transfected Ras-transformed NIH 3T3 cells: coexpression with catalytically inactive SHP-2 blocks responsiveness. Mol Endocrinol 1997; 11:1062. Sharma K, Patel YC, Srikant CB. Subtype selective induction of p53-dependent apoptosis but not cell cycle arrest by human somatostatin receptor 3. Mol Endocrinol 1996; 10:1688. Tannenbaum GS, Epelbaum J. Somatostatin. In: Costio JL, ed. Handbook of physiology, vol 5. Hormonal control of growth. Sect vii, The endocrine system. New York: Oxford, 1999:221. D'Alessio DA, Sieber C, Beglinger C, et al. A physiologic role for somatostatin-28 as a regulator of insulin secretion. J Clin Invest 1989; 84:857. Colturi TJ, Unger RH, Feldman M. Role of circulating somatostatin in regulation of gastric acid secretion, gastrin release and islet cell function. Studies in healthy subjects and duodenal ulcer patients. J Clin Invest 1984; 74:417. Gyr K, Beglinger C, Kohler E, et al. Circulating somatostatin: physiological regulator of pancreatic function? J Clin Invest 1987; 79:1595. Yamada T. Gut somatostatin. In: Reichlin S, ed. Somatostatin: basic and clinical status. New York: Plenum, 1987:221. Juarez RA, Rubinstein M, Chan EC, et al. Increased growth following normal development in middle aged somatostatin deficient mice. In: Program of the Annual Meeting of the Society for Neuroscience; October 1997; New Orleans. Abstract 659.10:1684. Shimon I, Taylor JE, Dong JZ, et al. Somatostatin receptor subtype specificity in human fetal pituitary culture. J Clin Invest 1997; 99:789. Rohrer SP, Birzin ET, Mosley RT, et al. Rapid identification of subtype-selective agonists of the somatostatin receptor through combinatorial chemistry. Science 1998; 282:737. Zheng H, Bailey A, Jiang MH, et al. Somatostatin receptor subtype 2 knock-out mice are refractory to growth hormone, negative feedback on arcuate neurons. Mol Endocrinol 1997; 11:1709. Krejs GJ, Orci L, Conlon JM, et al. Somatostatinoma syndrome: biochemical, morphological and clinical features. N Engl J Med 1979; 301:285. Marbach P, Bauer W, Bodmer D, et al. Discovery and development of somatostatin agonists. Pharm Biotechnol 1998; 11:183. Lamberts SWJ, Krenning EP, Reubi JC. The role of somatostatin and its analogs in the diagnosis and treatment of tumors. Endocr Rev 1991; 12:450. Lancranjan L, Bruns C, Grass P, et al. Sandostatin LAR: pharmacokinetics, pharmacodynamics, efficacy, and tolerability in acromegalic patients. Metabolism 1995; 44:18. Giusti M, Gussoni G, Cuttica CM, et al. Effectiveness and tolerability of slow release lanreotide treatment in active acromegaly: six-month report on an Italian multicenter study. J Clin Endocrinol Metab 1996; 81:2089. Tomassetti P, Migliori M, Gullo L. Slow-release lanreotide treatment in endocrine gastrointestinal tumors. Am J Gastroenterol 1998; 93:1468. Rubin J, Ajani J, Schirmer W, et al. Octreotide acetate long-acting formulation versus open-label subcutaneous octreotide acetate in malignant carcinoid syndrome. J Clin Oncol 1999; 17:600. Vikic-Topic S, Raisch KP, Kvols LK, Vuk-Pavlovic S. Expression of somatostatin receptor subtypes in breast carcinoma, carcinoid tumor, and renal cell carcinoma. J Clin Endocrinol Metab 1995; 80:2974. Schaer J-C, Waser B, Mengod G, Reubi JC. Somatostatin receptor subtypes sst1, sst2, sst3 and sst5 expression in human pituitary, gastroentero-pancreatic and mammary tumors: comparison of mRNA analysis with receptor autoradiography. Int J Cancer 1997; 70:530. Evans AA, Cook T, Laws SAM, et al. Analysis of somatostatin receptor subtype mRNA expression in human breast cancer. Br J Cancer 1997; 75:798. Kvols LK, Moertal OG, O'Connell MJ, et al. Treatment of the malignant carcinoid syndrome: evaluation of a long-acting somatostatin analogue. N Engl J Med 1986; 315:663. Gorden P, Comi RJ, Maton PN, et al. Somatostatin and somatostatin analog (SMS 201-995) in treatment of hormone secreting tumors of the pituitary and gastrointestinal tract and nonneoplastic diseases of the gut. Ann Int Med 1989; 110:35. Wynick D, Bloom SR. The use of the long-acting somatostatin analog octreotide in the treatment of gut neuroendocrine tumors. J Clin Endocrinol Metab 1991; 73:1. Maton PN, Avaraki RF. Therapeutic use of somatostatin and octreotide acetate in neuroendocrine tumors. In: Weil C, Muller EE, Thorner MO, eds. Somatostatin. Basic and clinical aspects of neuroscience series, vol 4. Berlin: Springer-Verlag, 1992:55. Sassolas G, Harris AG, James-Deidier A, et al. Long-term effect of incremental doses of the somatostatin analog SMS 201-995 in 58 acromegalic patients. J Clin Endocrinol Metab 1990; 71:391. Melmed S, Jackson I, Kleinberg D, Klibanski A. Current treatment guidelines for acromegaly. J Clin Endocrinol Metab 1998; 83:2646. Ezzat S, Snyder PJ, Young WF, et al. Octreotide treatment of acromegaly. Ann Intern Med 1992; 117:711. Barkan AL, Lloyd RV, Chandler WF, et al. Preoperative treatment of acromegaly with long-acting somatostatin analog SMS 201-995: shrinkage of invasive pituitary macroadenomas and improved surgical remission rate. J Clin Endocrinol Metab 1988; 67:1040. Beck-Peccoz P, Brucker-Davis F, Persani L, et al. Thyrotropin-secreting pituitary tumors. Endocr Rev 1996; 17:610. Sharma K, Srikant CB. Induction of wild type p53, bax and acidic endonuclease during somatostatin signaled apoptosis in MCF-7 human breast cancer cells. Int J Cancer 1998; 76:259. Weckbecker G, Tolcsvai L, Stolz B, et al. Somatostatin analogue octreotide enhances the antineoplastic effect of tamoxifen and ovariectomy on 7, 12-dimethylbenz(a)anthracene-induced rat mammary carcinomas. Cancer Res 1994; 54:6334. Bontenbal M, Foekens JA, Lamberts SWJ, et al. Feasibility, endocrine and antitumor effects of a triple endocrine therapy with tamoxifen, a somatostatin analogue and an antiprolactin in postmenopausal breast cancer: a randomized study with long-term follow-up. Br J Cancer 1998; 77:115. Burroughs AK, McCormick PA, Hughes MD, et al. Randomized double blind placebo controlled trial of somatostatin for variceal bleeding. Gastroenterology 1990; 99:1388. Torres A, Landa J, Moreno-Azcoita M, et al. Somatostatin in the management of gastrointestinal fistulas. Arch Surg 1992; 127:97. Geer RJ, Richards WO, O'Dorisio TM, et al. Efficacy of octreotide acetate in treatment of severe postgastrectomy dumping syndrome. Ann Surg 1990; 212:678. Soudah HC, Hasler WL, Owyang C. Effect of octreotide on intestinal motility and bacterial overgrowth in scleroderma. N Engl J Med 1991; 325:1461. Hoeldtke RD, Israel BC. Treatment of orthostatic hypotension with octreotide. J Clin Endocrinol Metab 1989; 68:1051. Lamberts SWJ, Bakker WH, Reubi J-C, et al. Somatostatin receptor imaging in the localization of endocrine tumors. N Engl J Med 1990; 323:1246. 1575 Hofland LJ, Breeman WAP, Krenning EP, et al. Internalization of [DOTA°, 125 I-Tyr 3] octreotide by somatostatin receptor-positive cells in vitro and in vivo: implications for somatostatin receptor-targeted radioguided surgery. Proc Assoc Am Phys 1999; 111:63. McCarthy KE, Woltering EA, Espenan GD, et al. In situ radiotherapy with 111 In-pentetreotide: initial observations and future directions. Cancer J Sci Amer 1998; 4:94.

CHAPTER 170 KININS Principles and Practice of Endocrinology and Metabolism

CHAPTER 170 KININS DOMENICO C. REGOLI The Kallikrein–Kinin System Chemistry Inactivation Measurement In Blood Kinin Receptors Bioactivities of Kinins Vascular Effects Inflammation Algogenic Effect Other Endocrine and Exocrine Effects Other Possible Physiologic Effects Pathophysiologic Aspects of Kinins Kinin Activation Kallikrein–Kinins and Hypertension Chapter References

THE KALLIKREIN–KININ SYSTEM CHEMISTRY The kinins are potent vasodilator peptides that are contained in large precursors (kininogens), from which they are released into the blood and biologic fluids through the action of proteolytic enzymes (kallikreins). In contrast to neuropeptides and many other endogenous hormonal agents that are synthesized and stored in nerve and endocrine cells, kinins are generated in the blood and tissues.1 The kallikreins, which are responsible for the formation of kinins, originate from the liver (plasma kallikreins), from exocrine glands (glandular kallikreins),2 from the kidney, and from other organs and tissues.3 The kallikreins exist in blood and tissues as precursors (prekallikreins) that are activated by various chemical and physical factors. Among the endogenous activators are the Hageman factor, several enzymes (trypsin, plasmin, factor XI), and, in vitro, glass surfaces, kaolin, collagen, and other charged substances, such as cellulose sulfate.4,5 Plasma prekallikrein (with a molecular mass of 130,000 daltons) is the precursor of kallikrein (95,000–100,000 daltons), which interacts with the high-molecular-mass kininogen (88,000–114,000 daltons) to release the nonapeptide bradykinin5 (Fig. 170-1). Glandular kallikreins are acidic glycoproteins (27,000–43,000 daltons) that release the decapeptide kallidin (see Fig. 170-1) from kininogens of low molecular mass (48,000–70,000 daltons).2,3,6 Kallikreins and kininogens have been cloned: Sequence structures of the enzymes and the substrates are known (see comprehensive review7,8).

FIGURE 170-1. Formation and degradation of kinins. (LMW, low molecular weight; HMW, high molecular weight.) (From Regoli D. Polypeptides et antagonistes. In: Giroud J-P, Mathé G, Meyniel G, eds. Pharmacologie clinique: bases de la thérapeutique, 2nd ed. Paris: Expansion Scientifique Française, 1988:691.)

INACTIVATION Kinins are rapidly inactivated by several proteolytic enzymes present in plasma and tissues that act either at the amino or at the carboxyl end of the kinins (see Fig. 170-1). The metabolic products of kinins are inactive except for the bradykinin that is released from kallidin by the aminopeptidase, and the desArg9 bradykinin or desArg10 kallidin that is released from bradykinin and kallidin by kininase I.9,10 The most efficient system for inactivating kinins and activating angiotensin I to angiotensin II is kininase II, a carboxydipeptidase that is widely distributed in plasma, endothelial cells, and various organs, such as the lung and the kidney.9 Kininase II is particularly active in the pulmonary vascular bed: 80% to 95% of kinins are eliminated during the short time in which the blood passes through the pulmonary circulation.11 Because of their rapid inactivation in the lung and in other tissues, the biologic half-lives of bradykinin and kallidin in the dog are 0.27 and 0.32 minutes, respectively. Thus, as a result of the balance between the simultaneous processes of production and inactivation, kinins circulate at very low concentrations. MEASUREMENT IN BLOOD BLOOD LEVELS IN HEALTHY SUBJECTS Various investigators have reported extreme variations (from 0.07–5.0 ng/mL) in circulating kinin concentrations in healthy volunteers, using either bioassays or radioimmunoassays.10 One group has reported a concentration of 0.025 ng/mL, a value 100 to 200 times lower than any previously reported.12 Such discrepancies are attributable largely to the fact that prekallikreins in blood are rapidly activated by contact with glass surfaces and by numerous other physical and chemical factors.1 Therefore, most of the data on the blood concentration of kinins reported since the 1970s reflect figures that are much too high compared to more recent results obtained with sensitive radioimmunoassays.4 When both types of assays have been used for the same plasmas, radioimmunoassay has yielded values two to six times higher than those derived from bioassay, probably because antibodies that are directed against the C-terminal part of the kinin molecules measure both the biologically active kinins and some inactive metabolites.4,10 BLOOD LEVELS IN DISEASE An increase of bradykinin-like material (measured by bioassay or radioimmunoassay) has been reported in the blood of patients affected by the dumping syndrome, postgastrectomy with or without dumping, dengue fever and shock, bronchial asthma, or hyperbradykininism, as well as in the synovial fluid of patients with rheumatoid and psoriatic arthritis.4,10 Again, the levels of circulating kinins reported by several of these authors are too high. More sensitive and specific radioimmunoassays13 and better-controlled experimental conditions for blood collection are needed to demonstrate significant changes of circulating kinins in pathophysiology. Alternatively, the activity of the kallikrein–kinin system can be evaluated by relating the activity of kallikreins to the rate of generation of kinins in vitro, in a manner similar to that used to estimate renin. Several researchers have used this approach to measure urinary kallikreins in an attempt to evaluate the possible role of renal kallikreins in human and experimental hypertension.14

KININ RECEPTORS Naturally occurring kinins and their potentially active metabolites exert various biologic actions and have different metabolic and pharmacologic features.10 Pharmacologic studies suggest that kinins may exert their numerous biologic effects by activating two different types of receptors.4,10 Differentiation between the two receptors has been accomplished by using classic criteria,15 and other modern criteria16 in relation to the B1 receptor. Thus, rabbit aorta, which contains the B1 receptor, is particularly sensitive to desArg9 bradykinin and desArg10 kallidin, the metabolites of bradykinin and kallidin released by kininase I. Conversely, the receptor of the rabbit jugular vein is much more sensitive to kallidin and bradykinin than to their metabolites. The discovery of specific and competitive antagonists for the B1 receptor, such as [Leu9]desArg10 kallidin and other similar compounds,10 as well as the finding that [Thi5,8, Phe7] bradykinin acts as a specific antagonist of B2 receptors,17 confirms the existence of two receptor types for the kinins. In fact, [Leu9]desArg10 kallidin is active on the B1 receptor and inactive on the B2 receptor, whereas [Thi5,8, Phe7] bradykinin is active on the B2 receptor. A selective B2-receptor antagonist, HOE 140 (D-Arg[Hyp3, Thi5,D-Tic7, Oic8]BK) has been reported18 and shown to be resistant to degradation and to have prolonged effects in vivo. The human B1 and B2 receptors have been cloned and shown to be rhodopsin-like proteins linked to G proteins.19,20 Two B2-receptor subtypes have been identified and characterized.21,22 Kininase II, the enzyme that releases desPhe8, desArg9 brady-kinin from bradykinin, and desPhe9 and desArg10 kallidin from kallidin, is indeed the most efficient inactivating system, because the metabolites are practically inactive in both preparations and, therefore, on both receptor types.

BIOACTIVITIES OF KININS VASCULAR EFFECTS The kallikreins and kinins were discovered mainly because of their potent cardiovascular effects.1,23 One of their prominent actions is the dilatation of peripheral vessels, resulting from the activation of B2 receptors on the vascular endothelium.24,24a, Similar to other agents, kinins may release a potent endogenous vasodilator from the endothelium that reduces the arterial smooth muscle tone, thereby increasing the blood flow to most of the peripheral organs and reducing the systemic blood pressure.25 Kinins also influence other blood vessels: They increase capillary permeability, probably by diminishing the size (possibly through contraction) of the endothelial cells10; they stimulate the veins; and they also modulate the functions of a variety of blood and tissue cells. INFLAMMATION Kinins are among the most potent activators of the arachidonic acid cascade,26 and they promote the release of prostaglandins, prostacyclins, and other similar compounds from various cells (see Chap. 172). Kinins also stimulate the release of histamine and, possibly, of 5-hydroxytryptamine from mast cells. Moreover, they modulate the motility and the function of leukocytes, macrophages, fibroblasts, and other cells.4,26 Kinins have also been implicated in the pathogenesis of inflammation, of tissue reactions to injury, and of tissue repair. In this context, two biologic effects of kinins are particularly important: the stimulation of the mitosis of T lymphocytes27 and the activation of protein formation and cell division (by both bradykinin and desArg9 bradykinin) in human lung fibroblasts.28 Collagen formation also appears to be stimulated by the kinins. The involvement of B2 and especially B1 receptors in acute or chronic pain and inflammation has been discussed.16,29 ALGOGENIC EFFECT Kinins are the most active natural substances that are algogenic (producing pain). When injected intraarterially or applied on a cutaneous “blister base,” these peptides evoke pain in both human subjects and animals. It is still uncertain whether kinins produce the pain by direct stimulation of the nerve endings or through the release of prostaglandins, which strongly potentiate kinin algogenic actions.26,29 Most of the effects mentioned earlier, including pain, are mediated by B2 receptors, which are constitutive: In pathologic conditions, the inducible (e.g., by interleukin-b16) B1 receptors sustain the acute effects initiated by the B2.16,29 OTHER ENDOCRINE AND EXOCRINE EFFECTS Kinins increase the release of both catecholamines from the adrenal glands and renin from the kidney. The physiologic significance of such effects is uncertain, considering the high concentrations (10-6 M) of kinins that are needed to stimulate the hormonal release. Kinins also promote the secretion of exocrine glands, such as the pancreas and the salivary glands, which have a very high kallikrein content.8,10 OTHER POSSIBLE PHYSIOLOGIC EFFECTS Receptors for kinins, generally of the B2 type, are found in various smooth muscles in the gastrointestinal and urinary tracts, the tracheobronchial tree, and the uterus. Except for the bronchi, the physiologic significance of such stimulatory effects of the kinins in all of these organs remains to be established. Through a complex mechanism involving the prostaglandins, angiotensin, and histamine, and also through direct effects on several types of renal cells (vascular, parenchymatous, interstitial), kinins act on the kidney to induce natriuresis and diuresis. They also antagonize the effect of vasopressin on water excretion.30,31 In the pancreas and the kidney, glandular kallikreins may act as activators of proinsulin and prorenin; such observations are interesting because they suggest that kallikreins may be involved in the conversion of prohormones to hormones. Moreover, these enzymes and their peptide products have other effects,32 since they promote sperm motility and cell proliferation. Furthermore, the b-nerve growth factor, endopeptidase, appears to be a kallikrein.33 Surveys of the many possible biologic roles of kallikreins and an overview of kallikrein and kininogens in general have been published.6,8

PATHOPHYSIOLOGIC ASPECTS OF KININS KININ ACTIVATION Apparently, kinins and kallikreins are hormones that are activated or generated by noxious stimuli to participate in the processes of tissue defense and repair. Probably the first of these processes is blood clotting and fibrinolysis. Both plasma kallikrein and the Hageman factor (kallikrein activator) promote clotting; deficiencies of these factors are the major causes of the Fitzgerald, the Williams, the Flanjeac, and also the Fugiwara traits. The interactions between the various factors involved have been summarized as follows. Contact with negatively charged surfaces converts Hageman factor (clotting factor XII) to its active form (XIIa). The active enzyme XIIa then initiates intrinsic and extrinsic blood clotting, fibrinolysis, and plasma kinin formation, by activating clotting factors VII and XI, prekallikrein, and plasminogen activator, respectively. The active factor XIa, plasmin, and plasma-kallikrein then convert more factor XII to XIIa. High-molecular-weight kininogen facilitates the interaction between XIIa and kallikrein, and also serves as a kinin-yielding substrate. Prekallikrein is also activated by plasmin.34 Such activations may occur in several pathologic states. According to one investigator,34 “significant activation may be produced by urate and pyrophosphate crystals in gout and pseudogout, by collagen in vascular injury, by glomerular basement membrane in some forms of glomerulonephritis, and by lipopolysaccharides in gramnegative infections.” Kinins contribute to the basic vascular and cellular events that occur in inflammation.35 Local activation of kallikrein and an increased production of kinins have been found in various inflammatory lesions produced experimentally, for example, by carrageenan, urate crystals, heat, and other noxious manipulations.10 The decrease of pH in inflamed tissues may activate kallikreins and reduce the degradation of kinins, thereby maintaining fairly high concentrations of these peptides during the acute and chronic phases of inflammation. In experimental shock, the kallikrein–kinin system is activated.36 Kinins contribute not only to rubor, dolor, calor, and tumor, but to tissue defense and repair by the activation of cell (polymorphonucleocytes, leukocytes) migration. Kallikrein may also influence the renal response to sepsis.37

KALLIKREIN–KININS AND HYPERTENSION The findings that high concentrations of kallikreins and traces of kinins are present in the urine of healthy people, and that kallikreins are decreased in some forms of hypertension, have caused great interest. Kallikreins are synthesized in the distal tubule, possibly to control water and electrolyte excretion. They have a functional link with the reninangiotensin system.38 By diffusing to the juxtaglomerular apparatus, kallikreins promote the conversion of prorenin to renin.3 The urinary kallikreins are influenced by variations in sodium and mineralocorticoids; they are increased in patients consuming a low-sodium diet and in those with primary aldosteronism, Bartter syndrome, or hyponatremia. They are reduced in essential hypertension and after the administration of spironolactone.39 This decrease in kallikrein excretion is found in both human and experimental hypertension, and it has been suggested that a decreased activity level of the kallikrein–kinin system in the kidney might generate favorable conditions for the development and maintenance of high blood pressure.12 Other implications of the kinins and their receptors in pathophysiology have been reviewed.16,40 CHAPTER REFERENCES 1. Rocha e Silva M. Kinin hormones. Springfield, IL: Charles C Thomas Publisher, 1970. 2. Pisano JJ. Chemistry and biology of the kallikrein-kinin system. In: Reich E, Rifkin DB, Shaw E, eds. Proteases and biological control, vol 2. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1975:199. 3. Levinsky NG. The renal kallikrein-kinin system. Circ Res 1979; 44:441. 4. Marceau F, Lussier A, Regoli D, Giroud J-P. Pharmacology of kinins: their relevance to tissue injury and inflammation. Gen Pharmacol 1983; 14:209. 5. Colman RW. Formation of human plasma kinin. N Engl J Med 1974; 291:509. 6. Schachter M. Kallikreins (kininogenases): a group of serine proteases with bioregulatory actions. Pharmacol Rev 1980; 31:1. 7. Hall JM. Bradykinin receptors: pharmacological properties and biological roles. Pharmacol Ther 1992; 56:131. 8. Bhoola KD, Figneroa CD, Worthy K. Bioregulation of kinins: kallikreins, kininogens and kininases. Pharmacol Rev 1992; 44:1. 9. Erdüs EG. Kininases. Handbook Exp Pharmacol 1979; 25(Suppl):427. 10. Regoli D, Barabé J. Pharmacology of bradykinin and related kinins. Pharmacol Rev 1980; 32:1. 11. Ferreira SH, Vane JR. The detection and estimation of bradykinin in the circulating blood. Br J Pharmacol 1967; 29:367. 12. Carretero OA, Scicli AG. Possible role of kinins in circulatory homeostasis. Hypertension 1981; 3:1. 13. Shimamoto K, Ando T, Tanaka S, et al. An improved method for the determination of human blood kinin levels by sensitive kinin radioimmunoassay. Endocrinol Jpn 1982; 29:487. 14. Carretero OA, Scicli AG. The renal kallikrein–kinin system. Am J Physiol 1980; 238:F247. 15. Schild HO. Receptor classification with special reference to beta adrenergic receptors. In: Rang HP, ed. Drug receptors. Baltimore: University Park Press, 1973:29. 16. Marceau F, Hess JF, Bachvarov DR. The B 1 receptor for kinins. Pharmacol Rev 1998;50:357. 17. Vavrek RJ, Stewart JM. Competitive antagonists of bradykinin. Peptides 1985; 6:161. 18. Wirth K, Hock FJ, Albus U, et al. HOE 140, a new potent and long-acting bradykinin antagonist: in vivo studies. Br J Pharmacol 1991; 102:774. 19. Hess JF, Borkowski JA, Young GS, et al. Cloning and pharmacological characterization of a human bradykinin (BK-2) receptor. Biochem Biophys Res Commun 1992; 184:260. 20. Menke JG, Borkowski JA, Bierilo KK, et al. Expression cloning of a human B1 bradykinin receptor. J Biol Chem 1994; 269:21583. 21. Regoli D, Jukic D, Rhaleb NE, Gobeil F. Receptors for bradykinin and related kinins. A critical analysis. Can J Physiol Pharmacol 1993; 71:556. 22. Regoli D, Gobeil F, Nguyen OT, et al. Bradykinin receptor types and subtypes. Life Sci 1994; 55:735. 23. Werle E, Berek U. Zur Kenntnis des Kallikreins. Z Angew Chemie 1948; 60A:53. 24. D'Orléans-Juste P, Dion S, Mizrahi J, Regoli D. Effects of peptides and nonpeptides on isolated arterial smooth muscles: role of endothelium. Eur J Pharmacol 1985; 114:9. 24a. Halle S, Gobeil F Jr, Ouellette J, et al. In vivo and in vitro effects of kinin B(1) and B(2) receptor agonists and antagonists in inbred control and cardiomyopathic hamsters. Br J Pharmacol 2000; 129:1641. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Thorgeirsson G. Endothelial autacoids. Acta Med Scand 1985; 217:453. Nasjletti A, Malik KU. Relationships between the kallikrein–kinin and prostaglandin systems. Life Sci 1979; 25:99. Perris AO, Whitfield JF. The mitogenic action of bradykinin on thymic lymphocytes and its dependence on calcium. Proc Soc Exp Biol Med 1969; 62:1198. Goldstein RH, Wall M. Activation of protein formation and cell division by bradykinin and desArg 9-bradykinin. J Biol Chem 1984; 259:9263. Dray A, Perkins M. Bradykinin and inflammatory pain. Trends Neurosci 1993; 16:99. Abe K, Irokawa N, Yasujima M, et al. The kallikrein–kinin system and prostaglandins in the kidney. Circ Res 1978; 43:254. Yamada K, Hasunuma K, Shiina T, et al. Inter-relationship between urinary kallikrein–kinins and arginine vasopressin in man. Clin Sci 1989; 76:13. Jauch KW, Hartl WH, Georgieff M, et al. Low-dose bradykinin infusion reduces endogenous glucose production in surgical patients. Metabolism 1988; 37:185. Bothwell MA, Wilson WH, Shooter EM. The relationship between glandular kallikrein and growth factor-processing proteases of mouse submaxillary gland. J Biol Chem 1979; 254:7287. Eisen V. Kinins in physiology and pathology. Trends Pharmacol Sci 1980; 1:212. Garcia-Leme J. Bradykinin system. Handbook Exp Pharmacol 1979; 50:464. Christopher TA, Xin-Liang MA, Gouthier TW, Lefer AM. Beneficial actions of CP-0127, a novel bradykinin receptor antagonist, in murine traumatic shock. Am J Physiol 1994; 266:H867. Cumming AD, Driedger AA, McDonald JW, et al. Vasoactive hormones in the renal response to systemic sepsis. Am J Kidney Dis 1988; 11:23. Cumming AD, Jeffrey S, Lambic AT, Robson JS. The kallikrein–kinin and renin–angiotensin systems in nephrotic syndrome. Nephron 1989; 51:185. Mills IH. Kallikrein, kininogen and kinins in control of blood pressure. Nephron 1979; 23:61. Stewart JM, Gera L, Chan DC, et al. Potent, long-acting bradykinin antagonists for a wide range of applications. Can J Physiol Pharmacol 1997; 75:719.

CHAPTER 171 SUBSTANCE P AND THE TACHYKININS Principles and Practice of Endocrinology and Metabolism

CHAPTER 171 SUBSTANCE P AND THE TACHYKININS NEIL ARONIN Tachykinins Biosynthesis of Substance P Distribution and Actions of Substance P in Endocrine Glands Anterior Pituitary Thyroid Adrenal Gland, Ovary, and Testis Pancreas Substance P in the Blood Substance P in Clinical Medicine Chapter References

Substance P, which is an 11–aminoacid peptide formed by cleavage of a larger precursor molecule, has a widespread and heterogeneous distribution in the central and peripheral nervous systems, the gastrointestinal tract, and the endocrine tissues.1,2,3,4 and 5 Consistent with its presence in multiple tissues, this peptide has diverse actions, stimulating or enhancing activity in some cells, while inhibiting activity in others. For example, in the dorsal horn of the spinal cord, substance P acts as an excitatory neurotransmitter in primary afferents conveying pain, whereas in the ventral horn of the spinal cord, substance P inhibits Renshaw cell activity. Detailed reviews on the discovery, localization, and physiologic role of substance P in the neural and gastrointestinal systems are available.3,4 and 5,5a Current research on substance P (and the related peptides, tachykinins) centers on their contributions to inflammation. The availability of tachykinin-specific receptor antagonists and knock-out mice has provided useful tools to study the potential role of substance P and tachykinins in the pathogenesis and treatment of a variety of diseases.

TACHYKININS Substance P shares a common carboxyl-terminus (C-terminus) sequence with other peptides. Together, these related peptides with common hypotensive properties are known as tachykinins (Fig. 171-1). A number of laboratories contributed to the discovery and naming of the tachykinin family of peptides. A uniform nomenclature has now been adopted. Substance P, neurokinin A, neuropeptide K, and neuropeptide g are all found in mammalian tissues. In general, substance P is the tachykinin of highest abundance. Physalaemin, which may be found in small amounts in mammalian brain, was first isolated from the skin of a South American amphibian. Eledoisin was discovered in extracts from cephalopod salivary glands. Neurokinin A (heretofore, substance K) was discovered in a prohormone that also contains substance P.6,7 Neuropeptide K (previously called neuromedin K) resembles other tachykinins structurally and shares some of their biologic effects, but is generated from a different tachykinin precursor.8 Many of the biologic activities of the tachykinins reside in the amidated C terminus.

FIGURE 171-1. Structures of substance P and some related peptides. (pGlu, pyroglutamic acid; PPT, preprotachykinin.)

BIOSYNTHESIS OF SUBSTANCE P Three protein precursors to substance P have been identified: a-, b-, and g-preprotachykinin A. The primary structures of these precursors have been derived from mammalian brain cDNA sequences.4,6,7,9 These tachykinin precursors are nearly homologous (Fig. 171-2) and differ in the inclusion (or exclusion) of specific exons in the cDNAs. Both substance P and neurokinin A are present in b-and g-preprotachykinin A. Only substance P is found in a-preprotachykinin A. All three preprotachykinin mRNAs originate from a single gene, preprotachykinin A, and are subsequently generated by alternative RNA splicing.6,9 A separate gene, preprotachykinin B, leads to the synthesis of the protein precursor to neurokinin B.8 A cDNA encoding b-preprotachykinin A has been isolated in a human carcinoid tumor containing a large amount of immunoreactive substance P.10 The human cDNA for b-preprotachykinin A has extensive sequence overlap with bovine cDNA b-preprotachykinin A.

FIGURE 171-2. The preprotachykinin-A (PPT-A) gene and alternatively spliced mRNA structures: a-, b-, g-preprotachykinin A mRNAs. Different tissues may have a preference for expressing one of the mRNAs. The unamidated substance P (SP), neurokinin A (NKA), and N terminally extended forms of NKA, neuropeptide K (b-PPT-A mRNA), and neuropeptide g (g-PPT-A mRNA) would then be cleaved enzymatically from the precursor protein and amidated to form bioactive peptides. (Reprinted from Nawa H, Kotani H, Nakanishi S. Tissue-specific generation to two preprotachykinin mRNAs from one gene by alternative RNA splicing. Nature 1984; 312:733, copyright 1984, Macmillan Journals Limited.)

The relative distribution of mRNAs for a-, b-, and g-preprotachykinin A varies in neural and nonneural tissues; their peptide products also vary.4,7 The finding of several, structurally related tachykinins in the same brain regions and peripheral tissues suggests that these peptides may interact on their target cells. There is little information on putative interactions, however. Substance P and neurokinin A share properties of the tachykinin family, but differ in potency.11 Clues to cellular effects of the tachykinins may be emerging from advances on tachykinin receptors. Until now, the designation of tachykinin receptors was based on relative affinities to physalaemin, eledoisin, and substance K (neurokinin A). It is now known that tachykinin receptors (to date, NK1, NK2, NK3) are each encoded by a separate gene. NK1 has highest affinity to substance P; NK2 to neurokinin A; NK3 to neurokinin B.12,13 However, each NK-receptor subtype binds other tachykinin peptides.13 The receptors couple to G protein (possibly through a pertussis toxininsensitive G protein) and appear to participate in the regulation of intracellular Ca2+ by affecting phosphoinositide hydrolysis.13,14 Specific actions of the receptor subtypes are under intense investigation, especially with use of newly developed pharmacologic tools such as peptide

and nonpeptide tachykinin receptor agonists and antagonists.15

DISTRIBUTION AND ACTIONS OF SUBSTANCE P IN ENDOCRINE GLANDS Substance P–like immunoreactivity is measurable in most endocrine tissues by radioimmunoassay and immunohisto-chemistry. However, despite studies suggesting that it may inuence the release of various hormones, specic roles for subuence the release of various hormones, specic roles for substance P in endocrine tissues have not yet been elucidated.16,17 Most of these studies examined effects of substance P, rather than the related tachykinins; the results can be attributed to substance P primarily. ANTERIOR PITUITARY Substance P may gain access to the regulation of the anterior pituitary at many sites along the hypothalamic-hypophysial axis. Substance P is detected in neural afferents to the hypothalamus, neuroendocrine cells of the hypothalamus, hypothalamic interneurons, the median eminence, neural fibers in the anterior pituitary, and cells composing the anterior pituitary.16,18,19 and 20 It is probably not surprising that effects of substance P on hormone release appear to depend (at least in part) on the route of administration. Thus, the substance P–induced increase in prolactin secretion is usually consistent, whether the tachykinin is administered intravenously or intracerebroventricularly, or in vitro. However, the substance P effects on growth hormone, luteinizing hormone, corticotropin, and thyroid-stimulating hormone secretion are less consistent and vary according to route of peptide administration, use of antipeptide antisera, overall hormonal state of the animal, and maturation of the animal. Results of intravenous injection of substance P on pituitary hormone release in human subjects do not necessarily correspond to findings in other experimental animals.17,21 There is, nonetheless, some enthusiasm for a paracrine role for substance P (or other tachykinins) in the anterior pituitary gland. Substance P is measurable in anterior pituitary extracts, its pituitary concentrations are regulated by thyroid hormone and gonadal steroids, and it is located primarily in somatotropes in male rats and more frequently in thyrotropes in female rats. 19,22,23 and 24 Substance P is synthesized in the anterior pituitary (including human), and the pituitary concentrations of several preprotachykinin A mRNAs are regulated by thyroid hormone and gonadal steroids.25 Direct effects of substance P on secretion of hormones in the anterior pituitary have been demonstrated.16,17,26 THYROID Substance P immunoreactivity in the thyroid has been demonstrated in neural fibers surrounding blood vessels and follicles. No effect of substance P on iodothyronine release has been observed, but in the rat, substance P is known to stimulate calcitonin secretion.27 ADRENAL GLAND, OVARY, AND TESTIS The substance P that is found in the adrenal medulla appears to be localized to nerve fibers that originate from the splanchnic nerve and terminate adjacent to chromaffin cells. Many studies have supported the concept of a modulatory role for substance P in catecholamine release; alone, this peptide has little or no direct influence on catecholamine secretion. There is evidence that substance P enhances catecholamine release after splanchnic nerve stimulation, by facilitating acetylcholine release pre-synaptically and postsynaptically to protect against nicotinic receptor desensitization.28,29 Substance P may also influence cortisol secretion from adrenocortical cells. In vitro, substance P increased cortisol release in a Ca2+-dependent manner in the adrenal cortex (where adrenocorticotropic hormone acts primarily through cyclic adenosine monophosphate–dependent signal transduction).30 A physiologic role for substance P in adrenal cortical control remains speculative. Likewise, substance P and neurokinin A mRNAs and tachykinin receptors are detectable in testis, albeit in low abundance.31 Substance P is distributed in nerve fibers in the ovary and also may be produced in both granulosa and luteal cells in culture, where it inhibits androstenedione and progesterone release, but stimulates estradiol secretion.32 The studies are few, however, and the results unconfirmed. PANCREAS In the pancreas, substance P is found in neural fibers; it influences secretion from the exocrine and endocrine pancreas. Arterial infusion of substance P increases total pancreatic outflow, and, therefore, secretion of bicarbonate and amylase.33 During meals or secretin administration, substance P inhibits amylase secretion. There is evidence that it promotes hyperglycemia, but direct effects on insulin and glucagon secretion (either enhancing or diminishing) are inconsistent. Because substance P and calcitonin generelated peptide (CGRP) are localized in pancreatic nervous tissue, neural stimulation in the pancreas may be attributable to one or both of these chemical messengers.34 Different responses also probably depend on differences in study designs. SUBSTANCE P IN THE BLOOD Substance P has been measured by radioimmunoassay in plasma extracts. In humans, reported normal values for immunoreactive substance P vary from 30 to more than 600 fmol/mL.35,36 and 37 As with many small peptides, plasma measurements are likely to be influenced by the method of extraction, as well as by nonspecific factors that alter antibody-peptide binding in the radioimmunoassay. Much of the circulating substance P appears to be derived from the intestine because the portal venous blood contains four times the concentration of the peptide in peripheral areas. The physiologic role of circulating substance P remains unknown. SUBSTANCE P IN CLINICAL MEDICINE Carcinoid tumors contain peptides, along with serotonin or its metabolites. Substance P immunoreactivity has been detected in carcinoid tumors originating in the midgut region, as well as in less common sites of origin, such as the testicle and ovary. High circulating levels of substance P have been found in patients with carcinoid tumors.38,39 This finding suggests that substance P may be a useful marker for the localization of these tumors. In one report in which selective catheterization and measurement of plasma serotonin failed to localize an ovarian carcinoid tumor, selective intraoperative blood sampling and measurement of substance P from the ovarian venous drainage revealed an increase in plasma concentrations of substance P.40 Occasionally, increased plasma concentrations of substance P have been found in patients with carcinoid tumors when simultaneous measurements of serotonin or serotonin metabolite levels were negative.33 These observations must be confirmed prospectively to determine whether plasma substance P can be used for diagnosing carcinoid, for precisely localizing the tumor, and for judging efficacy of treatment. Because substance P can mediate many vascular changes, including enhancement of vascular permeability and vasodilation, as well as increase intestinal motility, a role for substance P in the carcinoid diathesis has been postulated.33 In dogs, the arterial infusion of substance P in amounts sufficient to equal the plasma concentrations of substance P found in patients with carcinoid tumors reduced systemic vascular resistance, increased cardiac output, and enhanced the redistribution of blood flow to the muscular layers of the gastrointestinal tract.41 There are a few reports of the presence of substance P in other types of cancers. In some cases of medullary carcinoma of the thyroid, substance P immunoreactivity has been documented in the tumor, and increased levels have been detected in plasma extracts.32 Antagonists directed to tachykinin receptors were shown to inhibit the growth of small cell lung cancer transplanted in nude mice.42 Because antagonists to substance P receptors can have broad effects against other neuropeptide receptors (bombesin, gastrin-releasing peptide, vasopressin), a specific growth-promoting role for tachykinins in this experimental paradigm remains unsettled. A number of investigations support a role for tachykinins in the pathogenesis of diseases involving inflammation. There is evidence that the substance P that is released from peripheral nerve terminals may contribute to the development of rheumatoid arthritis. When incubated with substance P, synoviocytes from patients with rheumatoid arthritis were found to increase the generation of prostaglandin E2 and the release of collagenase.43 These effects may be observed with as little as 10-9 M of substance P, and are inhibited by a substance P antagonist. Furthermore, it has been proposed that substance P may mediate neural effects on the initiation or maintenance of the inflammatory process in the arthritic joint.43 Substance P, in partnership with proinflammatory cytokines, increases the expression of the adhesion molecule, vascular cell adhesion molecule-1 (VCAM-1), in fibroblast-like synoviocytes, thus promoting inflammation of the synovium.44 Tachykinins, especially substance P, may contribute to the pulmonary inflammation and airway hyperactivity found in experimental asthma.45,46 47 Agents that cause tachykinin release from nerve endings, such as capsaicin, can result in contraction of isolated tracheal smooth muscle, an effect mediated by NK1 and NK2 receptors. Pretreatment with capsaicin, which probably depletes neuropeptides from autonomic nerve endings, markedly attenuates carbachol-induced hyperactivity in mice that manifest tracheal hyperactivity.47 Altogether, tachykinins may modulate immune function and reactions by effects on cellular mediators of inflammation (mast cells, macrophages, T cells), by stimulating contractility of smooth muscle (broncho-constriction), and by affecting proliferation of smooth muscle cells, endothelial cells, and

fibroblasts.45,47 Use of NK1 knock-out mice has confirmed a role for substance P in pulmonary inflammation. The expected proinflammatory response to immune complex was abrogated in mice lacking NK1 receptors.48,49 50 Non-peptide tachykinin antagonists offer promise of reducing tachykinin-dependent plasma extravasation or bronchoconstriction after a variety of chemical challenges to the airway.51 To date, the efficacy of tachykinin antagonists in human pulmonary disease remains unknown. The original extracts in which substance P was found included intestinal tissues.1 Substance P and tachykinins are expressed throughout the gut—primarily in the enteric and extrinsic primary afferent nervous systems, but also in endocrine cells in the epithelium, immune cells, and endothelial cells.52 Diseases affecting the enteric or primary afferent nervous system in the intestines are associated with substance P reduction. These diseases include Hirschsprung disease and megacolon in patients with myotonic dystrophy. Increases in substance P receptor expression or sensitivity are reported in inflammatory bowel disease, implicating actions of tachykinins in their inflammatory and vascular abnormalities. Substance P may also prove useful in the early detection of diabetic sensory neuropathy of the skin. Exposure to noxious agents stimulates the release of transmitters from primary sensory neurons innervating skin. This release is followed by immediate vasodilatation, an increase in vascular permeability, and the release of histamine from local mast cells that causes a visible cutaneous “flare” and wheal. The phenomenon, which is called the axon response or triple response of Lewis, is thought to be mediated by a number of components, including substance P found in a subpopulation of primary sensory neurons. diabetic patients with clinical sensory neuropathy exhibit a reduced flare response after the intradermal injection of substance P, as compared to that of healthy subjects or diabetics without clinical evidence of neuropathy.53 Also, the impaired responses in diabetics have been found to occur in medial forearm skin, where sensory function is often normal by clinical examination (see Chap. 148). Thus, the substance P–provoked flare response may be a sensitive indicator of both the presence of cutaneous sensory neuropathy in diabetic patients and changes in sensory neural function after therapy. There is also clinical evidence that depletion of substance P (or other effects on the C fiber) after daily application of dilute capsaicin to the skin has salutary effects in the treatment of painful diabetic neuropathy.54 Not all sensory modalities may be improved, but it appears that up to half of diabetic patients with uncontrolled, painful neuropathy may benefit from this therapy. Long-term side effects of this treatment are not known. Continued use of the noncommittal term “substance P” seems appropriate. This hormone may have multiple roles in the endocrine system—as a neurotransmitter, neuromodulator, or paracrine agent—in keeping with its presence in neural inputs to endocrine glands, as well as its expression in endocrine cells. Substance P and its related tachykinins also serve as mediators at neural and immune junctions. Advances in the knowledge of tachykinin receptors and the availability of peptide and nonpeptide receptor antagonists offer considerable possibilities in the treatment of a wide variety of clinical disorders in which the tachykinins are implicated to play an important role. CHAPTER REFERENCES 1. 2. 3. 4. 5. 5a. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

von Euler US, Gaddum JH. An unidentified depressor substance in certain tissue extracts. J Physiol 1931; 72:74. Chang MM, Leeman SE, Niall HD. Amino acid sequence of substance P. Nature 1971; 232:86. Aronin N, DiFiglia M, Leeman SE. Substance P. In: Krieger DT, Brownstein MJ, Martin JB, eds. Brain peptides. New York: John Wiley and Sons, 1983:783. Krause JE, MacDonald MR, Takeda Y. The polyprotein nature of substance P precursors. Bioessays 1989; 10:62. Otsuka M, Yoshioka K. Neurotransmitter functions of mammalian tachykinins. Physiol Rev 1993; 73:229. Hokfelt T, Broberger C, Xu ZQ, et al. Neuropeptides—an overview. Neuropharmacol 2000; 39:1337. Nawa H, Hirose T, Takashima H, et al. Nucleotide sequences of cloned cDNAs for two types of bovine brain. Nature 1983; 306:32. Nawa H, Kotani H, Nakanishi S. Tissue-specific generation to two preprotachykinin mRNAs from one gene by alternative RNA splicing. Nature 1984; 312:729. Kotani H, Hoshimaru M, Nawa H, Nakanishi S. Structure and gene organization of bovine neuromedin K precursor. Proc Natl Acad Sci U S A 1986; 83:7074. Krause JE, Chirgwin JM, Caarter MS, et al. Three rat preprotachykinin mRNAs encode the neuropeptides substance P and neurokinin A. Proc Natl Acad Sci U S A 1987; 84:881. Harmer AJ, Armstrong A, Pascall JC, et al. cDNA sequence of human b-preprotachykinin, the common precursor to substance P and neurokinin A. FEBS Lett 1986; 208:67. Nawa H, Doteuchi M, Igano K, et al. Substance K: a novel mammalian tachykinin that differs from substance P in its pharmacological profile. Life Sci 1984; 34:1153. Masu Y, Nakayama K, Tamaki H, et al. cDNA cloning of bovine substance-K receptor through oocyte expression system. Nature 1987; 329,836. Watson S, Girdleston D. Receptor and ion channel nomenclature supplement. Trends Pharmacol Sci 1994; 15:40. Kwatra MM, Schwinn DA, Schreurs J, et al. The substance P receptor, which couples to Gq/11 , is a substrate of b-adrenergic receptor kinase 1 and 2. J Biol Chem 1993; 268:9161. Watling KJ, Krause J. The rising sun shines on substance P and related peptides. Trends Pharmacol Sci 1993; 14:81. Aronin N, Coslovsky R, Leeman SE. Substance P and neurotensin: their roles in the regulation of anterior pituitary function. Annu Rev Physiol 1985; 48:537. Jessop DS, Chowdrey HS, Larsen PJ, Lightman SL. Substance P: multifunctional peptide in the hypothalamo-pituitary system? J Endocrinol 1992; 132:331. Larsen PJ. Distribution of substance P-immunoreactive elements in the preoptic area and the hypothalamus of the rat. J Comp Neurol 1992; 316:287. Brown ER, Roth KA, Krause JE. Sexually dimorphic distribution of substance P in specific anterior pituitary cell populations. Proc Natl Acad Sci U S A 1991; 88:1222. Ju G, Liu S-J, Ma D. Calcitonin generelated peptide-and substance P-like immunoreactivity innervation of the anterior pituitary in the rat. Neuroscience 1993; 54:981. Coiro V, Capretti L, Volpi R, et al. Stimulation of ACTH/cortisol by intravenously infused substance P in normal men: inhibition by sodium valproate. Neuroendocrinology 1992; 56:459. Aronin N, Morency K, Leeman SE, et al. Regulation by thyroid hormone of the concentration of substance P in the rat anterior pituitary. Endocrinology 1984; 114:2138. Coslovsky R, Evans RW, Leeman SE, et al. The effects of gonadal steroids on the content of substance P in the rat anterior pituitary. Endocrinology 1984; 115:2285. Aronin N, Coslovsky R, Chase K. Hypothyroidism increases substance P concentrations in the heterotopic anterior pituitary. Endocrinology 1988; 122:2911. Jonassen JA, Mullikin-Kilpatrick D, McAdam A, Leeman SE. Thyroid hormone status regulates preprotachykinin-A gene expression in male rat anterior pituitary. Endocrinology 1987; 121:1555. Shamgochiam MD, Leeman SE. Substance P stimulates luteinizing hormone secretion from anterior pituitary cells in culture. Endocrinology 1992; 131:871. Ahren B, Grundiz T, Ekman R, et al. Neuropeptides in the thyroid gland: distribution of substance P and gastrin/cholecystokinin and their effects on the secretion of iodothyronine and calcitonin. Endocrinology 1983; 113:379. Livett BG, Kozousek V, Mizobe F, Dean RM. Substance P inhibits nicotinic activation of chromaffin cells. Nature 1979; 278:256. Zhou X-F, Livett BG. Substance P increases catecholamine secretion from perfused rat adrenal glands evoked by prolonged field stimulation. J Physiol 1990; 425:321. Yoshida T, Mio M, Tasaka K. Cortisol secretion induced by substance P from bovine adrenocortical cells and its inhibition by calmodulin inhibitors. Biochem Pharmacol 1992; 43:513. Chiwakata C, Brackmann B, Hunt N, et al. Tachykinin (substance-P) gene expression in Leydig cells of the human and mouse testis. Endocrinology 1991; 128:2441. Pitzel L, Jarry H, Wuttke W. Effects of substance-P and neuropeptide-Y on in vitro steroid release by porcine granulosa and luteal cells. Endocrinology 1991; 129:1059. Pernow B. Substance P. Pharmacol Rev 1983; 35:85. Carlsson P-O, Sandler S, Jansson L. Influence of the neurotoxin capsaicin on rat pancreatic islets in culture, and on the pancreatic islet blood flow of rats. Eur J Pharmacol 1996; 312:75. Nilsson G, Pernow B, Fisher GH, Folkers K. Presence of substance P-like immunoreactivity in plasma from man and dog. Acta Physiol Scand 1975; 94:542. Powell D, Skrabanek P, Cannon D, Balfe A. Substance P in human plasma. In: Marsan CA, Traczyk WX, eds. Neuropeptides and neural transmission. New York: Raven Press, 1980:105. Skrabanek P, Cannon D, Kirrane J, et al. Circulating immunoreactive substance P in man. Ir J Med Sci 1976; 145:399. Hakanson R, Bengmark S, Brodin E. Substance P-like immunoreactivity in intestinal carcinoid tumors. In: von Euler US, Pernow B, eds. Substance P (Nobel symposium). New York: Raven Press, 1977:55. Emson PC, Gilbert RFT, Martensson H, Nobin A. Elevated concentrations of substance P and 5-HT in plasma in patients with carcinoid tumors. Cancer 1984; 54:714. Strodel WE, Vinik AI, Jaffe BM, et al. Substance P in the localization of a carcinoid tumor. J Surg Oncol 1984; 27:106. Zinner MJ, Yeo CJ, Jaffe BM. The effect of carcinoid levels of serotonin and substance P on hemodynamics. Ann Surg 1984; 199:197. Langdon S, Sethi T, Ritchie A, et al. Broad spectrum neuropeptide antagonists inhibit the growth of small cell lung cancer in vivo. Cancer Res 1992; 52:4554. Lotz M, Carson DA, Vaughan JH. Substance P activation of rheumatoid synoviocytes: neural pathway in pathogenesis of arthritis. Science 1987; 235:893. Lambert N, Lescoulie PL, Yassine-Diab B, et al. Substance P enhances cytokine-induced vascular cell adhesion molecule-1 (VCAM-1) expression on cultured rheumatoid fibroblast-like synoviocytes. Clin Exp Immunol 1998; 113:269. Payan DG. Neuropeptides and inflammation: the role of substance P. Annu Rev Med 1989; 40:341. Ellis JL, Undem BJ. Inhibition by capsazepine of resiniferatoxin-and capsaicin- induced contractions of guinea pig trachea. J Pharmacol Exp Ther 1994; 268:85. Joos GF, Germonpre PR, Pauwels RA. Neural mechanism in asthma. Clin Exp Allergy 2000; 30:605. Colten HR, Krause JE. Pulmonary inflammation—a balancing act. N Engl J Med 1997; 336:1094. Reynolds PN, Holmes MD, Scicchitano R. Role of tachykinins in bronchial hyper-responsiveness. Clin Exp Pharmacol Physiol 1997; 24:273. Bozic CR, Lu B, Hopken UE, et al. Neurogenic amplification of immune complex inflammation. Science 1996; 273:1722. Bertrand C, Geppetti P. Tachykinin and kinin receptor antagonists: therapeutic perspectives in allergic airway disease. TIPS 1996; 17:255. Holzer P, Holzer-Petsche U. Tachykinins in the gut. Part II. Roles in neural excitation, secretion and inflammation. Pharmacol Ther 1997; 73:219. Aronin N, Leeman SE, Clements RS Jr. Diminished flare response in neuropathic diabetic patients: comparison of effects of substance P, histamine, and capsaicin. Diabetes 1987; 36:1139. Chad DA, Aronin N, Lundstrom R, et al. Does capsaicin relieve the pain of diabetic neuropathy? Pain 1990; 42:387.

CHAPTER 172 PROSTAGLANDINS, THROMBOXANES, AND LEUKOTRIENES Principles and Practice of Endocrinology and Metabolism

CHAPTER 172 PROSTAGLANDINS, THROMBOXANES, AND LEUKOTRIENES R. PAUL ROBERTSON Arachidonic Acid Metabolism Influence of Drugs Catabolism Assay Mechanism of Action Physiology and Pathophysiology Relevant to the Endocrine System Carbohydrate Metabolism and Diabetes Mellitus Lipolysis Bone Resorption and the Hypercalcemia of Malignancy Reproductive Physiology Other Endocrine Glands Physiology and Pathophysiology Relevant to Other Body Systems Platelets Influence on Ductus Arteriosus Renal Effects Gastrointestinal Effects Pulmonary Effects Immunoregulation and Inflammation Platelet-Activating Factor Conclusion Chapter References

Few scientific areas have surpassed the field of arachidonic acid metabolism in the accumulation of new biologic information over the past three decades. Arachidonic acid and one or more of its products—prostaglandins (PG), thromboxanes (TX), and leukotrienes (LT)—have been investigated as regulators or counterregulators of virtually every organ and tissue in the mammalian body. Moreover, roles for these fatty acids have emerged and continue to be discovered for many endocrine organs and their target tissues. PG, TX, and LT are not blood-borne hormones in the classic sense. Although it is true that they have many hormone-like effects, there is no evidence that they have hemocrine characteristics. Rather, they are paracrine hormones, or autacoids, that are synthesized by the tissue in which they act and primarily provide fine modulation of ongoing cellular activity.

ARACHIDONIC ACID METABOLISM PGs were the first arachidonic acid metabolites to be discovered and were so named because they were originally identified in seminal fluid and thought to be secreted by the prostate. Arachidonic acid is stored in the membranes of cells as part of phospholipids. Phospholipase A2 activation is the mechanism for cleavage of arachidonic acid from phospholipid and initiation of the arachidonic acid cascade. A phospholipase-activating protein termed PLAP has been identified in some cells. An alternate mechanism involves phosphatidylinositol and three enzymes: phospholipase C, a diacylglycerol lipase, and a monoglyceride lipase. The cyclooxygenase and the lipoxygenase pathways are two major routes by which arachidonic acid is oxygenated. products of the cyclooxygenase pathway include all of the PG and TX. The lipoxygenase pathway generates the LT and other substances such as hydroxyeicosatetraenoic acids (HETE). Several lipoxygenase pathways, including 5-lipoxygenase and 12-lipoxygenase, are important to human physiology. The first product of the cyclooxygenase pathway is the cyclic endoperoxide prostaglandin G2 (PGG2), which is converted to prostaglandin H2 (PGH2). PGG2 and PGH2 are the key intermediates in the formation of physiologically active prostaglandins (PGD2, PGE2, PGF2a, and PGI2) and thromboxane A2 (TXA2). The first product of the 5-lipoxygenase pathway is 5-hydro-peroxyeicosatetraenoic acid (5-HPETE), which is an intermediate in the formation of 5-hydroxyeicosatetraenoic acid (5-HETE) and the leukotrienes (LTA4, LTB4, LTC4, LTD4, and LTE4). All metabolites of arachidonic acid, collectively, are called eicosanoids. Arachidonic acid metabolites in the cyclooxygenase pathway carry the subscript 2; LT derivatives in the lipoxygenase pathway carry the subscript 4. These subscripts designate the number of double bonds between carbon atoms in the chains. Eicosanoids are not stored within cells; they are synthesized rapidly according to the needs of the tissue in which they originate. Fish oils contain omega-3 fatty acids, which can decrease production of some arachidonate metabolites and increase levels of prostanoids (cyclooxygenase derivatives) with the subscript 3. INFLUENCE OF DRUGS Several groups of drugs in clinical practice, including corticosteroids and nonsteroidal antiinflammatory drugs (NSAIDs), inhibit the arachidonic acid cascade (Fig. 172-1). Although these drugs have important clinical uses and historically were of key importance to understanding the relevance of arachidonic acid metabolism to mammalian physiology, none of them specifically inhibits the synthesis of single eicosanoids. The sites of action of these drugs occur early in the arachidonic acid cascade; consequently, they influence the production of many eicosanoids simultaneously. Therefore, it is difficult to ascribe the clinical effect of any of these drugs to the absence of a particular eicosanoid. Another major limitation to full elucidation of the roles of arachidonic acid metabolites has been the paucity of drugs available to serve as specific eicosanoid receptor antagonists.

FIGURE 172-1. Arachidonic acid metabolism. In this scheme, arachidonic acid is cleaved from phospholipid by phospholipase A2 and subsequently oxygenated predominantly by cyclooxygenase or lipoxygenase. Active metabolites of the former pathway include the prostaglandins (PG) and thromboxane A2 (TXA2), whereas active metabolites of the latter pathway include the leukotrienes (LT) and hydroxyeico-satetraenoic acid (HETE). Phospholipase-activating protein (PLAP) and 5-lipoxygenase-activating protein (FLAP) are found in some cells and activate, respectively, phospholipase and 5-lipoxygenase.

There are two forms of cyclooxygenase.1,2 and 3 COX-1, a basal or constitutive form, generally plays a role in modulating physiologic activities in tissues (Fig. 172-2). The other form, COX-2, is a regulated or inducible form of the enzyme that responds to specific stimuli and is usually associated with inflammatory or noxious events in tissues. NSAIDs, as a class, were found to affect the two enzymes with differing potencies,4 and some of these drugs affect the posttranscriptional mechanisms that form the enzymes as well as their enzymatic activities (Fig. 172-2, Table 172-1). These observations provide new insight into the sites of action for these drugs. For example, aspirin and sodium salicylate can decrease the mass of COX-2 protein as well as inhibit its activity. Drugs such as aspirin, indomethacin, ibuprofen, acetaminophen, and sodium salicylate are more potent in inhibiting COX-1 than COX-2, whereas diclofenac and naproxen are more potent in inhibiting COX-2 than

COX-1. Newer, more specific inhibitors of COX-2 include rofecoxib and celecoxib with improved gas trointestinal safety profile.4a,4b Thus, the undesirable side effects of NSAIDs are due to interference with normal physiologic processes dependent on COX-1, and specific COX-2 inhibitors may offer more specific treatment of inflammatory processes with fewer side effects. Sodium salicylate, specifically, may have yet another mechanism of action involving the transcription factor nuclear factor-kappaB (NF-kB), which induces the expression of many genes involved in inflammation and infection. Sodium salicylate and aspirin inhibit the activation of NF-kB,5 which may be involved in the molecular regulation of PG synthesis through its binding site on the COX-2 gene promoter.

FIGURE 172-2. Molecular regulation of prostaglandin synthesis. The COX-1 and COX-2 genes control synthesis of the COX-1 and COX-2 enzymes, respectively. The two genes are located on separate chromosomes and give rise to mRNAs of different sizes. The two enzymes are approximately the same size but have ~60% homology. Both enzymes use arachidonic acid as substrate to generate biologically active prostanoids (PGD2, PGE2, PGF2a, PGI2, and TXA2), here represented by PGE2. Once PGE2 is formed, it can interact with any one subtype or all four subtypes of EP receptors. Cytokines, such as interleukin-1 (IL-1), and epinephrine (EPI) increase gene expression of COX-2 mRNA but do not affect COX-1 mRNA levels. Dexamethasone (DEX) diminishes the induction of COX-2 gene expression. Acetylsalicyclic acid (ASA) and sodium salicylate (SS) decrease the mass of COX-2 enzyme translated and processed. These two nonsteroidal antiinflammatory drugs (NSAIDs) and others, represented by indomethacin (INDO), also inhibit activation of the COX-1 and COX-2 enzymes. None of the NSAIDs affect expression of the COX-1 gene or translation and processing of the COX-1 enzyme. (c-AMP, cyclic adenosine monophosphate.) (Adapted from Robertson RP. Molecular regulation of prostaglandin synthesis: implications for endocrine systems. TEM 1995; 6:293.)

TABLE 172-1. IC50 Values (µg/mL) for Nonsteroidal Antiinflammatory Drugs on COX-1 and COX-2 Activity

CATABOLISM Eicosanoids are catabolized rapidly in vivo. Prostaglandin E (PGE) and prostaglandin F (PGF) are degraded almost completely during a single passage through the liver or the lung. Nonmetabolized PGE2 in the urine reflects renal and seminal vesicle production, whereas urinary PGE2 metabolites represent total body synthesis. Both PGI2 and TXA2 are also rapidly catabolized in vivo, as are the LTs. However, most of the eicosanoids have much longer chemical half-lives when stored under proper laboratory conditions. ASSAY Six methods are available to measure eicosanoids in physiologic fluids: bioassay, radioimmunoassay, chromatography, receptor assay, enzyme-linked immunoassay, and mass spectrometry. Special precautions should be taken in handling samples because eicosanoid synthesis may be stimulated during the collection procedure. For example, PGE2 and TXA2 may be generated if blood is allowed to clot or if platelets are not carefully separated from plasma. This problem can be minimized by using an inhibitor of eicosanoid synthesis in the collection tube. Measurements of inactive metabolites of PGE2, PGI2, and TXA2 (e.g., 13,14-dihydro-15-keto-PGE2, 6-keto-PGF2a, and TXB2, respectively) are commonly used for physiologic fluids because the parent substances are so short-lived in vivo. Native eicosanoids can be reliably measured in studies using in vitro experiments, such as organ incubations or cell culture. MECHANISM OF ACTION Tissues and cells with receptors for eicosanoids include adipocytes, hepatocytes, pancreatic B cells, adrenal cortex and medulla, corpus luteum, uterus, kidney, stomach, ileum, thymus, skin, brain (including pineal gland), lung, platelets, red blood cells, neutrophils, macrophages, and monocytes. These receptors are found in the plasma membrane and the nucleus and are specific for a given type of eicosanoid. Receptor density can decrease or increase in response to local increases or decreases, respectively, of the ligand for that receptor with corresponding postreceptor effects.6,7 Current nomenclature refers to eicosanoid receptors as P receptors; a preceding letter indicates the prostanoid to which each receptor is most sensitive.3 Thus, the terms DP, EP, FP, IP, and TP are used. The EP1, EP2, EP3, and EP4 receptors are coupled to transduction systems that alter phosphoinositide hydrolysis or adenylate cyclase activity (Table 172-2). Current research is focused on identification and characterization of the various types and subtypes of prostanoid receptors in tissues throughout the human body. This should help to explain how a given PG might have opposite effects on enzymatic activity, such as for adenylate cyclase in the same or different tissues under differing physiologic or pathophysiologic conditions.

TABLE 172-2. Nomenclature for Receptor Subtypes of Cyclooxygenase-Derived Prostanoids and Their Mechanisms of Action

PHYSIOLOGY AND PATHOPHYSIOLOGY RELEVANT TO THE ENDOCRINE SYSTEM The areas of endocrinology and metabolism in which eicosanoids have been investigated most intensely include carbohydrate metabolism, lipolysis, bone resorption, and reproductive physiology. However, effects of eicosanoids and drugs that prevent their synthesis have also been investigated in several other endocrine and metabolic tissues. CARBOHYDRATE METABOLISM AND DIABETES MELLITUS Reports of the involvement of PGs in carbohydrate metabolism appeared as early as the later nineteenth century. Sodium salicylate was used in the late 1800s to decrease glycosuria in diabetic patients. Very little significance was attributed to this therapy, especially because insulin was discovered in the 1920s. However, the situation changed with the discovery that NSAIDs inhibit cyclooxygenase.8 Concomitantly, it was found that PGs of the E series inhibit glucose-induced insulin secretion.9,10 and 11 This PGE effect is mediated by G proteins. It is associated with stimulation of GTPase12 and a decreased production of cyclic adenosine monophosphate (AMP) that is preventable by pertussis toxin,13 an agent that inhibits G1 and G0 activity. This inhibitory effect is specific for glucose because the insulin response to other secretagogues is not influenced by PGE2. The pancreatic islet is an exception to the current dogma about the two cyclooxygenases. Only one COX mRNA has been identified in various nonstimulated and stimulated preparations of pancreatic islets and transformed B-cell lines.14 The single form is COX-2, which is expressed both basally and under stimulated conditions. COX-2 mRNA is appropriately decreased with dexamethasone, and a COX-2–specific inhibitor blocks all PG synthesis by islets. This COX-2 dominance is associated with high levels of NF-1L6, a transcription factor that stimulates the promoter region of the COX-2 gene. Thus, it appears that under physiologic conditions, the islet tonically synthesizes PGE2, a process known to be stimulated by glucose. Since PGE2 inhibits glucose-induced insulin secretion, this high basal activity of COX-2 might serve to modulate insulin release during physiologic stimulation with glucose. Type 2 diabetics characteristically lack first-phase insulin responses to intravenous glucose, but retain responsiveness to other secretagogues. Exogenous PGE2 inhibits first-phase insulin responses to glucose stimulation.9,10 In addition, all NSAIDs (except indomethacin) consistently enhance glucose-induced insulin secretion.15 Intravenous sodium salicylate infusion doubles the basal insulin level, partially restores defective first-and second-phase insulin responses to intravenous glucose, and accelerates glucose disappearance rates after intravenous glucose challenges in type 2 diabetes.10 Exogenous PGE2 reverses the augmentation of glucose-induced insulin secretion by NSAIDs.16 Consequently, it has been hypothesized that PGs of the E series may contribute to abnormal glucose-induced insulin secretion in diabetes mellitus.17 Conflicting reports fall into two categories. The first involves studies of the effects of PGE2 itself, rather than effects of PGE 2 on glucose stimulation of insulin secretion. Sometimes, PGE2 behaves as a weak stimulator of insulin secretion if glucose concentrations are held constant. This effect appears to depend on activation of islet adenylate cyclase.18 The second category involves the use of indomethacin as an inhibitor of islet cyclooxygenase. Indomethacin not only fails to augment insulin secretion, but usually inhibits it. Although indomethacin inhibits islet cyclooxygenase, it also affects other enzyme systems and calcium flux across membranes. This multiplicity of indomethacin's effects may explain why it is discordant with other structurally unrelated inhibitors of cyclooxygenase that augment glucose-induced insulin secretion. Increased circulating levels of PGE2 and PGI2 have been reported in patients in diabetic ketoacidosis19,20 and in animals made diabetic experimentally. Consequently, it has been hypothesized that increased PGE2 production may represent the host's attempt to counterregulate the accelerated lipolysis present in ketoacidosis, and that both PGE2 and PGI2 may play roles in the decreased vascular resistance and hypotension sometimes observed in diabetic ketoacidosis. LIPOXYGENASE PRODUCTS There has been much work examining the effect of lipoxygenase products on insulin secretion.21,22 Generally, drugs that inhibit lipoxygenase inhibit glucose-induced insulin secretion. 12-HETE appears to be the major lipoxygenase product in the pancreatic islet. 12-Hydroperoxyeicosatetraenoic acid (12-HPETE), the precursor to 12-HETE, augments glucose-induced insulin secretion, whereas many other lipoxygenase products, including 12-HETE, do not. Few clinical studies that assess the effects of lipoxygenase products or inhibitors on human insulin secretion are available. DUAL ACTION OF ARACHIDONIC ACID METABOLITES The prevailing hypothesis is that arachidonic acid pathways can exert both negative and positive modulatory effects on glucose-induced insulin secretion.17 The negative modulatory effect appears to be mediated by PGE2, whereas the positive modulatory effect appears to be mediated by 12-HPETE (Fig. 172-3). It is not yet established under what physiologic or pathophysiologic conditions one or the other of these two pathways dominate to cause net negative or net positive modulatory effects.

FIGURE 172-3. Hypothetical relationship between arachidonic acid metabolites and glucose-induced insulin secretion in the pancreatic islet. The cyclooxygenase pathway is shown to have a prostaglandin E2 (PGE2)–mediated negative modulatory effect on insulin secretion, whereas the lipoxygenase pathway has a 12-hydroxyperoxyeicosatetraenoic acid (HPETE)–mediated positive modulatory effect on insulin secretion. Which physiologic and pathophysiologic conditions regulate the cyclooxygenase and lipoxygenase pathways, respectively, to yield net negative and net positive modulation of glucose-induced insulin secretion are unknown.

EFFECTS ON HEPATIC GLUCOSE PRODUCTION Eicosanoids also have been implicated in the regulation of glucose production by the liver, a major component of carbohydrate homeostasis. Hepatocytes contain receptors that are specific for PGE, and down-regulation of these receptors is associated with heterologous desensitization of hepatocyte adenylate cyclase.23 Exogenous PGE2 inhibits glucagon-induced glucose production,24 although it also has been suggested that PGE exerts a positive modulatory role on glycogenolysis.25 Little information is available about the effects of lipoxygenase products on hepatic glucose production. LIPOLYSIS The ability of PGE2 to inhibit hormone-stimulated lipolysis was one of the first physiologic actions of PGs to be discovered26: PGE2 is as potent as insulin in this regard. Also, PGE receptors were first demonstrated in fat cells.27 An early hypothesis,28,29 still being evaluated, is that fat cells synthesize PGE2 intracellularly in response to incoming hormonal signals to provide local counterregulation of hormone-induced lipolysis (Fig. 172-4). Although the inhibitory effect of PGE2 on lipolysis is easily reproduced in vitro, inconsistent effects of NSAIDs on lipolysis have raised doubts about this hypothesis. Most of the drugs used in these older studies were not specific

inhibitors of PGE2 synthesis, and how effectively many of these agents inhibit cyclooxygenase in fat cells has not been assessed.

FIGURE 172-4. Hypothetical sequence of events in fat cells in which a hormone stimulates the formation of cyclic adenosine monophosphate (cAMP) and lipolysis but, at the same time, stimulates synthesis of prostaglandin E2 (PGE2), which then produces negative feedback on adenylate cyclase to counterregulate the generation of cAMP and lipolysis. (AA, arachidonic acid; FFA, free fatty acid; TG, triglyceride.)

BONE RESORPTION AND THE HYPERCALCEMIA OF MALIGNANCY PGE2 is equipotent to parathyroid hormone (PTH) as a stimulator of bone resorption.30 After this bone-resorptive effect of PGE2 had been demonstrated, two extensive series of experiments in two different animal models31,32 were conducted. They convincingly demonstrated that tumorbearing animals had increased synthesis of PGE2, and that treatment of the animals with corticosteroids and NSAIDs decreased both PGE2 synthesis and circulating levels of calcium. These experiments led to the hypothesis that hypercalcemia associated with certain malignancies in humans might be explained by increased levels of PGE2. Initially, it was thought that ectopic (paraneoplastic) production of PTH explained most of these cases. It has become apparent, however, that this is a rare phenomenon (see Chap. 59). Greater interest in PGE2 arose when it was observed that urinary levels of PGE metabolites were elevated in hypercalcemic patients who had solid malignancies, a finding not observed in normocalcemic patients with malignancies or in hypercalcemic patients with parathyroid adenomas.31 The suggestion that tumor-associated increased synthesis of PGE2 might account for the hypercalcemia was strengthened by observations that NSAIDs could lower elevated calcium levels in patients with malignancies.33 However, it has become evident that most subjects with hypercalcemia and malignancies do not respond to these drugs.34 Nevertheless, it appears that the responders were the patients with independent evidence for elevated PGE2 levels.35 Evidence now indicates that parathyroid hormone–related protein (PTHrP) is the most common mediator of hypercalcemia in malignancy (see Chap. 52). The source of the excess PGE2 levels is not clear. One possibility is that in the presence of lung metastases, venous drainage from the tumor containing PGE2 arrives at bone through the arterial circulation without first passing through lung and liver tissue, where PGE would be degraded. Another explanation is metastatic seeding of bone. That tumor cells can synthesize PGE2 in culture suggests that metastatic tumor cells in bone could synthesize PGE2, which could act locally to resorb bone. Although hypercalcemia in malignancy can occur in the absence of demonstrable bone metastases, radioisotope scans may not be sensitive enough to detect multiple, small metastases. Alternatively, circulating white cells that collect at metastatic sites may be involved in increased PGE2 production. It is possible that there is a linkage among PTHrP, cytokines, and PGE2. Evidence for this linkage includes the following: Cytokines, including interleukin-1 (IL-1), which is known to induce bone resorption, increase PTHrP levels; IL-2 increases PTHrP production and secretion by human T-cell leukemia virus–infected T cells36; several cytokines (epidermal growth factor, transforming growth factor-a, platelet-derived growth factor) increase PGE2 synthesis and bone resorption in vivo and in vitro37,38; and indomethacin inhibits epidermal growth factor (EGF)–induced hypercalcemia in mice. These observations suggest that PGE2 may play a role in cytokine-induced hypercalcemia. In this regard, PGE1 increases PTHrP mRNA levels and PTHrP secretion in the human T-cell leukemia virus–infected T-cell line.39 REPRODUCTIVE PHYSIOLOGY LUTEOLYSIS Hysterectomy in sheep during the luteal phase is associated with luteal maintenance, suggesting that the uterus produces a luteolytic substance. PGF2a can cause luteal regression. Consequently, experiments were designed40 to assess whether PGF2a in uterine venous drainage might reach the ovary without passing through the systemic and pulmonary circulations, where it would be degraded. After the infusion of radiolabeled PGF2a into the uterine vein of sheep, the amount of radioactivity was many times higher in the ovarian arterial plasma than in the iliac arterial plasma. Consequently, a countercurrent phenomenon between the uterine vein and ovarian artery may allow PGF2a from the uterus to reach the ovary and induce luteolysis. UTERINE EFFECTS: DYSMENORRHEA PGs of the E and F series are synthesized by human endometrium. These PGs stimulate uterine contractions; therefore, their administration by intravenous infusion or tablet form has been used to initiate labor.41 The hypothesis that PGE or PGF, or both, may participate in dysmenorrhea follows from the clinical observation that women have successfully used NSAIDs to treat the discomfort associated with this syndrome. Moreover, it has been observed that PGF and PGE levels in menstrual blood are decreased in patients taking drugs that inhibit PG synthesis. In controlled trials comparing PG synthesis inhibitors with placebo in women with dysmenorrhea, symptomatic improvement has been greater after therapy with these drugs.42 OTHER ENDOCRINE GLANDS Eicosanoids have been implicated in the control of hormone secretion by several other endocrine glands (Table 172-3). For example, a number of eicosanoids stimulate hormone release43,44,45,46 and 47 from the pituitary gland (see Table 172-3). Other studies assessing the effects of NSAIDs on the secretion of these hormones have been performed. Research in these areas is insufficient to determine to what extent eicosanoids might be important modulators of pituitary gland function or might be pathophysiologic agents in pituitary disease states.

TABLE 172-3. Stimulatory and Inhibitory Actions of Eicosanoids on Endocrine Organ Function*

The observations that PGE2 decreases48 and PGF2a· increases49 PTH secretion remain uncontested. Somewhat more work has been done on the potential roles of eicosanoids as regulators of adrenal cortex function.50,51 However, this area also needs substantially more work before many of the conflicting reports can be satisfactorily resolved.

PHYSIOLOGY AND PATHOPHYSIOLOGY RELEVANT TO OTHER BODY SYSTEMS 52,53 PLATELETS It has been proposed that a balance between PGI2 and TXA2 levels modulates platelet aggregation.54 Platelets synthesize TXA2, which is a potent stimulator of aggregation. PGI2 is synthesized by endothelial cells of blood vessels, and it potently antagonizes platelet aggregation. It is thought that TXA2 and PGI2 exert their opposing effects by decreasing and increasing, respectively, platelet generation of cyclic AMP. It follows that endothelial damage, and a concomitant decrease in PGI2 synthesis, may allow unbridled platelet aggregation at the site of vessel wall damage. It has been hypothesized that this situation would favor the eventual development of atherosclerosis. These considerations have led to the therapeutic use of aspirin to suppress platelet aggregation. Cyclooxygenase inhibition by a single dose of aspirin is of longer duration in platelets than in other tissues because the platelet, in contrast to nucleated cells that can synthesize new proteins, does not have the nuclear machinery to form new cyclooxygenase. Consequently, the effect of aspirin persists until newly formed platelets have been released from bone marrow. On the other hand, endothelial cells rapidly recover cyclooxygenase activity after aspirin ingestion, thereby restoring PGI2 production. INFLUENCE ON DUCTUS ARTERIOSUS Many eicosanoids are vasoactive substances. PGE2 and PGI2 are vasodilators, whereas PGF2, TXA2, and LTC4, LTD4, and LTE4 are vasoconstrictors in most vascular beds. PGE2 has been postulated to be a primary factor in the maintenance of physiologic patency of the ductus arteriosus in the fetus. Trials with NSAIDs have been undertaken in newborn human infants with patent ductus arteriosus that caused closure of this vessel.55 Infants younger than 35 weeks of age are most likely to respond, and some individuals require a second course of therapy. RENAL EFFECTS GENERAL EFFECTS Arachidonic acid metabolites influence both the reninangiotensin– aldosterone system and the vasopressin system56,57 (see Chap. 25 and Chap. 183). Both PGE2 and PGI2 stimulate renin secretion. They also decrease renal vascular resistance and increase blood flow. The NSAIDs decrease total renal blood flow and can lead to acute renal vasoconstriction and decreased renal function in some circumstances, such as in volume depletion, in edematous states, and in older patients. Indomethacin increases sensitivity to exogenous vasopressin, and, conversely, PGE2 decreases vasopressin-stimulated water transport. BARTTER SYNDROME Bartter syndrome is characterized by increased levels of plasma renin, aldosterone, and bradykinin; resistance to the pressor effect of angiotensin infusion; hypokalemic alkalosis; and renal potassium wasting in the presence of normal blood pressure (see Chap. 80). It has been postulated that excessive PGE2 or PGI2 synthesis plays a role in this syndrome.58 Both PGE2 and PGI2 stimulate the release of renin and blunt the pressor effects of angiotensin. Elevated levels of PGE2 and PGI2 metabolites have been found in the urine of patients with this syndrome. Therapeutic trials with NSAIDs have reversed virtually all of these clinical abnormalities except hypokalemia. Consequently, it has been concluded that a PG, presumably PGE2 or PGI2, may be responsible for mediating many of the manifestations of Bartter syndrome, although excessive PG levels themselves are not the primary defect. GASTROINTESTINAL EFFECTS PGE, taken orally, protects the gastrointestinal mucosa from several forms of injury by a direct, cytoprotective effect.59 It also inhibits gastric acid secretion. Because gastric acid secretion is excessive in patients with peptic ulcer disease, analogs of PGE2 have been used as therapeutic agents.54 These agents are more effective than placebo in relieving pain and decreasing gastric acid secretion in patients with peptic ulcer disease. The healing of ulcer craters is accelerated in patients treated with these analogs compared with placebo-treated patients.60 PULMONARY EFFECTS Arachidonic acid metabolites may play a role in the clinical manifestations of allergic and drug-associated asthma.61 Many arachidonic acid metabolites can be formed by lung tissue. PGF2a, TXA2, LTC4, LTD4, and LTE4 are potent bronchoconstrictors, whereas PGE2 is a potent bronchodilator. The effect of PGF 2 is antagonized by PGE2 and by catecholamines, but not by atropine, antihistamines, serotonin antagonists, or a-adrenergic antagonists. An interesting subset of asthmatic patients have symptoms precipitated by drugs such as aspirin and indomethacin—the syndrome of aspirin-sensitive asthma. If arachidonic acid metabolites play a role in this particular syndrome, it might involve the inhibition of cyclooxygenase and shunting of substrate to the LT pathway, thereby inducing formation of large amounts of bronchoconstrictor substances. IMMUNOREGULATION AND INFLAMMATION INFLAMMATORY RESPONSE Much information about potential roles of eicosanoids in the immune response continues to accumulate.62 It has been recognized that small amounts of PGE2 suppress stimulation of human lymphocytes by mitogens. Moreover, the inflammatory response usually is associated with the local release of arachidonic acid metabolites. This has led to the hypothesis that eicosanoids act as negative modulators of lymphocyte function. Release of PGE by mitogen-stimulated lymphocytes is envisioned as a negative-feedback control mechanism by which lymphocyte activity is regulated. Indomethacin augments lymphocyte responsiveness to mitogens. Several lines of evidence support a relationship between inflammation and the generation of arachidonic acid metabolites. Inflammatory stimuli, such as histamine and bradykinin, release PGs. LTC4, LTD4, and LTE4 are more potent than histamine as bronchoconstrictors. PGE2 and LTD4 are commonly present in areas of inflammation. During phagocytosis, polymorphonuclear cells release eicosanoids that are chemotactic for leukocytes. Vasodilatation induced by PGE is not abolished by antagonists of known mediators of the inflammatory response such as atropine, propranolol, methysergide, or antihistamines. Thus, it has been postulated that PGE and other eicosanoids may have direct inflammatory effects, and other mediators of inflammation may act by influencing eicosanoid release. Mastocytosis and other disorders related to mast cell activation (see Chap. 181) have been related to excessive productionof PGD2.63 Symptoms in these syndromes include flushing, tachycardia, hypotension, abdominal cramping, diarrhea, chest pain, headache, dyspnea, and itching. Use of NSAIDs in combination with antihistaminic drugs has been reported to be effective therapy for mastocytosis. However, a small subset of patients have attacks that are provoked by aspirin. RHEUMATOID ARTHRITIS The inflammatory response and bone resorption that accompany rheumatoid arthritis may depend, to some degree, on the local generation of eicosanoids. Rheumatoid synovia synthesize PGE2 in tissue culture, and media from these cultures promote bone resorption. The inclusion of indomethacin in the culture medium blocks this bone-resorptive capacity, but does not prevent bone resorption caused by exogenous PGE2. Hence, it has been postulated that PGE2 produced by the synovia may be responsible for the bone resorption seen in patients with rheumatoid arthritis.64 A site of action for glucocorticoids has been identified in experiments using synovial tissue from patients with rheumatoid arthritis; that is, COX-2 mRNA and COX-2 protein are markedly suppressed by dexamethasone.65 This observation is important

because previously glucocorticoids were thought to act on arachidonic acid metabolism exclusively by inhibiting phospholipase A2 activity. PLATELET-ACTIVATING FACTOR A related compound to arachidonic acid is platelet-activating factor (PAF). PAF is a phospholipid synthesized by the hydrolysis of arachidonic acid. This phospholipid autacoid has a wide range of physiologic actions, including platelet activation, increased vascular permeability, uterine contraction, and several effects on reproductive function.66

CONCLUSION Arachidonic acid metabolites are important modulators of physiologic activity in many tissues, and eicosanoids may play a role in the pathogenesis of human disease. These metabolites and their inhibitors may have important therapeutic applications (e.g., metabolites [peptic ulcer disease, ductus-dependent congenital heart disease, hypertension,67 induction of labor, excessive lipolysis] and inhibitors [rheumatoid arthritis, hypercalcemia of cancer, inflammation, fever, dysmenorrhea, Bartter syndrome, type 2 diabetes mellitus, anticoagulation, patent ductus arteriosus]). The development and use of drugs that specifically and selectively inhibit the synthesis of single eicosanoids and that selectively antagonize receptors for specific eicosanoids are beginning to yield new and important physiologic and pathophysiologic information. CHAPTER REFERENCES 1. Robertson RP. Molecular regulation of prostaglandin synthesis: implications for endocrine systems. TEM 1995; 6:293. 2. Smith WL, Dewitt DL. Prostaglandin endoperoxide H synthases-1 and 2. Adv Immunol 1966; 62:167. 3. Coleman RA, Smith WL, Narumiya S. VIII. International union of pharmacology classification of prostanoid receptors: properties, distribution, and structure of the receptors and their subtypes. Pharmacol Rev 1994; 46:205. 4. Mitchell JA, Akarasereenont P, Thiemermann C, et al. Selectivity of nonsteroidal antiinflammatory drugs as inhibitors of constitutive and inducible cyclooxygenase. Proc Natl Acad Sci U S A 1994; 90:11,693. 4a.

Clemett D, Goa KL. Celecoxib: a review of its use in osteoarthritis, rheumatoid arthritis, and acute pain. Drugs 2000; 59:957.

4b.

Cannon GW, Caldwell JR, Holt P, et al. Rofecoxib, a specific inhibitor of cyclooxygenase–2. Arthritis Rheum 2000; 43:978.

5. Kopp E, Ghosh S. Inhibition of NF-kB by sodium salicylate and aspirin. Science 1994; 265:956. 6. Robertson RP, Westcott KR, Storm DR, Rice MG. Down-regulation in vivo of prostaglandin E receptors and adenylate cyclase stimulation in rat liver plasma membranes. Am J Physiol 1980; 239:E75. 7. Robertson RP, Little SA. Down-regulation of prostaglandin E receptors and homologous desensitization of isolated adipocytes. Endocrinology 1983; 113:1732. 8. Vane JR. Inhibition of prostaglandin synthesis as a mechanism of action for aspirin-like drugs. Nature 1971; 231:232. 9. Robertson RP, Gavareski DJ, Porte D Jr, Bierman EL. Inhibition of in vivo insulin secretion by prostaglandin E1. J Clin Invest 1974; 54:310. 10. Robertson RP, Chen M. A role for prostaglandin E (PGE) in defective insulin secretion and carbohydrate intolerance in diabetes mellitus. J Clin Invest 1977; 60:747. 11. Burr IM, Sharp R. Effects of prostaglandin E 1 and of epinephrine on the dynamics of insulin release in vitro. Endocrinology 1974; 94:835. 12. Kowluru A, Metz SA. Stimulation of prostaglandin E 2 of a high-affinity GTPase in the secretory granules of normal rat and human pancreatic islets. Biochem J 1994; 297:399. 13. Robertson RP, Tsai P, Little SA, et al. Receptor-mediated adenylate cyclase-coupled mechanism for PGE2 inhibition of insulin secretion in HIT cells. Diabetes 1987; 36:1047. 14. Sorli CH, Zhang H-J, Armstrong MB, et al. Basal expression of cyclooxyge-nase-2 and NIF-1L6 are dominant and coordinately regulated by IL-1 in the pancreatic islet. Proc Natl Acad Sci U S A 1998; 95:1788. 15. Robertson RP. Arachidonic acid metabolite regulation of insulin secretion. Diabetes Metab Rev 1986; 2:261. 16. Metz SA, Robertson RP, Fujimoto WF. Inhibition of prostaglandin E synthesis augments glucose-induced insulin secretion in cultured pancreas. Diabetes 1981; 30:551. 17. Robertson RP. Eicosanoids as pluripotential modulators of pancreatic islet function. Diabetes 1988; 37:367. 18. Johnson DG, Fujimoto WY, Williams RH. Enhanced release of insulin by prostaglandins in isolated pancreatic islets. Diabetes 1973; 22:658. 19. McRae JR. Day RP, Metz SA, et al. Prostaglandin E2 metabolite levels during diabetic ketoacidosis. Diabetes 1985; 34:761. 20. Axelrod L, Shulman GI, Blackshear PJ, Bomstein W, et al. Plasma level of 13, 14-dihydro-15-keto-PGE2 in patients with diabetic ketoacidosis and in normal fasting subjects. Diabetes 1986; 35:1004. 21. Metz S, VanRollins M, Strife R, et al. Lipoxygenase pathway in islet endocrine cells. J Clin Invest 1983; 71:1191. 22. Yamamoto S, Nakadate T, Nakaki T, et al. Prevention of glucose-induced insulin secretion by lipoxygenase inhibitor. Eur J Pharmacol 1982; 78:225. 23. Garrity MJ, Andreason TJ, Storm DR, Robertson RP. PGE-induced heterologous desensitization of hepatic adenylate cyclase: consequences on the guanyl nucleotide regulatory complex. J Biol Chem 1983; 258:8692. 24. Brass EP, Garrity MJ, Robertson RP. Inhibition of glucagon-stimulated hepatic glycogenolysis by E-series prostaglandins. FEBS Lett 1984; 169:293. 25. Ganguli S, Sperling MA, Frame E, Christensen R. Inhibition of glucagon-induced hepatic glucose production by indomethacin. Am J Physiol 1979; 236:E58. 26. Steinberg D, Vaughn M, Nestel PJ, et al. Effects of the prostaglandins on hormone-induced mobilization of free fatty acids. J Clin Invest 1964; 43:1533. 27. Kuehl FA, Humes JL. Direct evidence for a prostaglandin receptor and its application to prostaglandin measurements. Proc Natl Acad Sci U S A 1972; 69:480. 28. Shaw JE, Ramwell PW. Release of prostaglandin from rat epididymal fat pad on nervous and hormonal stimulation. J Biol Chem 1968; 243:1498. 29. Christ EJ, Nugteren DH. The biosynthesis and possible function of prostaglandins in adipose tissue. Biochim Biophys Acta 1970; 218:296. 30. Klein DC, Raisz LG. Prostaglandins: stimulation of bone resorption in tissue culture. Endocrinology 1970; 86:1436. 31. Tashjian AH Jr, Voelkel EF, Levine L, Goldhaber P. Evidence that the bone resorption-stimulating factor produced by mouse fibrosarcoma cells is prostaglandin E 2: a new model for the hypercalcemia of cancer. J Exp Med 1972; 136:1329. 32. Voelkel EF, Tashjian AH Jr, Franklin R, et al. Hypercalcemia and tumor prostaglandins: the VX 2 carcinoma model in the rabbit. Metabolism 1975; 24:973. 33. Seyberth HW, Segre GV, Morgan JL, et al. Prostaglandins as mediators of hypercalcemia associated with certain types of cancer. N Engl J Med 1975; 293:1278. 34. Robertson RP, Baylink DJ, Metz SA, Cummings KB. Plasma prostaglandin E in patients with cancer with and without hypercalcemia. J Clin Endocrinol Metab 1976; 43:1330. 35. Metz SA, McRae JR, Robertson RP. Prostaglandins as mediators of paraneoplastic syndromes: review and update. Metabolism 1981; 30:299. 36. Ikeda K, Okazaki R, Inoue D, et al. Interleukin-2 increases production and secretion of parathyroid hormone-related peptide by human T cell leukemia virus type-I-infected T cells: possible role in hypercalcemia associated with adult T cell leukemia. Endocrinology 1993b; 132:2551. 37. Tashjian AH Jr, Hohmann EL, Antoniades HN, Levine L. Platelet-derived growth factor stimulates bone resorption via a prostaglandin-mediated mechanism. Endocrinology 1982; 111:118. 38. Tashjian AH Jr, Voelkel EF, Lloyd W, et al. Actions of growth factors on plasma calcium: epidermal growth factor and human transforming growth factoralpha cause elevation of plasma calcium in mice. J Clin Invest 1986; 78:1405. 39. Ikeda K, Okazaki R, Inoue D, et al. Transcription of the gene for parathyroid hormone-related peptide from the human is activated through a cAMP-dependent pathway by prostaglandin E 1 in HTLV-I-infected T cells. J Biol Chem 1993a; 268:1174. 40. McCracken JA, Baird DT, Goding JR. Factors affecting the secretion of steroids from the transplanted ovary in the sheep. Recent Prog Horm Res 1971; 27:537. 41. Casey C, Kehoe J, Mylotte MJ. Vaginal prostaglandins for the ripe cervix. Int J Gynaecol Obstet 1994; 44:21. 42. Budoff PW. Zomspirac sodium in the treatment of primary dysmenorrhea syndrome. N Engl J Med 1982; 307:714. 43. Ojeda SR, Harms PG, McCann SM. Central effect of prostaglandin E1 (PGE1) on prolactin release. Endocrinology 1974; 95:613. 44. Naor Z, Vanderhoek JY, Lindner HR, Catt KJ. Arachidonic acid products as possible mediators of the action of gonadotropin-releasing hormone. Adv Prostaglandin Thromboxane Leukotriene Res 1983; 12:259. 45. Brown MR, Hedge GA. In vivo effects of prostaglandins on TRH-induced TSH secretion. Endocrinology 1974; 95:1392. 46. Drouin J, Labrie F. Specificity of the stimulatory effect of prostaglandins on hormone release in rat anterior pituitary cells in culture. Prostaglandins 1976; 11:355. 47. Hedge GA. Hypothalamic and pituitary effects of prostaglandins on ACTH secretion. Prostaglandins 1976; 11:293. 48. Gardner DG, Brown EM, Windeck R, Aurbach GD. Prostaglandin E2 stimulation of adenosine 3',5'-monophosphate accumulation and parathyroid hormone release in dispersed bovine parathyroid cells. Endocrinology 1978; 103:577. 49. Gardner DG, Brown EM, Windeck R, Aurbach GD. Prostaglandin F 2a inhibits 3',5'-adenosine monophosphate accumulation and parathyroid hormone release from dispersed bovine parathyroid cells. Endocrinology 1979; 104:1. 50. Matsuoka H, Tan SY, Mulrow PJ. Effects of prostaglandins on adrenal steroidogenesis in the rat. Prostaglandins 1980; 19:291. 51. Carchman RA, Shen JC, Bilgin S, Rubin RP. Diverse effects of Ca2+ on the prostacyclin and corticotropin modulation of adenosine 3',5'-monophosphate and steroid production in normal cat and mouse tumor cells of the adrenal cortex. Biochem Pharmacol 1980; 29:2213. 52. Robertson RP, ed. Symposium on prostaglandins in health and disease. Med Clin North Am 1981; 65:711. 53. Zipser RD, Laffi G. Prostaglandins, thromboxanes and leukotrienes in clinical medicine. West J Med 1985; 143:485. 54. Moncada S, Vane JR. Arachidonic acid metabolites and the interactions between platelets and blood vessel walls. N Engl J Med 1979; 300:1142. 55. Van Overmeire B, Smets K, Lecoutere D, et al. A comparison of ibuprofen and indomethacin for closure of patent ductus arteriosus. N Engl J Med 2000; 343:674. 56. Ferris TF. Prostaglandins and the kidney. Am J Nephrol 1983; 13:139. 57. Orloff J, Handler JS, Bergstrom S. Effect of prostaglandin (PGE1) on the permeability response of the toad bladder to vasopressin, theophylline and adenosine 3',5'-monophosphate. Nature 1965; 205:397. 58. Gill JR. Frolich JC, Bowden RE, et al. Bartter's syndrome: a disorder characterized by high urinary prostaglandins and a dependence of hyperreninemia on prostaglandin synthesis. Am J Med 1976; 61:43. 59. Redfern JS, Feldman M. Role of endogenous prostaglandins in preventing gastrointestinal ulceration: induction of ulcers by antibodies to prostaglandins. Gastroenterology 1989; 96:596. 60. Vantrappen G, Janssens J, Popiela T, et al. Effect of 15(R)-15-methylprostaglandin E 2 (arbaprostil) on the healing of duodenal ulcer. Gastroenterology 1982; 83:357. 61. Hyman AL, Mathe AA, Lippton HL, Kadowitz PJ. Prostaglandins and the lung. Med Clin North Am 1981; 65:789. 62. Goetzl E, Scott WA, eds. Regulation of cellular activities by leukotrienes. J Allergy Clin Immunol 1984; 74:309. 63. Roberts LJ 2d. Carcinoid syndrome and disorders of systemic mastcell activation including systemic mastocytosis. Endocrinol Metab Clin North Am 1988; 17:415. 64. Robinson DR, Tashjian AH, Levine L. Prostaglandin-stimulated bone resorption by rheumatoid synovia. J Clin Invest 1975; 56:1181. 65. Crofford LJ, Wilder RL, Ristimaki AP, et al. Cyclooxygenase-1 and -2 expression in rheumatoid synovial tissues: effects of interleukin-b, phorbol ester, and corticosteroids. J Clin Invest 1994; 93:1095. 66. Venable ME, Zimmerman GA, Maclntyre TM, Prescott SM. Platelet-activating factor: a phospholipid autocoid with diverse actions. J Lipid Res 1993; 34:691. 67. Hoeper MM, Schwarze M, Ehlerding S, et al. Long-term treatment of primary pulmonary hypertension with aerosolized iloprost; a prostacyclin analogue. N Engl J Med 2000; 342:1866.

CHAPTER 173 GROWTH FACTORS AND CYTOKINES Principles and Practice of Endocrinology and Metabolism

CHAPTER 173 GROWTH FACTORS AND CYTOKINES DEREK LEROITH AND VICKY A. BLAKESLEY Insulin-Like Growth Factor System Insulin-Like Growth Factors Insulin-Like Growth Factor Receptors Insulin-Like Growth Factor–Binding Proteins Insulin-Like Growth Factors in Health and Disease Normal Growth and Development Growth Disorders Diabetes Cancer Nervous System Immune System Reproductive System Bone Potential Clinical Uses Other Growth Factors Vascular Endothelial Growth Factor Nerve Growth Factor Family Platelet-Derived Growth Factors Epidermal Growth Factor Family Inhibin, Activin, and Müllerian-Inhibiting Hormone Fibroblast Growth Factor Family Cytokines Interleukins Interferons Cytokine Receptors Chapter References

INSULIN-LIKE GROWTH FACTOR SYSTEM This chapter on growth factors begins with a discussion of the insulin-like growth factors (IGFs), their receptors and binding proteins. This format reflects the authors' belief that the IGF system represents an excellent paradigm for both growth factors and cytokines, and, furthermore, is one of the more clinically relevant growth factor families for the academic and practicing endocrinologist. Descriptions are given of structural and functional aspects of the ligands, receptors, and binding proteins, followed by examples of clinically relevant aspects of the IGF system.1 Subsequently, other growth factors and cytokines are discussed, with emphasis on their potential roles in clinically related disorders. A list of the families of growth factors is presented in Table 173-1. A more complete list of the cytokine families can be found in Chapter 174.

TABLE 173-1. Growth Factor Families

INSULIN-LIKE GROWTH FACTORS IGF-I and IGF-II are small (~70 amino acids) secreted peptides that are structurally related to proinsulin. In mammals they are encoded by large, complex, single-copy genes, homologues of which are found in all vertebrate species.2 As with most secreted proteins, they are synthesized as precursors containing N- and C-terminal extensions. The N-terminal signal peptide directs the IGFs to the constitutive secretory pathway, and the C-terminal E-peptide moiety of the prohormone is cleaved during transit through the Golgi network to yield the mature molecule. The mature IGF-I peptide consists of B and A domains, of which the amino-acid sequences are homologous to the B and A chains of insulin, linked by a short C domain. IGFs also contain a short, C-terminal D domain for which no analogy exists in the mature insulin molecule. Most circulating IGFs occur in the fully processed form, without the E peptide. Under certain conditions, the prohormone can be detected in the serum with the E peptide as part of the IGF molecule (Fig. 173-1A).

FIGURE 173-1. A, The insulin-like growth factor (IGF) family of peptides. Insulin, IGF-I, and IGF-II are processed from precursor molecules. In the case of insulin, the C peptide is removed, whereas the IGF-I and IGF-II molecules retain the C peptide and, in addition, contain a D extension at the carboxyl terminal ends of their A chains. The E peptide found in the proIGF-I and proIGF-II molecules is cleaved during processing. A chains are in bold, C peptides and D extensions are shown in double lines, and the B chain is represented as a single line. B, The insulin-like growth factor receptors. Schematic representation of the IGF-I receptor. The IGF-I receptor is a heterotetrameric transmembrane tyrosine kinase receptor. Each hemireceptor is comprised of an a and a b subunit covalently bound by disulfide bonds. A doublet of hemireceptors constitutes the active IGF-I receptor. The a subunit is entirely extracellular and contains the ligand (IGF-I or -II)–binding domain. The b subunit spans the membrane and is divided into the extracellular, transmembrane, juxtamembrane, tyrosine kinase, and C-terminal domains. The essential adenosine triphosphate (ATP)–binding site is at lysine 1003 within the tyrosine kinase domain. Also within the catalytic domain is a triple tyrosine cluster (tyrosines 1131, 1135, and 1136).

IGF-I and IGF-II messenger ribonucleic acid (mRNA) or protein can be detected at some developmental stage in every tissue examined. In rodents, the expression of

the IGF-II gene is highest prenatally, with expression decreasing dramatically postnatally. IGF-I gene expression, while present in fetal tissues, tends to increase dramatically postnatally, especially during the peripubertal period when growth hormone (GH) responsiveness develops.3 In humans, however, circulating IGF-II levels remain high even in adults. These distinct patterns of expression suggest that IGF-I and IGF-II serve unique functions during development. In humans, GH exerts a major influence on IGF-I gene expression, particularly in the liver, consistent with a classic endocrine pathway. The production of IGF-I and IGF-II by most tissues is consistent with their roles as autocrine/paracrine growth factors, in addition to their obvious roles as classic endocrine agents. In support of this view, IGF-I gene expression is also modulated by numerous hormones, nutrients, and molecules. The control of IGF-II gene expression is similarly complex, with various mRNA species differentially expressed at different stages of development and in different tissues. The IGF-II promoter is imprinted, with only one allele normally being expressed. It is also controlled by tumor-suppressor gene products (e.g., Wilms tumor WT1) and the adjacent H19 gene. Dysregulation of these control mechanisms may explain its frequent overexpression by multiple tumors. Both IGF-I and IGF-II gene expression are subject to complex control mechanisms consistent with their postulated roles of fetal and postnatal growth factors.4 The importance of the IGFs in normal growth and development has been addressed by an elegant series of gene-targeting studies in which the IGF-I and IGF-II genes, singly and in combination, have been inactivated.5 Homozygous IGF-I– and IGF-II– deficient mice exhibited a 30% reduction in body growth during embryonic development.6 While the IGF-II–deficient mice continued to grow at this reduced level postnatally, IGF-I– deficient mice died immediately after birth. Double mutants (generated by crossing) exhibited an additive 60% decrease in growth. Both IGF-I and IGF-II appear to be necessary for full growth potential. As demonstrated by more recent models of IGF-I–deficient mice, IGF-I also plays an important role for specific tissue function. These IGF-I–deficient mice show improved postnatal survival but lack the normal peripubertal growth spurt due to absent responsiveness to GH-induced IGF-I production. The mice are also infertile. These in vivo models show that the IGFs are important growth regulators for which synthesis is subject to complex control mechanisms. INSULIN-LIKE GROWTH FACTOR RECEPTORS The actions of the IGFs require their interaction with specific cell-surface receptors. The family of receptors involved in the bioactions of the IGFs include the insulin receptor, the IGF-I receptor (type I IGF receptor), the IGF-II/mannose-6-phosphate receptor (type II IGF receptor), the insulin-receptor–related receptor (IRR), and hybrid receptors (consisting of an insulin a and b subunit and an IGF-I receptor a and b subunit).7 The relative binding affinities of these receptors for their potential ligands are delineated in Table 173-2.

TABLE 173-2. Relative Affinities of Insulin-Like Growth Factor Ligands for Their Receptors

The IGF-I and insulin receptors are structurally similar molecules encoded by distinct genes.8 Both consist of two extracellular glycosylated a subunits containing cysteine-rich regions. These cysteine-rich regions in the IGF-I receptor are primarily responsible for IGF binding, but ligand binding to the insulin receptor requires regions flanking the cysteine-rich domain. The a subunits are joined to each other and to the transmembrane b subunits by disulfide bonds that are formed after proteolytic cleavage of the proreceptor into a and b subunits in the Golgi apparatus. The b subunit consists of a short extracellular domain, a transmembrane region, a cytoplasmic juxtamembrane region in which tyrosine-containing motifs important for receptor internalization and interaction with endogenous substrates occur, a tyrosine kinase domain that includes an adenosine triphosphate (ATP)–binding motif, a cluster of tyrosine residues subject to autophosphorylation, and a C-terminal domain containing several tyrosine residues that are also auto-phosphorylated following activation of the receptor. The tyrosine kinase domain is the most highly conserved (~87%), while the C-terminal domain is the least conserved between the IGF-I and insulin receptors. The IGF-I and insulin receptor are members of a receptor tyrosine kinase family that includes the epidermal growth factor (EGF) and platelet-derived growth factor (PDGF) receptors, but they are atypical in that they occur in a heterotetrameric structure and are considered to be “pre-dimerized,” unlike the other receptors of this family that require ligand binding for dimerization and activation (Fig. 173-1B). The IGF-II/mannose-6-phosphate receptor is unrelated structurally to the IGF-I and insulin receptors, because it contains a long extracellular domain consisting of numerous repeats of a short amino-acid sequence, a transmembrane domain, and a short cytoplasmic tail.9 The extracellular domain contains binding sites for IGF-II and mannose-6-phosphate, and ligand binding at one site influences ligand binding at the other. This receptor is important in the uptake and intracellular trafficking of mannose-6-phosphate– containing lysosomal enzymes and may play a role in the clearance of IGF-II from the circulation. The cytoplasmic tail lacks any obvious catalytic activity or classic binding sites for known endogenous substrates and—in the absence of any convincing evidence to the contrary—has generally been considered to lack signaling properties. Proteolytic cleavage of the extracellular domain releases a circulating form of the receptor, which may then act as a binding protein for IGF-II. Since the IGF-I receptor is considered to mediate the actions of the IGFs, functional aspects of this receptor are dealt with in more detail in the following section. INSULIN-LIKE GROWTH FACTOR-I RECEPTOR The IGF-I receptor gene is expressed ubiquitously in most tissues.10 The gene contains 21 exons, and two mRNA species can be found in human tissues, one of ~11 kilobases (kb) and a second of ~7 kb.11 The mRNA contains an ~1-kb 5‹ untranslated (UT) region, a coding region of ~5 kb, and a 3‹ UT region of ~5 kb. Both the 5‹ UT and 5‹ flanking regions of the gene are enriched in GC sequences, and the expression is regulated by SP1 and WT1 transcription factors that bind specifically to these GC-rich boxes in the promoter region.12 This region is also responsible for the regulation of the IGF-I receptor gene by growth factors.13,14 EGF, PDGF, and bFGF up-regulate gene expression, whereas IGF-I down-regulates its own receptor at the transcriptional level.15,16 The 3‹ UT region most likely is involved in the stability of the mRNA. The sequence of events initiated by ligand binding to receptor tyrosine kinases has been elucidated for a number of these receptors.17 In the case of the IGF-I receptor, ligand binding to the a subunit alters the conformation of the extracellular and transmembrane portions of the receptor, leading to a change in the cytoplasmic domain, which in turn results in activation of the tyrosine kinase and phosphorylation of a cluster of three tyrosine residues in the kinase domain itself.18,19 Subsequently, tyrosines in the juxtamembrane and C-terminal domains are phosphorylated along with other substrates.20 The insulin receptor substrate (IRS) family of proteins, of which four have been characterized together with the protein Shc, represent the most well studied of the substrates involved in IGF-I and insulin-receptor signal transduction.21 Their primary site of interaction is the phosphorylated tyrosine at residue 960 within the juxtamembrane domain of the IGF-I receptor. IGF-I–receptor activation results in tyrosine phosphorylation of IRS (1–4) and Shc.22,23 and 24 Phosphorylated tyrosine residues on the IRS molecules are found in consensus sequences for the binding of Src homology domain (SH2) domains of several proteins of the signal transduction pathways.25 These include the p85 subunit of phosphoinositide3‹kinase (PI3‹K), growth factor receptor bound protein-2 (Grb-2), the tyrosine phosphatase Syp, Crk adapter proteins, and Nck26 (Fig. 173-2). PI3‹K activation leads to pathways involved in preventing apoptosis (via protein kinase B/Akt), signaling to focal adhesion sites and even mitogenesis.27 Grb-2 is an adapter protein that interacts via its SH2 domain with both tyrosine phosphorylated IRS and Shc.28 Bound Grb-2 then interacts via its SH3 domain with the mammalian homologue of son-of-sevenless (mSOS), which contains the appropriate proline-rich region required for SH3 domain interaction.29,30 This interaction recruits mSOS to the plasma membrane where mSOS, a guanine nucleotide exchange factor, activates p21ras by facilitating the displacement of guanine diphosphate (GDP) from the guanine nucleotide-binding site of p21ras and its replacement with guanine triphosphate (GTP). GTP-p21ras interacts with raf proteins and thereby activates the mitogen-activated protein (MAP) kinase/Erk pathway. The final result is the transport of active molecules into the nucleus and enhancement of specific gene transcription, leading to increased mitogenic signals and cellular proliferation. These major pathways are the ones that have been most clearly characterized for IGF-I–receptor signaling. Additionally, a host of other substrates and pathways have been delineated.

FIGURE 173-2. Signal transduction by the activated insulin-like growth factor-I (IGF-I) receptor. Similar to most tyrosine kinase transmembrane receptors, the IGF-I receptor has multiple substrates and pathways via which it mediates its cellular responses. Important substrates include SHC and the insulin receptor substrate (IRS) family of proteins. These interact with downstream substrates to result eventually in cellular proliferation and/or inhibition of apoptosis. Akt, a serine/threonine kinase, also known as PKB (protein kinase B); Fyn, an adapter protein encoded by c-fyn oncogene; IGF-IR, insulin-like growth factor-I receptor; IRS 1-4, insulin receptor substrates 1-4; GDP, guanosine diphosphate; GLUT4, glucose transporter 4; GSK3, glycogen synthase kinase 3; GTP, guanosine triphosphate; JNK, c-Jun NH2-terminal kinase, also known as SAPK1 (stess-activated protein kinase 1); MAPK, mitogen-activated protein kinase; MEK or MAPK/Erk 1, MAPK/extracellular signal-regulated protein kinase 1; Nck, a SH2- and SH3-containing adapter protein; p38 MAPK, p38 mitogen-activated protein kinase, also known as SAPK2 (stess-activated protein kinase 2); PDK1 and 2, phosphatidylinositol-dependent kinase 1 and 2; PI3-kinase, phosphoinositide 3-kinase; PI-3,4-P2, phosphatidylinositol-3,4-bisphosphate; PI-3,4,5-P3, phosphatidylinositol-3,4,5-trisphosphate; PI-4,5-P2, phosphatidylinositol-4,5-bisphosphate; PP2A, protein phosphatase 2A; Raf, a serine/threonine kinase encoded by c-raf oncogene; Ras, a guanine nucleotide binding protein encoded by H-ras oncogene; SEK1, stress signaling kinase 1, also known as MKK4 (MAP kinase kinase 4); Shc, src homology/collagen protein; Syp, a SH2-containing protein tyrosine phosphatase, also known as SH-PTP2 or PTP1D (src homology 2-binding tyrosine phosphatase 1D); SOS, a guanylnucleotide exchange factor that is the mammalian homologue of son-of-sevenless.

The Crk family of adapter proteins, CrkI and II and CrkL, while lacking catalytic activity, play important roles in the signal transduction pathways of growth factor receptors including the IGF-I receptor.31,32 IGF-I–receptor activation results in tyrosine phosphorylation of both CrkII and CrkL, which probably occurs by the association of these molecules with members of the IRS family of proteins. CrkII is involved in the differentiated function of the IGF-I receptor, whereas CrkL results in the more transforming phenotype.33,34 and 35 Examples of recently described substrates involved in IGF-I receptor-signaling cascades include the vav protooncogene in hematopoietic cells, two isoforms of the 14-3-3 family of proteins that interact directly with the IGF-I receptor, protein kinase C isoforms (i.e., PKC a), the G protein Gbg, focal adhesion kinase, Janus kinase 1 (JAK1), signal transducers and activators of transcription (Stat) 3, and suppressor of cytokine signaling (SOCS-2). IGF-I–receptor activation also results in enhanced expression of a large number of genes. These include vascular endothelial growth factor (VEGF), elastin, myogenin, cyclin D1, c-myc, c-fos, c-jun, tubulin, and neurofilament. The wide array of genes affected by IGF-I–receptor signaling attests to the pleiotropic responses to this receptor, which range from cell-cycle progression and cellular proliferation to differentiated functions in specialized tissues. INSULIN-LIKE GROWTH FACTOR–BINDING PROTEINS A third, yet important, component of the IGF system is the family of IGF-binding proteins (IGFBPs).36,37 IGFBPs-1 through -6 are encoded by a gene family and are characterized by cysteine-rich N and C termini with less-conserved central regions.38 IGFBPs are produced by a variety of biologic tissues and are found in serum and other biologic fluids.39 Some of the general characteristics are shown in Table 173-3.40

TABLE 173-3. Characteristics of the Human Insulin-Like Growth Factor–Binding Proteins

REGULATION The production of IGFBPs is regulated by several factors. All of the IGFBPs are developmentally regulated from fetal levels through the aging process. IGFBP-1 gene expression and circulating levels are regulated by nutritional status, insulin, and gluco-corticoids. GH, on the other hand, increases the levels of both IGFBP-3 and the acid-labile subunit (ALS), both of which form the major serum complex that binds the majority of circulating IGFs. The effect of GH on IGFBP-3 is probably mediated by IGF-I itself. At the tissue level, the IGFBPs are also regulated by other hormones, including retinoic acid and cytokines, and intracellular messengers such as cyclic adenosine monophosphate (AMP). Several IGFBPs are subject to phosphorylation and glycosylation. Most IGFBPs are present in soluble and cell-associated forms via the Arg-Gly-Asp (RGD) sequences that may interact with integrins.41 Most IGFBPs are subject to proteolytic cleavage by specific proteases, which themselves are subject to regulation.42 Thus, the production and activity of the IGFBPs are controlled at multiple levels, including synthesis, degradation, posttranslational modification, and cell association. FUNCTIONS In human serum, the major pool of IGFs circulate as an ~150-kDa (IGFBP-3/ALS) complex, a smaller pool as an ~50-kDa complex, and ~1% as free IGF. Bound IGFs in the serum have a half-life of up to 15 hours when bound to the IGFBP-3/ALS complex, but only ~20 minutes when they circulate in the free form or in the smaller 50-kDa complex. The 150-kDa complex can serve as a reservoir from which IGFs can be dissociated into the free form or transferred directly to other circulating IGFBPs. The small 50-kDa complex can leave the bloodstream and deliver IGF-I to target tissues. The major bound form of IGFs also prevents the hypoglycemia that would occur if the high concentrations of IGFs were in the free form and able to bind, albeit with a low affinity, to insulin receptors in muscle and liver.43 Locally produced IGFBPs affect the interaction of the IGFs with the target tissues. Under certain circumstances, IGFBPs may enhance IGF action by effectively increasing the local concentration of IGF, by presentation of IGF to the receptor, or possibly by triggering IGF-independent events that contribute to IGF action at a postreceptor level. Generally, the IGFBPs demonstrate a higher affinity for the IGFs compared to the IGF-I receptor, especially when in the soluble form or matrix-bound.44 This affinity falls dramatically when cleaved by proteases, phosphorylated, or partitioned, that is, cell-surface bound.45 These changes may explain the ability of certain IGFBPs either to inhibit or to potentiate the actions of the IGFs at the target. 46 INDEPENDENT ACTION Current evidence strongly suggests that some of the IGFBPs may function in an IGF-independent manner.47 IGFBP-1 and -2 both contain RGD sequences that may enable them to bind to integrins and promote cellular migration and possibly proliferation. IGFBP-3 binds to the cell surface by an undetermined mechanism. Further-more, IGFBP-3 has been shown to inhibit proliferation and even induce apoptosis via an IGF-I receptor–independent mechanism.48 Binding of IGFBP-3 to the cell surface may be the signal for this antiproliferative effect. No specific receptor for IGFBP-3 (or for any IGFBP) has been described; thus, the exact mechanisms

whereby the IGFBPs affect cellular processes remain to be determined.

INSULIN-LIKE GROWTH FACTORS IN HEALTH AND DISEASE NORMAL GROWTH AND DEVELOPMENT The IGFs are essential for normal growth and development. The ligands, receptors, and certain IGFBPs are expressed in the prenatal stage.49,50 When individual genes encoding certain members of the IGF system are inactivated by mutation or by homologous recombinant gene-targeting techniques, the consequences for fetal development are extremely serious.51,52 Loss of the IGF-II/mannose-6-phosphate receptor due to the Tme-deletion mutant results in fetal death, whereas homologous deletion of IGF-I, IGF-II, or the IGF-I–receptor genes results in overall growth retardation of the fetus, generalized retardation in organ development, and, in the case of IGF-I and the IGF-I receptor gene deletion, in perinatal lethality. GROWTH DISORDERS GH-deficient dwarfs have reduced circulating IGF-I levels. Serum IGF-I levels return to normal on replacement with recombinant human GH. Laron-type dwarfs have similar reductions in circulating IGF-I levels despite elevated serum GH levels, due to mutations in the GH receptor and resultant GH resistance. Patients with malnutrition and/or poorly controlled type 1 diabetes have lowered circulating IGF-I levels due partly to lower insulin levels and partly to postreceptor abnormalities in the tissues, leading to GH resistance. To date there has been only one report of a growth-retarded male patient with an IGF-I–gene deletion. Growth retardation is also seen in rare patients with deletion of the IGF-I receptor due to a ring form of chromosome 15. DIABETES The IGF system has been implicated in the development of many of the complications of diabetes. A delayed pubertal growth spurt in poorly controlled patients with type 1 diabetes is associated with reduced circulating levels of IGF-I secondary to insulinopenia that causes GH resistance at the level of the liver.53 Diabetic microangiopathy and macroangiopathy may partially be the result of the mitogenic effects of IGFs on the vascular smooth muscle cells. IGF-I enhances the gene expression of VEGF in these cells. IGF-I appears to have an indirect angiogenic effect by increasing VEGF gene expression. Retinal pigment epithelial cells release VEGF into the culture media in response to IGF-I with a resultant stimulation of capillary endothelial cell proliferation in mixed cultures of RPE and endothelial cells.54 Enlargement of the kidneys at early stages of diabetes is associated with increased hemodynamic changes secondary to IGF-I. Recombinant human IGF-I (rhIGF-I) has been tested in diabetics in an attempt to overcome the severe insulin resistance.55 Subcutaneous injection of rhIGF-I reduces the requirement for insulin administration in both types 1 and 2 diabetic patients and is associated with enhanced insulin-induced disposal of glucose, inhibition of hepatic glucose output, and improvement in the metabolic status of these patients. Systemically administered rhIGF-I suppresses GH (thereby reducing insulin resistance), suppresses glucagon (thereby abrogating the glucagon-enhanced hepatic glucose output), and, in the case of type 2 diabetics, suppresses insulin levels (thereby preventing the “down-regulation” of insulin receptors on peripheral target cells that occurs in response to elevated insulin levels).56 These effects of administered rhIGF-I may explain the improvement in the glucose homeostasis of these patients. It remains to be determined if rhIGF-I therapy will become a viable treatment for diabetic patients. 56a CANCER Many components of the IGF system appear to play important roles in cancer.57,58,58a Numerous tumors express high levels of IGFs, both types of IGF receptors, and many of the IGFBPs in various combinations.59 IGF-II is commonly expressed at high levels in many tumors due to loss of imprinting, or mutations of tumor-suppressor gene products, such as WT1, which normally control its expression.60 In the case of the IGF-II gene (chromosome 11p15), studies have shown that the maternal allele, which is imprinted in normal tissue, may be expressed in Wilms tumors without loss of heterozygosity (LOH).61 This loss of imprinting(LOI) results in the expression of both maternal and paternal IGF-II alleles. Similar LOI can be seen with various lung, breast, and colon cancers. LOI is also associated with microsatellite instability (MSI). Imprinting has been ascribed to methylation of the gene. Hypermethylation appears to favor imprinting and it is postulated that hypomethylation is associated with expression of the gene. Expression of the IGF-II gene may also be due to hypomethylation of the neighboring inhibitory H19 gene. Many nonislet cell tumors, primarily of mesenchymal origin, overexpress an incompletely processed prohormone form of IGF-II (termed “big IGF-II”), which leads to clinically apparent hypoglycemia.62 Big IGF-II is incompletely neutralized by the 150-kDa IGFBP-3 complex and associates preferentially with the smaller 50-kDa IGFBP-2 complex.63,64 This complex rapidly transfers the ligand out of the circulation and delivers it to the target tissues, thereby inducing hypoglycemia.65 The IGF-I receptor has intrinsic tyrosine kinase activity and mediates cell proliferation.66 It is commonly overexpressed in tumors, thereby enhancing tumor growth, presumably by increasing responsiveness to the IGFs.67 Circulating levels of IGF-I are increased in patients with prostatic, colon, breast, and lung cancer.68 Results of prospective studies suggest an increased risk for the development of these and other cancers in the setting of elevated IGF-I levels. Elevated IGF-I levels have also been measured in cancer patients in case-controlled studies. Whether the increased IGF-I levels are merely markers for cancer development or intimately involved in the development of tumor growth remains to be answered. Further study is also needed to determine the origin(s) of these increased IGF-I levels in the circulation. NERVOUS SYSTEM The IGF system is involved in the growth and development of the nervous system. IGFs acting through the IGF-I receptor modulate growth and differentiation of neurons and glial cells. IGFs stimulate neurite outgrowth, synaptogenesis in sympathetic precursors, and commitment of progenitors to the oligodendrocytic lineage. The important role of the IGFs in adult nervous tissue centers primarily on the control of apoptosis. During development, apoptosis is essential to the removal of excess neurons, whereas, in adult tissue, anti-apoptosis is essential to prevent the loss of neurons after certain injuries and neurodegenerative diseases. IGFs signaling via the IGF-I receptor are powerful antiapoptotic agents. The major signaling pathways impacting apoptosis have been elucidated in cell culture systems. Activation of the PI3‹K system in turn activates protein kinase B/Akt. In some specific cell types, the MAP kinase pathways are also involved in control of apoptosis. In vivo models of hypoxic brain injury corroborate the importance of IGF-I in the control of apoptosis. Hypoxic/ischemic brain infarction is associated with a large area of apoptosis, which occurs in the surrounding tissue and may lead to a greater degree of cerebral dysfunction than the ischemic region itself. Local injection of rhIGF-I immediately after the infarction greatly reduces the neuronal loss and preserves function.69,70 IMMUNE SYSTEM Differentiation of T cells, induction of granulocytic differentiation by granulocyte and macrophage colony-stimulating factor (GM-CSF), and differentiation of hematopoietic progenitors by erythropoietin (EPO) are associated with local production of IGF-I, which acts through the IGF-I receptors that are widely expressed on all myeloid and lymphoid cells. In addition to stimulating proliferation and differentiation of these cells, IGF-I exerts chemoattractant effects on T-cell progenitors migrating from hematopoietic tissues to the thymus, where they are further differentiated by cytokines.71 At sites of inflammation, macrophage- derived IGF-I, together with interleukin-1 (IL-1), enhances migration of T lymphocytes to the site. These cytokine-activated T cells undergo clonal expansion in response to IGF-I.72 In nonhuman primates, administration of rhIGF-I alone, or with rhGH, increases T cells in the spleen with a concomitant increase in the CD4/CD8 ratio.73 The potential use of IGF-I as a therapy for immunodeficiency states requires further study. REPRODUCTIVE SYSTEM Delayed puberty is commonly associated with GH deficiency and is assumed to be the result of reduced IGF-I levels, which play an important role in ovarian, uterine, and testicular physiology.74 Gonadotropins and sex steroids enhance local IGF-I production and IGF-I–receptor expression. Conversely, IGF-I modulates follicle-stimulating hormone (FSH) action on the ovary and enhances steroidogenesis in the ovary and testes.75 Patients undergoing hormonal manipulation for in vitro fertilization show an improved response when GH is administered systemically, presumably by enhancing hemocrine and local IGF-I production. Estrogen enhances uterine IGF-I expression, which in turn affects endometrial physiology; moreover, IGF-I may also be important for embryo implantation in the uterus.76 BONE The pubertal growth spurt is mediated by IGF-I, either from the circulation or more likely from local autocrine/paracrine IGF-I production.77 Osteoblasts produce IGF-I, and this expression is enhanced by parathyroid hormone, prostaglandin E2 (PGE2), and estrogen.78,79 In cooperation with other growth factors, IGFs affect proliferation

and differentiation of bone cells. These actions may be extremely important in callus formation after a fracture and may be therapeutically relevant in cases of osteoporosis. IGFBPs are also expressed by osteoblasts, and the level of IGFBP-4 that inhibits IGF-I action on proliferation and differentiation is increased by 1,25(OH)2 vitamin D3. IGFBP-5 enhances IGF-I effects on bone cells; IGF-I enhances formation of multinucleated osteoclastic cells, thereby coupling bone formation and resorption.80 POTENTIAL CLINICAL USES Recombinant human IGF-I has been extremely useful clinically in reversing the negative nitrogen balance in several catabolic states and reducing insulin resistance in types 1 and 2 diabetes.81 It induces growth in Laron-type dwarfism, for which it has been approved.82 Other potential uses include treatment for diabetic neuropathy, amyotrophic lateral sclerosis, bony fractures, osteoporosis, and wound healing. It also may be useful in reactivating the immune response in acquired immunodeficiency syndrome (AIDS). Its usefulness may be enhanced by the coadministration of IGFBP-3. Yet to be confirmed are its potential long-term side effects, including enhanced mitogenesis and tumor growth.83

OTHER GROWTH FACTORS VASCULAR ENDOTHELIAL GROWTH FACTOR Angiogenesis, a process of new blood vessel formation, is a fundamental requirement for organ development and differentiation during embryogenesis. This has been confirmed by homologous recombination null mutants for VEGF and its receptors that are invariably fatal during embryogenesis due to a failure of blood vessel development.84 It is also implicated in numerous pathologic processes, including tumor growth, metastases, and diabetic retinopathy.85 Although there are a number of angiogenic factors, including the fibroblast growth factors (FGFs), tumor necrosis factor a (TNF-a), transforming growth factor-a and -b (TGF-a and-b, and leptin,86 the most powerful known regulator of normal and tumor angiogenesis is VEGF. There are a number of VEGF isoforms, each with a different pattern of secretion. Some isoforms are retained on cell surfaces, and others are sequestered in the extracellular matrix by heparan sulfate proteoglycans. The larger VEGF family of peptides includes placenta-derived growth factor, and VEGF A, B, and C. These three, well-characterized, transmembrane VEGF receptors are expressed almost exclusively by vascular endothelial cells and contain extracellular immunoglobulin repeats and a split tyrosine kinase domain in the cytoplasmic portion of the molecule. Activation of the receptor kinase leads to signaling via phospholipase C-g (PLC-g), PI3‹K, and ras GTPase-activating protein (GAP), Fyn and Yes (Src family of proteins). VEGF is expressed by a variety of tumors, and the degree of vascularization of the malignancy correlates with the level of VEGF mRNA. Antibodies to VEGF can inhibit tumor growth, suggesting an important role in tumor progression and supporting the theory that there may indeed be a role for an antagonist to VEGF in treating tumors and restenosis after angioplasty.87 NERVE GROWTH FACTOR FAMILY Nerve growth factor (NGF) is part of a family of neurotropins that includes brain-derived neurotropic factor (BDNF), neurotropin- 3 (NT-3), NT-4, and NT-5. NGF is a highly conserved, 118–amino-acid protein exhibiting more than 70% homology across all vertebrate species. The various members of the family (i.e., NGF, BDNF, and the NTs) are also conserved in certain regions and also have similar predicted tertiary structures. RECEPTORS Initial studies demonstrated the existence of low-affinity receptors that bound NGF, BDNF, and NT-3, and high-affinity receptors were found that were specific for NGF or BDNF. The low-affinity receptor was shown to be a 75-kDa intrinsic membrane protein (p75). A 140-kDa tyrosine kinase receptor encoded by the protooncogene Trk (or TRK-A) binds NGF with high affinity and mediates NGF effects. TRK-B is a TRK-related gene product that binds BDNF and NT-3, and TRK-C binds NT-3 with high affinity. TRK-A, TRK-B, and TRK-C contain tyrosine kinase activity, but p75 does not. A model for the functional interaction of the NGF-related peptides and functional receptors suggests that p75, TRK-A, and TRK-B individually are low-affinity receptors, and the association of p75 with TRK-A or TRK-B generates a high-affinity, functional receptor. The TRK-A tyrosine kinase receptor is responsible for mediating NGF signaling. The signaling pathways include PLC-g, PI3‹K, ras GAP, Shc, and the MAP kinase/Erk1 pathways.88 FUNCTIONS NGF-related peptides stimulate survival and differentiation of a range of target neurons. TRK-B expression is widespread throughout the central nervous system, serving more general functions. TRK and p75 are colocalized to the medial septal nucleus and the nucleus of the Broca diagonal band, which contains the NGF-responsive magnocellular cholinergic neurons projecting to the hippocampus and cerebral cortex. NGF is the prototypic growth factor that controls cell survival. Neurons that project to an inappropriate target are automatically eliminated, because they fail to be stimulated by neurotropic factors, but neurons projecting to the appropriate target cell are innervated accordingly. The other members of the family have similar functions but act on different subsets of neurons. Selective degeneration of magnocellular cholinergic neurons in the nucleus basalis of Meynert is seen in patients with Alzheimer disease. Clinical trials have begun using NGF directly intraventricularly in patients with Alzheimer dementia or to support transplanted adrenal medullary tissue in Parkinson disease. PLATELET-DERIVED GROWTH FACTORS STRUCTURE PDGFs are a family of proteins consisting of disulfide-bonded dimers of A and B chains.89 The A and B chains are encoded by separate genes that apparently arose by gene duplication and divergence. They have retained ~60% similarity in their amino-acid sequences, and their eight cysteine residues are perfectly conserved. The major form of PDGF in humans is the AB heterodimer, which is expressed primarily by platelets in adults. The BB homodimer (homologous to the viral sis oncogene) is found in other species, and the AA homodimer is expressed by certain tissues and various tumors (i.e., osteosarcomas, melanomas, and glioblastomas). PDGFs are also produced by macrophages, mesangial cells, placental cytotrophoblasts, smooth muscle cells, endothelial cells, fibroblasts, neurons, glial cells, and embryonic cells. PDGF receptors are expressed by vascular smooth muscle cells, fibroblasts, and glial cells, but the receptors are not expressed by most hematopoietic, epithelial, or endothelial cells. Two subtypes have been identified. The a receptor binds all three PDGF isoforms (i.e., AA, AB, and BB), and the b receptor binds only PDGF-BB with high affinity. Both receptor sub-types contain five immunoglobulin-like domains, a single transmembrane sequence, and an intracellular protein tyrosine kinase region that is split by a kinase insert region. Ligand binding leads to receptor dimerization, activation of the kinase, and subsequent association and activation of numerous endogenous substrates. These include phospholipase A2, PLC-g, PI3‹K, and ras GAP. These substrates interact directly with phosphotyrosine residues on the PDGF receptor by means of their SH2 domains. FUNCTIONS PDGF induces cell proliferation in mesenchymal cells, including fibroblasts, osteoblasts, arterial smooth muscle cells, and brain glial cells. PDGF is a competence factor, allowing cells to enter the G0/G1 phase of the cell cycle, but further progression is the function of other growth factors. Wound Healing. PDGF is synthesized and released at the site of injury by platelets, vascular cells, monocyte-macrophages, fibroblasts, and skin epithelial cells. PDGF, by means of a paracrine mechanism, induces proliferation and chemotaxis of connective tissue cells and production of extracellular matrix, thereby enhancing healing. Osteogenesis. PDGF plays an important role in bone formation and metabolism by stimulating DNA synthesis and collagen synthesis by osteoblasts. In addition to normal bone development, PDGF is potentially capable of enhancing new bone formation after fractures. Atherosclerosis. Atherosclerosis is an abnormal proliferation of arterial smooth muscle cells, increased number of macrophages, and excessive deposition of connective tissue. In addition to platelet-derived PDGF, vascular endothelial cells and activated macrophages express PDGF. PDGF-receptor expression is increased in the cells proximal to the macrophages, the intimal smooth muscle cells. PDGF is one of the most important growth factors involved in atherogenesis.

The restenosis that follows balloon angioplasty is also associated with increased PDGF and receptor expression, particularly by neointimal smooth muscle cells. Fibrosis. PDGF and receptors play a role in many fibrotic diseases, including myelofibrosis, scleroderma, and pulmonary fibrosis, by stimulating connective tissue cell proliferation, chemotaxis, and collagen synthesis, which are all pathognomonic features of these diseases. Neoplasia. The acutely transforming simian sarcoma virus genome contains a retroviral homologue of the cellular gene encoding PDGF-B. This finding led to speculation that the cell-cycle competence factor PDGF may be involved in tumor cell growth. Tumors such as gliomas, sarcomas, melanomas, mesotheliomas, carcinomas, and hematopoietic cell-derived tumors over-express PDGF. Many tumors express PDGF receptors that may be activated in the absence of the ligand. Transforming effects of v-sis may occur by this mechanism. Furthermore, chronic myelomonocytic leukemia is associated with a fusion protein produced by chromosomal translocation. This fusion protein consists of the PDGF kinase domain fused to an ets-like gene product. The fusion product serves to autodimerize the receptor, thereby activating it in the absence of ligand.90 Although PDGF may be involved in pathologic states, its value therapeutically in treating wounds and bone fractures, to name a few examples, remains to be explored. EPIDERMAL GROWTH FACTOR FAMILY The EGF family of growth factors includes EGF, TGF-a, amphiregulin, heparin-binding EGF, schwannoma-derived growth factor, and the vaccinia virus growth factor.91 They are each synthesized as a much larger membrane-bound glycosylated precursor before being processed into a smaller mature peptide. EPIDERMAL GROWTH FACTOR The 1217–amino-acid, membrane-bound EGF precursor (Pre-proEGF) and the 53–amino-acid mature peptide are capable of interacting with cell-surface EGF receptors. This interaction causes dimerization of the single-chain EGF receptors, activation of the cytoplasmic tyrosine kinase domain, and subsequent biologic responses. EGF is a potent stimulator of cell multiplication, and it modulates the differentiation and specialized functions of various cells. EGF's effects on development include eyelid opening, teeth eruption, lung maturation, and skin development. In keeping with its initial isolation from salivary glands, EGF has been shown to protect the gastric mucosa by inhibiting gastric acid secretion. In addition to its angiogenic and proliferative ability, EGF displays a chemotactic activity for inducing migration of fibroblasts into the wound area.92 The overexpression of EGF and the EGF receptor occurs in certain carcinomas. The EGF receptor (c-erb1) is homologous to the avian viral oncogene v-erbB, strongly supporting the notion that overexpression may be involved in tumorigenesis. A correlation has been demonstrated between amplification of the EGF-receptor gene and poor prognosis in breast, lung, and bladder cancers. TRANSFORMING GROWTH FACTOR-a TGF-a resembles EGF structurally and functionally. It is synthesized as part of a 160–amino-acid cell-surface precursor, and the mature 50–amino-acid TGF-a is proteolytically cleaved and released from the extracellular domain. It is a potent mitogen, acting through the EGF receptor. It is expressed in preimplantation embryos and the fetus and is essential in normal development. In adults, it is expressed by the anterior pituitary, brain, decidual cells, skin keratinocytes, bronchus, kidney, and genital tract and has been implicated in wound healing and inflammation, angiogenesis, and bone resorption. TGF-a is overexpressed in many cancers, in which it is implicated, along with the EGF receptor, as being tumorigenic.93 TRANSFORMING GROWTH FACTOR-b FAMILY TGF-b1 is a disulfide-linked dimer of two identical chains of 112 amino acids. The chains are synthesized as 390–amino-acid precursor molecules. The cleaved proregion remains associated with the mature TGF-b1 dimer forming a biologically latent complex, which becomes active on disassembly of this complex. This family of growth factors in humans includes TGF-b, the activin-inhibin family, and müllerian inhibitory hormone. BIOSYNTHESIS The TGF-b peptides represent three separate gene products expressed by many normal cells and tissues. Expression is active throughout embryonic development and into adulthood. Gene expression is regulated by multiple factors at the level of transcription, and except in the case of platelets, in which TGF-b is stored in a-granules, the TGF-bs are released from cells through a constitutive pathway. TGF-b1 released from cells is in the “latent” form and the proregion contains mannose-6-phosphate, which binds to IGF-II/mannose-6-phosphate receptors, and an RGD sequence, which may be important for its interaction with integrin receptors. The role of these residues in regulating the release of active TGF-b remains undefined; however, proteases play a definitive role in conversion from the latent to the active form. RECEPTORS There are at least two distinct receptors that mediate the effects of TGF-b. The type I receptor is expressed only by hematopoietic cells that are growth-inhibited by TGF-b and is expressed with the type II receptor by many different cells and tissues in which growth inhibition, extracellular matrix protein synthesis, and differentiation are the major responses. In addition to these two classic receptors, betaglycan (the type III TGF-b receptor), an abundant membrane proteoglycan, also binds TGF-b. A soluble form of betaglycan is found in the extracellular matrix. Betaglycan binding of TGF-b has been invoked as important in the storage of TGF-b in the matrix, presentation of the ligand to the receptor, and even clearance of TGF-b, whereby the membrane form internalizes with the ligand. Type V TGF-b receptor contains an amino-acid sequence that shows homology to the type II TGF-b receptor.94 This receptor also appears to play a role in internalizing the ligand. Type I, II, and V TGF-b receptors are members of the family of transmembrane serinethreonine kinases. Members of this family include receptors for activin and müllerian inhibiting substance. Type I TGF-b receptors cannot bind ligand independently. Type I, II, III, and V receptors undergo autophosphorylation.95 One of the most important effects of TGF-b is its inhibitory effect on cell proliferation. Inhibition of phosphorylation of the retinoblastoma gene product is apparently the mechanism whereby TGF-b inhibits cell-cycle progression. ACTIONS The bioactions of TGF-b are extensive and varied. TGF-b can inhibit or stimulate proliferation, depending on the culture conditions. In the presence of mitogen-rich medium, TGF-b inhibits; in the presence of mitogen-free medium, TGF-b can enhance proliferation by inducing PDGF.96 TGF-b causes enhanced cell-cell adhesion in mesenchymal and epithelial cells and various cell lines. This process is accompanied by increased extracellular matrix production and expression of cell-adhesion receptors. These effects explain, at least in part, the role of TGF-b in wound healing, tissue repair, and angiogenesis. TGF-b is expressed by activated macrocytes and macrophages at sites of wound healing or inflammation and is a chemoattractant for these cells. Bone remodeling is enhanced by locally produced TGF-b. TGF-b has also been invoked as a causative mediator in diseases, including acute mesangial proliferative glomerulosclerosis, fibrotic diseases such as lung fibrosis, liver cirrhosis, arterial restenosis after angioplasty, and myelofibrosis. Because TGF-b forms have antiproliferative effects on T- and B-lymphocytes, they have potent immunosuppressive effects in vivo. The elevated TGF-b expression in lymphocytes may be one explanation for the general immunosuppressive effects of the AIDS virus despite the limited number of lymphocytes that are actually infected. The potential benefits of the antiinflammatory and immunosuppressive effects of TGF-b in systemic disease, such as rheumatoid arthritis, await further experimentation. As a suppressor of cell proliferation, the absence of TGF-b or its receptor may result in oncogenesis. Only retinoblastoma cells have been reported to be devoid of TGF-b receptors, which may enable increased oncogenic potential. INHIBIN, ACTIVIN, AND MÜLLERIAN-INHIBITING HORMONE Inhibins and activins are dimeric polypeptides composed of similar subunits (see Chap. 16, Chap. 113 and Chap. 114). Inhibins inhibit production of FSH in pituitary cells, sex steroid in the gonads, and many placental hormones. The activins stimulate the production of all these hormones. Activins also induce differentiation of

erythroleukemia cells. Müllerian-inhibiting substance (MIH) induces müllerian duct regression, inhibits oocyte maturation, and is capable of inhibiting production of desaturated phosphatidylcholine, a component of surfactant.97 FIBROBLAST GROWTH FACTOR FAMILY Seven members of the FGF family are known.98 Acidic FGF (FGF-a) and basic FGF (FGF-b) have different tissue preferences for expression and function. The family includes a newly described member: keratinocyte growth factor (KGF). The FGFs support the survival of neural cells and stimulate proliferation of many types of cells, including fibroblasts, endothelial cells, smooth muscle cells, hepatocytes, and skeletal myoblasts. Moreover, mesoderm induction is FGF dependent. Two classes of FGF-binding sites have been characterized. The low-affinity, high-capacity receptors are cell-surface proteoglycans containing heparan sulfate side chains. The high-affinity tyrosine kinase receptors are important for transducing the signals. These high-affinity receptors represent a family of gene products. The extracellular domains have three immunoglobulin-like loops, which are involved in ligand binding and are highly conserved. The high- and low-affinity receptors collaborate in FGF binding. The low-affinity, heparan sulfate–containing receptors bind the FGF molecule, allowing it to dimerize so that it can bind to the high-affinity receptors. This results in activation of tyrosine kinase activity and activation of phospholipase Cg 1, one of the major receptor substrates in the signal transduction pathway.

CYTOKINES The cytokine family contains a diverse collection of proteins that, by activation of specific cell-surface receptors, regulate many cellular processes, including the immune and inflammatory systems (see Chap. 227), and differentiation processes such as hematopoiesis and leukopoiesis. This family includes the interleukins (IL), the tumor necrosis factors (TNF-a and -b), the interferons (IFN-a, -b, and -g), the macrophage and granulocyte colony-stimulating factors (M-CSF/CSF-1, G-CSF, and GM-CSF), and several other molecules. In this section, the major subgroups of this family are discussed, including the ILs, TNFs, IFNs, and CSFs. Cytokines are closely allied with growth factors because they are often synthesized by multiple cell types. Exceptions to this are IL-2 through IL-5 and IFN-g, which are produced specifically by lymphoid cells. IL-3, for example, is predominantly produced by activated T cells. Also similar to many growth factors, cytokines modulate the activity of several types of cells rather than just one specific target cell. INTERLEUKINS INTERLEUKIN-1 IL-1a and IL-1b are encoded by similar but distinct genes. Both proteins are synthesized as part of larger precursor proteins. IL-1 is produced by many cell types, including monocytes and macrophages, neutrophils, astrocytes and microglia, endothelial cells, fibroblasts, T and B lymphocytes, and platelets. IL-1 is one of the most pleiotropic interleukins in that it induces the widest array of responses. IL-1b is involved in the immune and inflammatory responses, and in hematopoiesis. It has been shown to be cytotoxic for isolated pancreatic cells and may play a causative role in the development of autoimmune or type 1 diabetes. IL-1b induces the apoptotic response of many cells by activating the caspase cascade of events. Almost all of the effects of IL-1 are seen with TNF, and synergy with TNF can be demonstrated in many cases. INTERLEUKIN-2 IL-2 is an important component in the immune response.99 It is produced by antigen-induced T cells, and its subsequent interaction with IL-2 receptors, which are also antigen induced, leads to clonal expansion of effector T-cell populations. In addition to proliferative effects, IL-2 stimulates differentiation, inducing the production of IFN-g and IL-4 by T cells. IL-2 also affects B cells, natural killer cells (NK), lymphokine-activated killer (LAK) cells, monocytes, macrophages, and oligodendrocytes. The T-cell–activating capacity has promoted its consideration as a form of immune replacement therapy in immunodeficient (AIDS) or immunocompromised patients. Its effect on LAK cell function has led to its use as a possible anti-tumor therapy (IL-2/LAK) in patients with renal carcinoma and malignant melanoma.100 The major dose-limiting toxicity of IL-2 is due to extravasation of fluid and protein into the interstitium.101 This toxicity seems to be due to activation of the endothelium subsequent to the IL-2–induced release of IL-1, TNF-a, and interferon; it may be reduced by depletion of NK cells. Pharmacologically, this side effect may be reduced by concomitant steroid therapy and lymphocyte depletion, but these modalities reduce the antitumor efficacy of the primary treatment. Endocrine toxicity of IL-2 therapy includes hypothyroidism, associated with the development of antimicrosomal and antithyroglobulin antibodies. In rare instances, adrenal insufficiency occurs after IL-2 therapy. INTERLEUKIN-3 IL-3 is produced primarily by activated T lymphocytes. Its major function is in the hematopoietic system, where it stimulates the growth of bone marrow stem cells and the differentiation of myeloid stem cells into many different lineages, including platelets, mast cells, eosinophils, basophils, and erythroid cells. Thus, it links the lymphoid system with the hematopoietic system. It has been proposed that IL-3 may be important in the recovery after myelotoxicity especially of the platelets. However, IL-3 may be involved in allergy production by enhancing mast cell production. Thus, its potential use in enhancing hematopoietic cell differentiation will need further evaluation given its effect on the allergic system.102 INTERLEUKIN-4 IL-4 is produced by T cells, primarily TH2 as well as NK cells, and leads to a preferential stimulation of humoral immunity by its effect on B-cell development.103 IL-4 enhances the antigen-presenting capacity of B cells toward T cells. It induces activated B cells to produce IL-6 and TNF cytokines, which play an important role in the activation and clonal expansion of activated T cells. Thus, IL-4 favors T-B cell interactions. IL-4 inhibits the production of IFN-g by activated T cells and, in doing so, favors the generation of T cells of the TH2 type. Blood mononuclear cells from atopic patients display an increased capacity to produce IL-4. The deregulated immunoglobulin E (IgE) synthesis by cells from atopic patients can be inhibited by anti–IL-4 antibodies. In patients with severe atopic dermatitis, circulating and skin-derived T cells produce higher levels of IL-4 and less IFN-g. IL-4 is a potent antitumor agent in mice. When expressed by plasmacytoma cell lines, IL-4 inhibits growth; this is reversed by anti–IL-4 antibodies. IL-4 has been tested in humans with melanoma and renal cell cancer.104 Nasal congestion and gastritis due to histamine release are specific side effects. When used with IL-2, some remission has been shown (e.g., spleen and lymph node reductions in patients with low-grade lymphoma; even complete remission in cases of Hodgkin lymphoma). Complete remission of pulmonary metastases in cases of renal carcinoma has been seen. Similarly to IL-3, while it may have an antitumor effect, it also is involved in allergy development.105 INTERLEUKIN-5 IL-5 is produced by leukocytes. Within the lymphocyte lineage, T helper cells produce IL-5, whereas among the myeloid lineages, mast cells and eosinophils are the major producers of IL-5. IL-5—as well as being an antiapoptotic agent for B cells that, thus, enlarges the B-cell population—affects B-cell differentiation and causes B cells to produce antibodies.106 IL-5 also enhances eosinophil development and activation, as well as basophil histamine release. Thus, IL-5 may be involved in the pathogenesis of inflammation in allergies (e.g., the chronic asthmatic response). It has been implicated in the eosinophilic response to helminth infections in which IL-5–mediated eosinophilia may evoke an immunologic protection against helminthes. IL-5 injection in murine models of sarcoma and lymphoma led to an eosinophilic infiltration of the malignancy and tumor rejection. Eosinophilic infiltration of allografts is associated with increased expression of IL-5 by the allograft-infiltrating cells and may result in allograft rejection.107 INTERLEUKIN-6 IL-6 is a multifunctional cytokine with a broad range of biologic actions. IL-6 inclusion with IL-1 and glucocorticoids is primarily responsible for the liver induction of the acute-response proteins after severe injury and inflammation. IL-6 also affects the hypothalamic– pituitary–adrenal axis by increasing the production and release of corticotropin-releasing hormone (CRH), which causes increased adrenocorticotropic hormone (ACTH) and glucocorticoid release in response to stress (see Chap.

229). The major source of IL-6 in the bone marrow is macrophages, where it synergistically, with other cytokines, stimulates proliferation of progenitor hematopoietic cells. Activated B cells express IL-6 receptors that respond with enhanced antibody production. IL-6 activates NK cells, but simultaneously enhances plasmacytoma and myeloma growth. Circulating levels of IL-6 are elevated in autoimmune diseases (i.e., systemic lupus erythematosus [SLE], rheumatoid arthritis, and AIDS), implicating IL-6 in these diseases. IL-6 induces the differentiation of numerous nervous system cell lineages and stimulates the secretion of prolactin (PRL), GH, and luteinizing hormone (LH) from the pituitary gland. IL-6, which normally is produced by osteoblasts, is overproduced in ovariectomized mice, leading to enhanced osteoclast development, suggesting a role for IL-6 in the production of postmenopausal osteoporosis, as well as its putative roles in cancer and inflammatory/immune diseases.108 INTERLEUKIN-7 In addition to its traditional role in promotion of growth of B-cell progenitors, IL-7 affects thymopoiesis, myelopoiesis, and neuronal cell survival. In addition, it has profound immuno-modulatory effects ranging from augmentation of immunotherapeutically effective T-cell responses to tumor antigens—such as its ability to enhance the generation of cytolytic T lymphocytes (CTL) and to enhance immune responses to virally mediated pathologic conditions (e.g., influenza A). Moreover, IL-7 improves immune responses to parasitic infections. For example, IL-7 treatment of murine macrophages, which are infected with Leishmania major, reduces the number of infected cells by 50% and reduces the parasite burden per cell.109 INTERLEUKIN-8 More than 15 chemokines, chemoattractants for inflammatory cells, have been discovered thus far. Structurally, they are divided into two major subgroups, C-X-C and C-C chemokines, depending on the presence of one amino acid between the first two cysteine residues. Intradermal injection of IL-8 induces massive local neutrophil infiltration; both lymphocytes and eosinophils infiltrate in response to IL-8. Since many diseases are associated with neutrophil infiltration, it is not surprising that IL-8 may be responsible. Thus, it may play a role in neutrophil infiltration into the synovial fluids of rheumatoid arthritis, osteoarthritis, and gout.110 In acute respiratory distress syndrome, increased concentrations of IL-8 are found in bronchial lavage specimens. Urinary IL-8 levels are increased in cases of glomerulonephritis, IgA nephropathy, lupus nephritis, and so forth. INTERLEUKIN-9 IL-9 functions as a growth factor for helper T cells and as an erythroid colony stimulating factor (CSF). Its lymphocytic activity appears to be specific since it is active with CD4T cells but is inactive with cytotoxic CD8T cells. INTERLEUKIN-10 IL-10 was initially identified as a “cytokine synthesis inhibitory factor,” but, in fact, it demonstrates both inhibitory and stimulatory effects, depending on the local context in which it is produced. IL-10 is produced by TH2 lymphocytes and by B cells, macrophages, and keratinocytes. IL-10, when secreted by TH2 cells, inhibits TH1-cell cytokine synthesis. Parasitic infections (e.g., Schistosoma mansoni infection) result in a strong TH2 response with concomitant suppression of the TH1 response. The reduction in IFN-g allows the infection to occur, and this effect can be blocked with the administration of anti–IL-10 monoclonal antibodies. In addition, nitric oxide production by macrophages and subsequent killing of parasites is inhibited by IL-10. Similar results were noted with mycobacterial infections. On the other hand, in SLE, glomerulonephritis can be prevented by anti–IL-10 monoclonal antibody administration, probably via elevated TNF-a levels. IL-10 has also been shown to function in B-cell growth and differentiation into immunoglobulin-secreting cells.111 In the skin, ultraviolet (UV) radiation causes increased IL-10, which may mediate the immune tolerance after UV radiation and participate in the development of tumors.112 INTERLEUKIN-18 IL-18 induces IFN-g; it is structurally related to IL-1 and demonstrates bioactions shared with IL-1 and IL-12. The IL-18 receptor signals via NFkB and the ras/raf/MAP kinase pathways. It shows some anticancer effects when administered with IL-12. Aberrant regulation of IL-18 may account for the TH1-dependent autoimmunity seen in the NOD mice that develop diabetes; furthermore, one genomic locus associated with type 1 diabetes in NOD mice (namely, Idd2) maps in close proximity to the IL-18 gene.113,114 and 115 TUMOR NECROSIS FACTOR AND LYMPHOTOXIN TNF-a (cachectin) and TNF-b (lymphotoxin) are potent cytokines encoded by genes that are near each other within the major histocompatibility complex on human chromosome 17.116 TNF-a is produced by monocytes, neutrophils, fibroblasts, NK cells, vascular endothelial cells, mast cells, glial cells, astrocytes, smooth muscle cells, and certain cancers. TNF has extremely powerful cytotoxic effects. The effect on cellular killing may be either via apoptosis or via necrosis, depending on the cell type. Apoptosis is mediated via a mechanism similar to Fas ligand and FAS activation.117 Crucial for the immunomodulatory effects of TNF is its ability to alter cell-surface proteins (e.g., human leukocyte antigen [HLA] class I and II antigen, intercellular adhesion molecule [ICAM], E-selectin, and vascular adhesion molecule [VCAM]), to alter production of cytokines (e.g., IL-6 and IL-1), and to alter production of receptors (e.g., EGF and IL-2). TNF is a costimulator for immune cells (e.g., T and B cells), macrophages, and monocytes; its effect on osteoblasts and osteoclasts leads to reduced bone formation and increased bone resorption, thus promoting osteoporosis. TNF was originally described as a cachectic factor after infections. TNF may down-regulate lipoprotein lipase at the transcriptional level and may up-regulate lipolytic and glycogenolytic pathways in muscle.118 Lipopolysaccharide (LPS) stimulates the production of TNF, which synergistically along with the other cytokines is responsible for the septic shock syndrome. Many viral infections induce TNF production by macrophages and monocytes; in turn, TNF is an antiviral agent. However, some evidence indicates that TNF, like some other cytokines, may stimulate the proliferation of viruses (i.e., human immunodeficiency virus [HIV]). In multiple sclerosis, the intracranial injection of anti-TNF antibodies has shown some promise in controlling the disease; in rheumatoid arthritis, systemic anti-TNF antibody injections have shown efficacy.119 TNF has been tested as an anticancer therapy; alone it has had no beneficial effect; however, in some tumors (i.e., sarcomas and melanomas), when used with cytostatics or IFN-g, some benefit has been achieved. Generally, its initial promise as an anticancer agent has not been realized; its greatest limitation lies in its systemic toxicity effects.120 Unlike TNF, the cellular production of lymphotoxin is limited to the B-lymphocyte lineage. IL-1, IL-2, IL-4, and IL-6, as well as viruses, can induce the production and release of lymphotoxin. While lymphotoxin exerts its physiologic actions through the same receptor as TNF, its effects differ. Although it is unknown how these cytokines exert similar but distinct functions, it is possible that this occurs because of different binding affinities and domains for the same receptor. Missense mutations of the 55-kDa TNF receptor (TNFR1) have been demonstrated in six of seven families with autosomal dominant periodic fever. These mutations, found in the trans-membrane domain, result in reduced cleavage and impaired down-regulation of membrane TNFR1. Thus, the periodic syndrome(s) may be due to constitutively activated receptors.121 Lymphotoxin is antiproliferative for certain tumors, acts as a growth factor for normal B cells and B-cell lymphomas, and induces proliferation and differentiation of hematopoietic cells. Thus, it causes differentiation of human myeloid leukemic cells without any cytotoxic effect. Like TNF it interacts with neutrophils and activates them for antibody-dependent cell-mediated cytotoxicity. It modulates endothelial cell function, including increased permeability, and activates fibroblasts to produce GM-CSF, M-CSF, and IL-6. It is cytotoxic to keratinocytes and causes bone resorption. It can kill virally infected cells and, like TNF, induces lipolysis in adipocytes. In addition, it also acts as an antitumor agent. Lymphotoxin may also play a role in autoimmune diseases (e.g., the insulinitis seen in type 1 diabetes). INTERFERONS IFNs are a family of proteins that are related by their ability to protect cells from viral infections.122 Based on their physical properties and source, the IFNs were subsequently divided into type I or type II. Type I IFNs include IFN-a (originally known as leukocyte IFN) and IFN-b (originally known as fibroblast IFN). IFN-g, known as type II IFN or “immune” IFN, is produced only by T lymphocytes and NK cells. IFN-a and - are very powerful antiviral agents, whereas IFN-g is less potent as an antiviral agent and more potent as an immunomodulator.123,124 Differential effects are seen with IFN-a and IFN-b. For example, certain tumor-derived cell lines are inhibited by one and not the other. On the other hand, their antiviral and immune effects are mediated similarly. In general, IFN-b is produced by fibroblasts and epithelial cells, whereas IFN-a is produced by blood leukocytes. IFNs have been biosynthesized by recombinant technology; however, when administered they have a number of side effects. Acute effects are flulike, whereas chronic administration results in neurologic (depression)

and cardiac toxicity, autoimmune thyroiditis, and hematologic suppression. The development of antibodies after repeated use may limit its usefulness. Clinical trials in multiple sclerosis (MS) patients have not been particularly successful. Condylomas, cervical neoplasia, and herpes simplex skin lesions have proved more responsive. Acute and chronic hepatitis C and B have been treated with IFNs, and the response of some tumors seems promising. Transgenic mice expressing IFN-a within the pancreatic islets develop diabetes, and IFN-a expression increases in the islets of mice with streptozotocin-induced diabetes and in the pancreases of patients with type 1 diabetes. On the other hand, IFN-a administration to NOD mice, which usually develop diabetes, prevents the development of the disease. This paradoxic response may be due to a dose- and time-response effect. IFN-a can augment or suppress cellular and humoral immunity; its effect depends on the time course of the disease. Similarly, IL-1 and IL-12 can either prevent or accelerate diabetes in NOD mice. The inhibition of diabetes is associated with the absence of insulinitis. IFN-g is produced by all CD8 and subsets of CD4 T lymphocytes. The primary physiologic stimulus is an antigen in the context of either major histocompatibility complex (MHC) class II for CD4 cells, or MHC class I for CD8 cells. NK-stimulatory factor or IL-12, a product of B cells, can induce IFN-g production and together with TNF-a can induce production of IFN-g in NK cells. This latter effect is important in the early response to microbial infections, as the IFN-g can rapidly activate macrophage populations until T-cell–dependent sterilizing immunity can be generated. IFN-g regulates MHC class I and class II protein expression in a variety of immunologically important cell types (especially macrophage/monocytes), thereby promoting antigen presentation during the inductive phase of immune responses. IFN-g appears to be necessary for TH1 cell development, leading to cell-mediated immune responses. When IL-10 inhibits IFN-g production, TH2 cells develop to initiate a humoral response. IFN-g is the major physiologic macrophage-activating factor and as such is the major cytokine responsible for inducing nonspecific cell-mediated mechanisms of host defense. It plays an important role in promoting inflammatory reactions and increases TNF-a and TNF-receptor expression, thereby enhancing the LPS-mediated tissue damage. The action of IL-12 as an antitumor agent is mediated by IFN-g. One possible mechanism for this is that IFN-g up-regulates the presentation of tumor antigens to CD4 and CD8 cells, resulting in tumor-cell recognition and regression. CYTOKINE RECEPTORS The cytokine/hematopoietin receptor superfamily includes receptors for EPO, granulocyte-CSF, GM-CSF, ciliary neurotrophic factor (CNTF), IL-2, IL-3, IL-4, IL-5, IL-6, IL-7, IL-9, IL-10, IL-11, IL-12, thrombopoietin, interferons, leptin, GH, and PRL. 86,125,126 The signaling cascade is activated after binding of the ligand to the extracellular domain of one of the receptor subunits, followed by heterodimerization, recruitment, and activation of a member of the Janus family of tyrosine kinases (JAKs).127 JAKs phosphorylate a family of transcription factors (Stats) that translocate to the nucleus and induce gene expression.128 In addition, many other cascades of events are activated by the various cytokine receptors. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56.

LeRoith D. Insulin-like growth factors. N Engl J Med 1997; 336:633. Daughaday WH, Rotwein P. Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and gene structures, serum, and tissue concentrations. Endocr Rev 1989; 10:68 Mathews LS, Norstedt G, Palmiter RD. Regulation of insulin-like growth factor I gene expression by growth hormone. Proc Natl Acad Sci U S A 1986; 83:9343. Pao CI, Farmer PK, Begovic S, et al. Expression of hepatic insulin-like growth factor-I and insulin-like growth factor-binding protein-1 genes is transcriptionally regulated in streptozotocin-diabetic rats. Mol Endocrinol 1992; 6:969. Baker J, Liu JP, Robertson EJ, Efstratiadis A. Role of insulin-like growth factors in embryonic and postnatal growth. Cell 1993; 75:73. Powell-Braxton L, Hollingshead P, Warburton C, et al. IGF-I is required for normal embryonic growth in mice. Genes Dev 1993; 7:2609. Treadway JL, Morrison BD, Goldfine ID, Pessin JE. Assembly of insulin/insulin-like growth factor-1 hybrid receptors in vitro. J Biol Chem 1989; 264:21450. Ullrich A, Gray A, Tam AW, et al. Insulin-like growth factor I receptor primary structure: comparison with insulin receptor suggests structural determinants that define functional specificity. Embo J 1986; 5:2503. Morgan DO, Edman JC, Standring DN, et al. Insulin-like growth factor II receptor as a multifunctional binding protein [published erratum appears in Nature 1988; 20(7):442]. Nature 1987; 329:301. Lowe WL Jr, Adamo M, Werner H, et al. Regulation by fasting of rat insulin- like growth factor I and its receptor. Effects on gene expression and binding. J Clin Invest 1989; 84:619. Abbott AM, Bueno R, Pedrini MT, et al. Insulin-like growth factor I receptor gene structure. J Biol Chem 1992; 267:10759. Werner H, Re GG, Drummond AI, et al. Increased expression of the insulin-like growth factor I receptor gene, IGF1R, in Wilms tumor is correlated with modulation of IGF1R promoter activity by the WT1 Wilms tumor gene product. Proc Natl Acad Sci U S A 1993; 90:5828. Werner H, Bach MA, Stannard B, et al. Structural and functional analysis of the insulin-like growth factor I receptor gene promoter. Mol Endocrinol 1992; 6:1545. Werner H, Woloschak M, Adamo M, et al. Developmental regulation of the rat insulin-like growth factor I receptor gene. Proc Natl Acad Sci U S A 1989; 86:7451. Hernandez-Sanchez C, Werner H, Roberts CT, et al. Differential regulation of insulin-like growth factor-I (IGF-I) receptor gene expression by IGF-I and basic fibroblastic growth factor. J Biol Chem 1997; 272:4663. Rubini M, Werner H, Gandini E, et al. Platelet-derived growth factor increases the activity of the promoter of the insulin-like growth factor-1 (IGF-1) receptor gene. Exp Cell Res 1994; 211:374. Cheatham B, Kahn CR. Insulin action and the insulin signaling network. Endocr Rev 1995; 16:117. Kato H, Faria TN, Stannard B, et al. Essential role of tyrosine residues 1131:1135, and 1136 of the insulin-like growth factor-I (IGF-I) receptor in IGF-I action. Mol Endocrinol 1994; 8:40. Garrett TP, McKern NM, Lou M, et al. Crystal structure of the first three domains of the type-1 insulin-like growth factor receptor. Nature 1998; 394:395. LeRoith D, Werner H, Beitner-Johnson D, Roberts CT Jr. Molecular and cellular aspects of the insulin-like growth factor I receptor. Endocr Rev 1995; 16:143. Bruning JC, Winnay J, Cheatham B, Kahn CR. Differential signaling by insulin receptor substrate 1 (IRS-1) and IRS-2 in IRS-1-deficient cells. Mol Cell Biol 1997; 17:1513. Lavan BE, Fantin VR, Chang ET, et al. A novel 160-kDa phosphotyrosine protein in insulin-treated embryonic kidney cells is a new member of the insulin receptor substrate family. J Biol Chem 1997; 272:21403. Lavan BE, Lane WS, Lienhard GE. The 60-kDa phosphotyrosine protein in insulin-treated adipocytes is a new member of the insulin receptor substrate family. J Biol Chem 1997; 272:11439. Isakoff SJ, Yu YP, Su YC, et al. Interaction between the phosphotyrosine binding domain of Shc and the insulin receptor is required for Shc phosphorylation by insulin in vivo. J Biol Chem 1996; 271:3959. Patti ME, Sun XJ, Bruening JC, et al. 4PS/insulin receptor substrate (IRS)-2 is the alternative substrate of the insulin receptor in IRS-1-deficient mice. J Biol Chem 1995; 270:24670. Kuhne MR, Pawson T, Lienhard GE, Feng GS. The insulin receptor substrate 1 associates with the SH2-containing phosphotyrosine phosphatase Syp. J Biol Chem 1993; 268:11479. Parrizas M, Saltiel AR, LeRoith D. Insulin-like growth factor 1 inhibits apoptosis using the phosphatidylinositol 3‹-kinase and mitogen-activated protein kinase pathways. J Biol Chem 1997; 272:154. Pronk GJ, McGlade J, Pelicci G, et al. Insulin-induced phosphorylation of the 46- and 52-kDa Shc proteins. J Biol Chem 1993; 268:5748. Skolnik EY, Lee CH, Batzer A, et al. The SH2/SH3 domain-containing protein GRB2 interacts with tyrosine-phosphorylated IRS1 and Shc: implications for insulin control of ras signalling. Embo J 1993; 12:1929. Skolnik EY, Batzer A, Li N, et al. The function of GRB2 in linking the insulin receptor to Ras signaling pathways. Science 1993; 260:1953. Beitner-Johnson D, Blakesley VA, ShenOrr Z, et al. The protooncogene product c-Crk associates with insulin receptor substrate-1 and 4PS. Modulation by insulin growth factor-I (IGF) and enhanced IGF-I signaling. J Biol Chem 1996; 271:9287. Beitner-Johnson D, LeRoith D. Insulin-like growth factor-I stimulates tyrosine phosphorylation of endogenous c-Crk. J Biol Chem 1995; 270:5187. ten Hoeve J, Morris C, Heisterkamp N, Groffen J. Isolation and chromosomal localization of CRKL, a human crklike gene. Oncogene 1993; 8:2469. Koval AP, Karas M, Zick Y, LeRoith D. Interplay of the protooncogene proteins CrkL and CrkII in insulin-like growth factor-I receptor-mediated signal transduction. J Biol Chem 1998; 273:14780. Koval AP, Blakesley VA, Roberts CT Jr, et al. Interaction in vitro of the product of the c-Crk-II protooncogene with the insulin-like growth factor I receptor. Biochem J 1998; 330:923. Lamson G, Giudice LC, Rosenfeld RG. Insulin-like growth factor binding proteins: structural and molecular relationships. Growth Factors 1991; 5:19. Jones JI, Clemmons DR. Insulin-like growth factors and their binding proteins: biological actions. Endocr Rev 1995; 16:3. Rajaram S, Baylink DJ, Mohan S. Insulin-like growth factor-binding proteins in serum and other biological fluids: regulation and functions. Endocr Rev 1997; 18:801. Jones JI, Gockerman A, Busby WH Jr, et al. Extracellular matrix contains insulin-like growth factor binding protein-5: potentiation of the effects of IGF-I. J Cell Biol 1993; 121:679. Rechler MM. Insulin-like growth factor binding proteins. Vitam Horm 1993; 47:1. Jones JI, Gockerman A, Busby WH Jr, et al. Insulin-like growth factor binding protein 1 stimulates cell migration and binds to the a 5 b 1 integrin by means of its Arg-Gly-Asp sequence. Proc Natl Acad Sci U S A 1993; 90:10553. Maile LA, Xu S, Cwyfan-Hughes S, et al. Active and inhibitory components of the insulin-like growth factor binding protein-3 protease system in adult serum, interstitial, and synovial fluid. Endocrinology 1998; 139:4772. Zapf J. Physiological role of the insulin-like growth factor binding proteins. Eur J Endocrinol 1995; 132:645. Collett-Solberg PF, Cohen P. The role of the insulin-like growth factor binding proteins and the IGFBP proteases in modulating IGF action. Endocrinol Metab Clin North Am 1996; 25:591. Giudice LC. IGF binding protein-3 protease regulation: how sweet it is! J Clin Endocrinol Metab 1995; 80:2279. De Mellow JS, Baxter RC. Growth hormone-dependent insulin-like growth factor (IGF) binding protein both inhibits and potentiates IGF-I-stimulated DNA synthesis in human skin fibroblasts. Biochem Biophys Res Commun 1988; 156:199. Oh Y, Muller HL, Lamson G, Rosenfeld RG. Insulin-like growth factor (IGF)-independent action of IGF-binding protein-3 in Hs578T human breast cancer cells. Cell surface binding and growth inhibition. J Biol Chem 1993; 268:14964. Cohen P, Lamson G, Okajima T, Rosenfeld RG. Transfection of the human insulin-like growth factor binding protein-3 gene into Balb/c fibroblasts inhibits cellular growth. Mol Endocrinol 1993; 7:380. Bondy CA, Werner H, Roberts CT Jr, LeRoith D. Cellular pattern of insulin-like growth factor-I (IGF-I) and type I IGF receptor gene expression in early organogenesis: comparison with IGF-II gene expression. Mol Endocrinol 1990; 4:1386. Bondy C, Werner H, Roberts CT Jr, LeRoith D. Cellular pattern of type-I insulin-like growth factor receptor gene expression during maturation of the rat brain: comparison with insulin-like growth factors I and II. Neuro-science 1992; 46:909. Liu JL, Grinberg A, Westphal H, et al. Insulin-like growth factor-I affects perinatal lethality and postnatal development in a gene dosage-dependent manner: manipulation using the Cre/loxP system in transgenic mice. Mol Endocrinol 1998; 12:1452. Liu JP, Baker J, Perkins AS, et al. Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 1993; 75:59. LeRoith D, Clemmons D, Nissley P, Rechler MM. NIH conference. Insulin-like growth factors in health and disease. Ann Intern Med 1992; 116:854. Punglia RS, Lu M, Hsu J, et al. Regulation of vascular endothelial growth factor expression by insulin-like growth factor I. Diabetes 1997; 46:1619. Schoenle EJ, Zenobi PD, Torresani T, et al. Recombinant human insulin-like growth factor I (rhIGF I) reduces hyperglycaemia in patients with extreme insulin resistance. Diabetologia 1991; 34:675. Moses AC, Young SC, Morrow LA, et al. Recombinant human insulin-like growth factor I increases insulin sensitivity and improves glycemic control in type II diabetes. Diabetes 1996; 45:91.

56a. Clemmons DR, Moses AC, McKay MJ, et al. The combination of insulin-like growth factor I and insulin-like growth factor-binding protein-3 reduces insulin requirements in insulin-dependent type 1 diabetes: evidence for in vivo biological activity. J Clin Endocrinol Metab 2000; 85(4):1518. 57. Aaronson SA. Growth factors and cancer. Science 1991; 254:1146. 58. Macaulay VM. Insulin-like growth factors and cancer. Br J Cancer 1992; 65:311. 58a. Khandwala HM, McCutcheon IE, Flyvbjerg A, Friend KE. The effects of insulin-like growth factors on tumorigenesis and neoplastic growth. Endocr Rev 2000; 21(3):215. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128.

LeRoith D, Baserga R, Helman L, Roberts CT Jr. Insulin-like growth factors and cancer [see comments]. Ann Intern Med 1995; 122:54. Drummond IA, Madden SL, Rohwer-Nutter P, et al. Repression of the insulin- like growth factor II gene by the Wilms tumor suppressor WT1. Science 1992; 257:674. DeChiara TM, Robertson EJ, Efstratiadis A. Parental imprinting of the mouse insulin-like growth factor II gene. Cell 1991; 64:849. Lowe WL Jr, Roberts CT, LeRoith D, et al. Insulin-like growth factor-II in nonislet cell tumors associated with hypoglycemia: increased levels of messenger ribonucleic acid. J Clin Endocrinol Metab 1989; 69:1153. Zapf J. Role of insulin-like growth factor (IGF) II and IGF binding proteins in extrapancreatic tumour hypoglycaemia. J Intern Med 1993; 234:543. Baxter RC, Daughaday WH. Impaired formation of the ternary insulin-like growth factor-binding protein complex in patients with hypoglycemia due to nonislet cell tumors. J Clin Endocrinol Metab 1991; 73:696. Eastman RC, Carson RE, Orloff DG, et al. Glucose utilization in a patient with hepatoma and hypoglycemia. Assessment by a positron emission tomography. J Clin Invest 1992; 89:1958. Werner H, LeRoith D. The insulin-like growth factor-I receptor signaling pathways are important for tumorigenesis and inhibition of apoptosis. Crit Rev Oncog 1997; 8:71. Werner H, Karnieli E, Rauscher FJ, LeRoith D. Wild-type and mutant p53 differentially regulate transcription of the insulin-like growth factor I receptor gene. Proc Natl Acad Sci U S A 1996; 93:8318. Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma insulin-like growth factor-I and prostate cancer risk: a prospective study. Science 1998; 279:563. Gluckman P, Klempt N, Guan J, et al. A role for IGF-1 in the rescue of CNS neurons following hypoxic-ischemic injury. Biochem Biophys Res Commun 1992; 182:593. Guan J, Williams C, Gunning M, et al. The effects of IGF-1 treatment after hypoxic-ischemic brain injury in adult rats. J Cereb Blood Flow Metab 1993; 13:609. Landreth KS, Narayanan R, Dorshkind K. Insulin-like growth factor-I regulates pro-B cell differentiation. Blood 1992; 80:1207. Fu YK, Arkins S, Wang BS, Kelley KW. A novel role of growth hormone and insulin-like growth factor-I. Priming neutrophils for superoxide anion secretion. J Immunol 1991; 146:1602. LeRoith D, Yanowski J, Kaldjian EP, et al. The effects of growth hormone and insulin-like growth factor I on the immune system of aged female monkeys. Endocrinology 1996; 137:1071. Levy MJ, Hernandez ER, Adashi EY, et al. Expression of the insulin-like growth factor (IGF)-I and -II and the IGF-I and -II receptor genes during postnatal development of the rat ovary. Endocrinology 1992; 131:1202. Adashi EY, Resnick CE, Svoboda ME, Van Wyk JJ. Somatomedin-C enhances induction of luteinizing hormone receptors by follicle-stimulating hormone in cultured rat granulosa cells. Endocrinology 1985; 116:2369. Kapur S, Tamada H, Dey SK, Andrews GK. Expression of insulin-like growth factor-I (IGF-I) and its receptor in the peri-implantation mouse uterus, and cell-specific regulation of IGF-I gene expression by estradiol and progesterone. Biol Reprod 1992; 46:208. Okazaki R, Riggs BL, Conover CA. Glucocorticoid regulation of insulin-like growth factor-binding protein expression in normal human osteoblast-like cells. Endocrinology 1994; 134:126. McCarthy TL, Ji C, Shu H, et al. 17b-estradiol potently suppresses cAMP-induced insulin-like growth factor-I gene activation in primary rat osteoblast cultures. J Biol Chem 1997; 272:18132. Ernst M, Rodan GA. Estradiol regulation of insulin-like growth factor-I expression in osteoblastic cells: evidence for transcriptional control. Mol Endocrinol 1991; 5:1081. Birnbaum RS, Wiren KM. Changes in insulin-like growth factor-binding protein expression and secretion during the proliferation, differentiation, and mineralization of primary cultures of rat osteoblasts. Endocrinology 1994; 135:223. Dunger DB, Cheetham TD, Crowne EC. Insulin-like growth factors (IGFs) and IGF-I treatment in the adolescent with insulin-dependent diabetes mellitus. Metabolism 1995; 44:119. Backeljauw PF, Underwood LE. Prolonged treatment with recombinant insulin-like growth factor-I in children with growth hormone insensitivity syndrome—a clinical research center study. GHIS Collaborative Group. J Clin Endocrinol Metab 1996; 81:3312. Butler AA, Blakesley VA, Tsokos M, et al. Stimulation of tumor growth by recombinant human insulin-like growth factor-I (IGF-I) is dependent on the dose and the level of IGF-I receptor expression. Cancer Res 1998; 58:3021. Dumont DJ, Jussila L, Taipale J, et al. Cardiovascular failure in mouse embryos deficient in VEGF receptor-3. Science 1998; 282:946. Ferrara N, Davis-Smyth T. The biology of vascular endothelial growth factor. Endocr Rev 1997; 18:4. Sierra-Honigmann MR, Nath AK, Murakami C, et al. Biological action of leptin as an angiogenic factor. Science 1998; 281:1683 Klagsbrun M, D'Amore PA. Vascular endothelial growth factor and its receptors. Cytokine and Growth Factor Reviews 1996; 7:259. Kaplan DR, Stephens RM. Neurotrophin signal transduction by the Trk receptor. J Neurobiol 1994; 25:1404. Heldin CH. Structural and functional studies on platelet-derived growth factor. Embo J 1992; 11:4251. Golub TR, Barker GF, Lovett M, Gilliland DG. Fusion of PDGF receptor b to a novel ets-like gene, tel, in chronic myelomonocytic leukemia with t(5;12) chromosomal translocation. Cell 1994; 77:307. Massague J, Czech MP, Iwata K, et al. Affinity labeling of a transforming growth factor receptor that does not interact with epidermal growth factor. Proc Natl Acad Sci U S A 1982; 79:6822. Buckley-Sturock A, Wodward SC, Senior RM, et al. Differential stimulation of collagenase and chemotactic activity in fibroblasts derived from rat wound repair tissue and human skin by growth factors. J Cell Physiol 1989; 138:70. de Larco JE, Todaro GJ. Growth factors from murine sarcoma virus-transformed cells. Proc Natl Acad Sci U S A 1978; 75:4001. O'Grady P, Liu Q, Huang SS, Huang JS. Transforming growth factor b (TGF-b) type V receptor has a TGF-b-stimulated serine/threonine-specific autophosphorylation activity. J Biol Chem 1992; 267:21033. O'Grady P, Kuo MD, Baldassare JJ, et al. Purification of a new type high molecular weight receptor (type V receptor) of transforming growth factor beta (TGF-beta) from bovine liver. Identification of the type V TGF-beta receptor in cultured cells. J Biol Chem 1991; 266:8583 Laiho M, DeCaprio JA, Ludlow JW, et al. Growth inhibition by TGF-b linked to suppression of retinoblastoma protein phosphorylation. Cell 1990; 62:175. Behringer RR, Finegold MJ, Cate RL. Müllerian-inhibiting substance function during mammalian sexual development. Cell 1994; 79:415. Rosenthal SM, Brown EJ, Brunetti A, Goldfine ID. Fibroblast growth factor inhibits insulin-like growth factor-II (IGF-II) gene expression and increases IGF-I receptor abundance in BC3H-1 muscle cells. Mol Endocrinol 1991; 5:678. Weil-Hillman G, Schell K, Segal DM, et al. Activation of human T cells obtained pre- and post-interleukin-2 (IL-2) therapy by anti-CD3 monoclonal antibody plus IL-2: implications for combined in vivo treatment. J Immunother 1991; 10:267. Lotze MT, Chang AE, Seipp CA, et al. High-dose recombinant interleukin 2 in the treatment of patients with disseminated cancer. Responses, treatment- related morbidity, and histologic findings. JAMA 1986; 256:3117. Rosenstein M, Ettinghausen SE, Rosenberg SA. Extravasation of intravascular fluid mediated by the systemic administration of recombinant interleukin J Immunol 1986; 137:1735. Donahue RE, Seehra J, Metzger M, et al. Human IL-3 and GM-CSF act synergistically in stimulating hematopoiesis in primates. Science 1988; 241:1820. Banchereau J, de Paoli P, Valle A, et al. Long-term human B cell lines dependent on interleukin-4 and antibody to CD40. Science 1991; 251:70. Parronchi P, De Carli M, Manetti R, et al. Aberrant interleukin (IL)-4 and IL-5 production in vitro by CD4+ helper T cells from atopic subjects. Eur J Immunol 1992; 22:1615. Obiri NI, Hillman GG, Haas GP, et al. Expression of high affinity interleukin- 4 receptors on human renal cell carcinoma cells and inhibition of tumor cell growth in vitro by interleukin-4. J Clin Invest 1993; 91:88. Huston MM, Moore JP, Mettes HJ, et al. Human B cells express IL-5 receptor messenger ribonucleic acid and respond to IL-5 with enhanced IgM production after mitogenic stimulation with Moraxella catarrhalis. J Immunol 1996; 156:1392. Martinez OM, Ascher NL, Ferrell L, et al. Evidence for a nonclassical pathway of graft rejection involving interleukin 5 and eosinophils. Transplantation 1993; 55:909. Black K, Garrett IR, Mundy GR. Chinese hamster ovarian cells transfected with the murine interleukin-6 gene cause hypercalcemia as well as cachexia, leukocytosis and thrombocytosis in tumor-bearing nude mice. Endocrinology 1991; 128:2657. Lynch DH, Miller RE. Interleukin 7 promotes long-term in vitro growth of antitumor cytotoxic T lymphocytes with immunotherapeutic efficacy in vivo. J Exp Med 1994; 179:31. Endo H, Akahoshi T, Takagishi K, et al. Elevation of interleukin-8 (IL-8) levels in joint fluids of patients with rheumatoid arthritis and the induction by IL-8 of leukocyte infiltration and synovitis in rabbit joints. Lymphokine Cytokine Res 1991; 10:245. Katsikis PD, Chu CQ, Brennan FM, et al. Immunoregulatory role of interleukin 10 in rheumatoid arthritis. J Exp Med 1994; 179:1517. Enk AH, Angeloni VL, Udey MC, Katz SI. Inhibition of Langerhans cell antigen-presenting function by IL-10. A role for IL-10 in induction of tolerance. J Immunol 1993; 151:2390. Okamura H, Tsutsi H, Komatsu T, et al. Cloning of a new cytokine that induces IFN-g production by T cells. Nature 1995; 378:88. Parnet P, Garka KE, Bonnert TP, et al. E. IL-1Rrp is a novel receptor-like molecule similar to the type I interleukin-1 receptor and its homologues T1/ST2 and IL-1R AcP. J Biol Chem 1996; 271:3967. Rothe H, Jenkins NA, Copeland NG, Kolb H. Active stage of autoimmune diabetes is associated with the expression of a novel cytokine, IGIF, which is located near Idd2. J Clin Invest 1997; 99:469. Scheurich P, Thoma B, Ucer U, Pfizenmaier K. Immunoregulatory activity of recombinant human tumor necrosis factor (TNF)-a : induction of TNF receptors on human T cells and TNF-a-mediated enhancement of T cell responses. J Immunol 1987; 138:1786. Aggarwal BB, Graff K, Samal B, et al. Regulation of two forms of the TNF receptors by phorbol ester and dibutyryl cyclic adenosine 3‹,5‹-monophosphate in human histiocytic lymphoma cell line U-937. Lymphokine Cytokine Res 1993; 12:149. Grell M, Scheurich P, Meager A, Pfizenmaier K. TR60 and TR80 tumor necrosis factor (TNF)-receptors can independently mediate cytolysis. Lymphokine Cytokine Res 1993; 12:143. Aderka D, Wysenbeek A, Engelmann H, et al. Correlation between serum levels of soluble tumor necrosis factor receptor and disease activity in systemic lupus erythematosus. Arthritis Rheum 1993; 36:1111. Fuchs P, Strehl S, Dworzak M, et al. Structure of the human TNF receptor 1 (p60) gene (TNFR1) and localization to chromosome 12p13 [published erratum appears in Genomics 1992; 13(4):1384]. Genomics 1992; 13:219. McDermott MF, Aksentijevich I, Galon J, et al. Germline mutations in the extracellular domains of the 55-kDa TNF receptor, TNFR1, define a family of dominantly inherited autoinflammatory syndromes. Cell 1999; 97:133. Sen GC, Lengyel P. The interferon system. A bird's eye view of its biochemistry. J Biol Chem 1992; 267:5017. Uze G, Lutfalla G, Bandu MT, et al. Behavior of a cloned murine interferon a/b receptor expressed in homospecific or heterospecific background. Proc Natl Acad Sci U S A 1992; 89:4774. Uze G, Lutfalla G, Gresser I. Genetic transfer of a functional human interferon a receptor into mouse cells: cloning and expression of its cDNA. Cell 1990; 60:225. Bazan JF. Structural design and molecular evolution of a cytokine receptor superfamily. Proc Natl Acad Sci U S A 1990; 87:6934. Bazan JF. A novel family of growth factor receptors: a common binding domain in the growth hormone, prolactin, erythropoietin and IL-6 receptors, and the p75 IL-2 receptor b-chain. Biochem Biophys Res Commun 1989; 164:788. Hunter T. Signal transduction. Cytokine connections. Nature 1993; 366:114. Stahl N, Farruggella TJ, Boulton TG, et al. Choice of STATs and other substrates specified by modular tyrosine-based motifs in cytokine receptors. Science 1995; 267:1349.

CHAPTER 174 COMPENDIUM OF GROWTH FACTORS AND CYTOKINES Principles and Practice of Endocrinology and Metabolism

CHAPTER 174 COMPENDIUM OF GROWTH FACTORS AND CYTOKINES BHARAT B. AGGARWAL General Features of Cytokines Conclusion Chapter References

GENERAL FEATURES OF CYTOKINES Cytokines are polypeptide hormones, which, when secreted by a cell, affect the growth and metabolism of either the same (autocrine) or a neighboring cell (paracrine). Growth factors, which are also peptide hormones, induce cellular proliferation. The appellations, cytokines and growth factors, have been used interchangeably. Although the hormones produced by endocrine organs are not considered cytokines, extensive research has revealed no definitive criteria that can distinguish between cytokines, growth factors, and endocrine hormones, except that not all endocrine hormones are polypeptides. Perhaps the first cytokine to be discovered was interferon (IFN)—found in 1957 by Issacs and Lindenmann as a soluble factor produced by cells after exposure to heat-inactivated influenza virus. Among the growth factors, the nerve growth factor (NGF) and epidermal growth factor (EGF) were the first to be identified. Within the last two decades, an avalanche of novel cytokines and growth factors have been reported, and several monographs and reviews have been written on this subject.1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27 and 28 In this chapter, the author provides a compendium of cytokines and growth factors, with most information summarized in tables. For originalreferences, the reader should consult the monographs listed at the end of the chapter (see also Chap. 173, Chap. 180 and Chap. 212). Within the last two decades, as many as 200 cytokines have been identified, and each year more are added to the list. This rapid development is most likely due to the availability of the human genome sequence, which allows searches for structurally homologous genes. Some of the general features of most cytokines are given in Table 174-1. Although it was initially believed that most cytokines are secreted, it now appears that several of them are transmembrane proteins. For instance, of the 17 known members of the tumor necrosis factor (TNF) superfamily, all except lymphotoxin (also called TNF-b) are type II transmembrane proteins (i.e., its carboxyl terminus is extracellular as opposed to the amino terminus in type I). Why cytokines exist in transmembrane form is not clear, but one reason might be the localized action of the cytokine. Once secreted by proteolytic cleavage, the cytokine could have pathologic effects. Thus, the secretion of cytokines appears to be a highly regulated process.

TABLE 174-1. Common Features of Most Cytokines

Although cytokines and growth factors were initially thought to be produced by specific cell types, sensitive methods of detection have now revealed that most cell types have the potential to express the gene for the various cytokines. Table 174-2 lists some of the known examples, indicating that the same cytokine can be produced by a wide variety of different cell types. For example, TNF—once thought to be produced only by macrophages—later was found to be expressed by a wide variety of cells, tissues, and organs. Why a cytokine should be produced by so many different cell types is unclear, but the diversity may explain the multipotential activities assigned to this cytokine.

TABLE 174-2. Cytokine Production by Different Cell Types

Various observations indicate that biologic systems are highly redundant and overlapping; this may serve the purpose of protection against system-wide failure in the event that an individual component fails. Table 174-3 lists cytokines that have antiviral, inflammatory, growth modulatory, chemotactic, angiogenic, and antiangiogenic activities. At least four different gene products, referred to as IFN, have been reported to have antiviral activities. Why several cytokines exhibit similar activities is unclear, but perhaps this is a mechanism developed by nature to ensure the normality of the function in case a given cytokine gene is deleted or mutated.

TABLE 174-3. Cytokines with Multiple Biologic Responses

The physicochemical characteristics of most cytokines are very similar. As listed in Table 174-3, the genes for most cytokines have mRNA of ~1 to 2 kb encoding a protein with molecular mass 10 µU/mL. Typical means used to accomplish this are either a 72-hour fast with sequential measurements of plasma glucose, insulin, and C-peptide or similiar measurements before and during induced hypoglycemia52,53 (see Chap. 158). Insulinomas may often be localized by sequential measurement of hepatic vein insulin gradients after calcium infusion into pancreatic arteries54 The initial therapy for insulinoma should be a distal pancreatectomy that removes ~85% of the gland55,56 This procedure is favored for three reasons. First, multicentric insulin-secreting tissue is present in 92% to 100% of patients with MEN155,57 compared with only 10% in those with sporadic tumors. Second, 5% to 15% of the tumors are malignant5,6 Finally, complications of pancreatic endocrine and exocrine insufficiency are greatly reduced compared with the frequency after total pancreatectomy (see Chap. 160). If hypoglycemia persists after surgery, diazoxide may be used; if metastatic disease is present, streptozocin, dacarbazine, or a somatostatin analog58 may be used. A glucagon-secreting tumor (see Chap. 220) without the necrolysis syndrome has been documented in a euglycemic woman with MEN1,59 and hyperglucagonemia has been reported in five of six patients with MEN1.60 Nonfunctioning A-cell adenomas also have been observed. Many reports of functioning glucagonomas are poorly documented, and in most suspected cases, cortisol or growth hormone excess or diabetes after pancreatectomy explains the hyperglycemia. Increased vasoactive intestinal peptide levels have been demonstrated, and one of the patients in the original Verner and Morrison40 series who had watery diarrhea and hypokalemia (see Chap. 182 and Chap. 220) had MEN1. PITUITARY DISEASE The incidence of pituitary disease in MEN1 ranges from 47% (clinical) to 94% (autopsy)5,8 There is a spectrum, ranging from hyperplasia through adenoma to cancer. In a large study, the tumor types and their frequencies were as follows: chromophobe, 44% (one cancer); eosinophilic, 35%; basophilic, 9%; mixed, 11%; and miscellaneous, 5%5 Manifestations of pituitary tumor accounted for only 4% of the presenting complaints in patients in that review, but may be as high as 32%. Tumor mass effects may produce headache, visual field abnormality, and monotropic or general hypopituitarism (see Chap. 11, Chap. 17 and Chap. 19). The secretion of hormones by the tumor produces clinically identifiable endocrine manifestations: Prolactin causes galactorrhea-amenorrhea,61,62 growth hormone causes acromegaly, and adrenocorticotropic hormone (ACTH) causes Cushing disease. Nonfunctioning pituitary tumors are in the minority. Presumably, many of the hyperplasias result from ectopic secretion of hypothalamic-releasing hormones by the pancreas or other primary syndrome tumors63 The diagnosis and treatment of pituitary disease in MEN1 are similar to those of sporadic disease (see Chap. 17). The disease is confirmed by finding high or inappropriately low blood concentrations of prolactin, growth hormone, ACTH, or gonadotropins associated with an abnormality identified by gadoliniumenhanced MRI. Patients with prolactin-secreting microadenomas (see Chap. 13) may be treated with cabergoline, bromocriptine or pergolide unless subsequent mass effects dictate pituitary adenectomy. Macroadenomas, whether they secrete prolactin, growth hormone, ACTH, or nothing, require hypophysectomy, often followed by radiotherapy and medical therapy—such as octreotide for acromegaly. Larger or invasive tumors may require several treatment modalities. Endocrine deficiency states secondary to the tumor or its treatment necessitate replacement therapy. PARATHYROID DISEASE Parathyroid disease is the most highly penetrant as well as earliest presenting manifestation of MEN1, affecting 87% to 99% of patients. In many families, this component is detected first, with the discovery of MEN1 occurring subsequently. This chronology may occur because the parathyroid disease matures earlier than other components; because screening serum calcium determinations may be used more frequently than other markers; and because kindred members may be incompletely studied. There is universal agreement that parathyroid hyperplasia, diffuse or asymmetric, is the rule in MEN1. Older patients may have super-imposed adenomatous change. The parathyroid hyperplasia may be caused by a growth factor found in the plasma of patients with MEN1 that is highly mitogenic for bovine parathyroid cells64 Although involvement of the parathyroid gland is the most frequent component, symptoms of hyperparathyroidism occur in only 9% to 30% of patients with MEN1.5,6 The symptoms, signs, and complications of hyperparathyroidism are not significantly different from those in sporadic, familial, or MEN2–associated hyperparathyroidism (see Chap. 58). The diagnosis of parathyroid disease is established by finding increased concentrations of serum calcium and parathyroid hormone. Not all patients with familial hypercalcemia have MEN1. The differential diagnosis also includes other diseases inherited in an autosomal dominant pattern: familial hyperparathyroidism, MEN2A, and familial hypercalcemic hypocalciuria. The last disorder is suggested by an absence of clinical signs, normal concentrations of parathyroid hormone, onset of hypercalcemia in the first decade, hypermagnesemia, and a low urinary calcium/creatinine clearance ratio60 Because localization scans do not uniformly image all involved glands, they are not recommended. The indications for parathyroidectomy in patients with MEN1 are similar to those in patients with sporadic disease: bone disease (low bone mineral density, increased bone alkaline phosphatase, radiographic evidence of periosteal resorption or osteitis fibrosa cystica fracture), renal disease (increasing creatinine, active metabolic stone disease, nephrocalcinosis), a serum calcium value of 11.0 mg/dL or higher, or other complications suspected to be secondary to the hypercalcemia (e.g., cardiac arrhythmias, depressive symptoms, fatigue). A lesser indication might be detection of hypercalcemia at an early age, which would create a longer period at risk for the development of complications. Operation to restore normal serum calcium and thus reduce gastrin and gastric acid secretion has also been advocated. The surgical procedure of choice for all patients with MEN1 is subtotal parathyroidectomy with preservation of 30 to 50 mg parathyroid tissue or total parathyroidectomy with autotransplantation of some removed tissue. In patients with hyperplasia for whom an operation is performed, 93% are cured initially, but 23% also have permanent parathyroprivic hypoparathyroidism,65 and about half of those who have initially successful operations develop recurrence in the long term. Because of the high rate of hypoparathyroidism, autotransplantation and cryo-preservation of parathyroid tissue are recommended (see Chap. 62). In those who become permanently hypoparathyroid, human parathyroid hormone 1-34 treatment should be helpful65a ASSOCIATED DISEASES Carcinoid tumors may occur in the bronchus, gastroduodenum, and thymus of 5% to 9% of patients with MEN1.66 They may be locally invasive, multifocal, and occasionally metastatic. Clinical symptoms are rare,26 although occasional affected patients have a high urinary content of 5-hydroxyindoleacetic acid. Twelve to 18% of patients have thyroid disease, such as follicular adenoma, nodular hyperplasia, colloid goiter, diffuse enlargement with thyrotoxicosis, Hashimoto disease, or malignancy5,26 This does not represent an increased prevalence compared with the general population. Adrenocortical disease, including adenoma, diffuse hyperplasia, nodular hyperplasia, and, rarely, cancer, occurs in 10% to 38% of patients67 In one report, hypertension was associated with hyperaldosteronism5 Most of the adrenal abnormalities are nonfunctional and a few may be a reflection of pituitary or paraneoplastic ACTH secretion. Multiple subcutaneous, pleural, and retroperitoneal lipomas occur in 13% to 38% of patients68 The NIH group has noted that 88% of their patients have facial angiofibromas and 72% have skin collagenomas. SCREENING FOR MULTIPLE ENDOCRINE NEOPLASIA TYPE 1 Primary relatives of index cases with MEN1 and patients having certain diseases should be screened for MEN1. These diseases include ZES, insulinoma, or multiple gland hyperparathyroidism not associated with other known familial disease. MEN1 has been found in 26% to 33% of patients with ZES,43,45 4% to 10% of patients with

insulinoma,55 14% of families with familial hyperparathyroidism, and 16% of patients with parathyroid hyperplasia60 Before adulthood, major morbidity is rare and cancer related to MEN1 has not been reported; therefore, screening beginning at age 18 seems appropriate26 A positive family history is helpful; however, a negative family history does not rule out this entity. The screening history and physical examination should emphasize symptoms and signs of hypercalcemia, peptic ulcer disease, hypoglycemia, galactorrhea-amenorrhea, acromegaly, Cushing syndrome, visual changes, and hypopituitarism. In patients without evidence of disease, it is sufficient to measure serum calcium, gastrin, and prolactin69 The author is content to repeat core studies at intervals of 5 years in asymptomatic persons at risk. In such screening programs, asymptomatic hypercalcemia is the most frequently described abnormality. Routine screening with pancreatic polypeptide, growth hormone, cortisol, and MRI of the pituitary gland is inadvisable because of the low yield. Commercial testing for the MEN1 germline mutation is not available at this writing. When it is, identification of the specific mutation in an index case should be done first. This will allow selective search by the laboratory for only 1 of the 32 currently known mutations; this should reduce cost. Identification of carrier status will focus subsequent core screening measurements. on the affected and excuse the unaffected. This will provide a sense of relief and reduce laboratory test expenses for the unaffected.70 For the 6% of MEN1–affected families who have no detected mutation as of yet, routine testing at selected intervals must go on. In MEN1 carriers, the author measures core serum calcium, prolactin, gastrin, growth hormone, and cortisol at intervals of 2 years. After an initially negative pituitary MRI scan, this test is repeated for clinical suspicion or at 5-year intervals.

MULTIPLE ENDOCRINE NEOPLASIA TYPE 2 MEN2 has three subsets (see Table 188-1). Patients with MEN2A have a normal phenotype and commonly have parathyroid disease. Patients with MEN2B have the ganglioneuroma phenotype and only rarely have parathyroid disease. Patients with FMTC have MTC only. The incidence of principal organ involvement in patients with MEN2A has been reported in several publications and is summarized in Table 188-3. Precursor stages exist for all organ components: C-cell hyperplasia precedes MTC (Fig. 188-2); diffuse nonnodular adrenomedullary hyperplasia precedes pheochromocytoma, and occult, normocalcemic parathyroid hyperplasia is observed29,30 and 31 Generally, asynchronous clinical manifestations lead to detection, often beginning with the third decade. MTC is found in all patients for two reasons: its clinical presentation occurs earlier; and measurement of calcitonin, a tumor marker for MTC, provides an exquisitely sensitive means for the early detection of occult disease in family members at high risk. Both MTC and pheochromocytoma may cause death secondary to metastatic spread, and pheochromocytoma may cause hypertensionor hypotension-associated death. Occasional patients with MEN2A or FMTC may have cutaneous lichen amyloidosis, as well as Hirschsprung disease73,74 and 75

TABLE 188-3. Multiple Endocrine Neoplasia Type 2A: Combinations of Thyroid, Adrenal, and Parathyroid Disease

FIGURE 188-2. The spectrum of C-cell disease in a family with MEN2A. Upper panel, Microscopic example of the multicentric C-cell hyperplasia (arrow) found above the thyroid follicle in a 12-year-old girl. Middle panel, Surgical specimen showing 1.5-cm medullary thyroid carcinoma (MTC; arrow) in the right upper lobe and smaller lesions (arrows) in the left upper lobe of the girl's mother, age 29 years. Lower panel, Surgical specimen showing large masses of MTC in both lobes of the mother's 38-year-old brother. Larynx and trachea have been sketched.

MULTIPLE ENDOCRINE NEOPLASIA TYPE 2B The main features of the MEN2B syndrome are clearly established.76,77,78,79 and 80 Hypotheses about the development of this pheno-type, and the relationship of this “sympathoenteric lineage” to the RET mutation have been presented81 A marfanoid habitus is recognized by excessive limb length, loose-jointedness, scoliosis, and anterior chest deformities. However, these patients do not have the ectopia lentis and cardiovascular abnormalities of Marfan syndrome (see Chap. 189). The diffuse ganglioneuroma-tosis may present as yellow or white nodular or diffuse thickening of the tarsal plates of the eyes, associated with thickened corneal nerve fibers; pink, yellow, or translucent hemispherical nodules studding the tip and anterior third of the tongue; elongated projections posterior to each orolabial commissure; and nodules within the lips, causing them to appear everted, patulous, and lumpy (Fig. 188-3). These neural tumors are called ganglioneuromas rather than neuromas; they contain ganglion cells and tortuous nerve trunks with internal structural disarray. Functionally, alimentary ganglioneuromatosis produces gastrointestinal symptoms (constipation, diarrhea, dysphagia) and signs (megacolon, diverticulosis). The neuromuscular abnormalities range from localized (particularly to the peroneal muscles) to diffuse muscle weakness and may also be associated with sensory abnormalities78

FIGURE 188-3. Abnormal facies, thickened lips, and ganglioneuromas of tongue in four affected members of a family with multiple endocrine neoplasia type 2B (MEN2B). Left to right, Grandfather, two sons, and 2-year-old granddaughter. (From Sizemore GW, Heath H III, Carney JA. Multiple endocrine neoplasia type 2. Clin Endocrinol Metab 1980; 9:299.)

MEDULLARY THYROID CARCINOMA MTC originates from parafollicular, or C cells (so called because of their synthesis, storage, and secretion of calcitonin). The tumor has a solid, nonfollicular histologic pattern with amyloid in the stroma and a high incidence of lymph node metastases. In patients with MEN2, the MTC is located predominantly in the upper and middle thirds of the lateral thyroid lobes and, in contrast to sporadic MTC, is bilateral, multicentric, and almost universally associated with C-cell hyperplasia. CLINICAL MANIFESTATIONS One-third of all cases of MTC in patients with MEN2 present as a thyroid nodule, goiter, or mass, or as cervical lymphadenopathy, in the third to sixth decades. Less common clinical presentations include recognition because of an associated process, such as hyperparathyroidism, pheochromocytoma, or the ganglioneuroma phenotype in MEN2B. Rarely, there is an associated hereditary form of cutaneous lichen amyloidosis or Hirschsprung disease. Paraneoplastic syndromes occur: Cushing syndrome secondary to secretion of ACTH by the MTC; diarrhea, which is presumed to be caused by calcitonin-induced gastrointestinal secretion of water and electrolytes in persons with a large tumor volume; spontaneous or alcohol-induced flushing presumed to be caused by tumor secretion of other humoral substances—probably the vasodilating calcitonin gene–related peptide65,66 and 67 The remaining cases of MTC in patients with MEN2 are detected at earlier stages and ages through the screening of asymptomatic relatives at high risk. CALCITONIN TESTING MTC secretes calcitonin, a biologic marker that is uniquely useful for detection and follow-up of MTC14,74 The measurement of this hormone is widely available and cost effective, making it applicable to large numbers of patients. It is a highly sensitive discriminator that detects, quantitatively, a difference between those with and those without MTC; when sensitive assays are used properly, there are few false-negative or false-positive responses. Several caveats apply to the use of calcitonin measurement in these patients. Clinicians are advised to ensure that referral laboratories have sufficiently standardized test information in normal persons so that the test they plan to use will allow detection of the majority of MTC-affected patients at a reasonable age. Hypercalcitonemia is not specific for MTC; for reviews of this subject, see Chapter 40 and Chapter 53. Stimulation testing should not be done in patients with unequivocally high calcitonin values, because rapid increases in the concentration of calcitonin or, probably, calcitonin gene–related peptide, have occasionally caused vasodilatation and shock in such patients. Most patients with sporadic MTC have high basal concentrations of calcitonin17 However, 27% of patients with familial MTC—young patients with minimal tumor—have basal concentrations within the normal range, and repeated basal measurements alone are not helpful83 To identify C-cell hyperplasia or precursor forms of MTC, provocative tests of calcitonin secretion are needed to increase the marker's sensitivity. Tests using infusions of calcium, pentagastrin, or both appear to be best. Pentagastrin stimulation is particularly useful83,84 In this method, a bolus intravenous injection of 0.5 µg/kg is given over 5 seconds in 3.0 to 5.0 mL of 0.9% sodium chloride, and blood samples are taken at 0, 1.5 to 2.0, and 5 minutes for measurement of immunoreactive calcitonin. This test, when used with a well-standardized calcitonin assay, is a sensitive means of detecting and following MTC. The author's experience indicates that pentagastrin testing uniformly detects familial MTC before thyroidectomy in the 27% of patients who have normal basal calcitonin concentrations as well as the 15% who have false-negative responses to calcium infusion. In 20 years, the author has had only two initial false-negative results (which converted to positive at older ages) and two false-positive results. Others have obtained comparable results with a combined calcium-pentagastrin test85 In the latter procedure, a 1-minute injection of calcium, 2 mg/kg per minute, is followed by an injection of pentagastrin, 0.5 µg/kg in 5 seconds, with blood sampling during 15 minutes. At this time, pentagastrin is unavailable in the United States—a marketing decision. Although it remains available in Europe, many clinicians in the United States will have to use infusions of calcium as outlined in referral laboratory catalogs for stimulation testing86,87 SCREENING FOR MEDULLARY THYROID CARCINOMA MULTIPLE ENDOCRINE NEOPLASIA TYPE 2 KINDREDS Genetic analysis of RET protooncogene mutations has been a major advance toward the goal of early detection of affected patients with presymptomatic and premetastatic MTC. Detection of the mutations causing MTC in patients within kindreds having MEN2 or FMTC allows gene carrier status to be established at birth in the majority of affected members in kindreds having risk, and should allow the same in persons who are at risk because of associated diseases. An international consortium that studied the mutations in 477 families in which two or more members had MTC has reported excellent results88 MEN2A mutations were correctly identified in 98% of affected families, MEN2B mutations were identified in 95% of affected families, and FMTC mutations were identified in 88% of affected families. The consortium also found that between 2% and 12% of these MTC-affected families, respectively, were not detected by genetic analysis. Perhaps other mutations will be found that explain the present false-negative percentage. This group and others have not yet reported false-positive results. The study also demonstrated a relationship between specific RET protooncogene mutations and the various MTC phenotypes (Table 188-4). Initially, it was expected that all patients with MEN2 and a RET mutation would have medullary carcinoma or its precursor and that calcitonin testing would detect all. That is not the case. In MEN2, the clinical penetrance of the gene is not complete by age 70, and the prevalence of MTC detected by calcitonin testing through age 31 is 93%89 These delayed expressions may be explained by slow growth of MTC in some patients, insensitive calcitonin assays, or, more probably, unknown epigenetic factors that translate the genetic message into a functional MTC. An analogy should help. Although it is known that persons who have the BRCA 1 or 2 mutations have higher than normal risk for breast carcinoma, all persons with these mutations do not develop breast cancer within a few years of birth. The expression is delayed by unknown factors. Affected persons are advised to have long-term clinical follow-up and repeated mammograms. Gene mutations measure one thing, and stimulation tests or mammograms measure another. Therefore, direct comparisons seem somewhat futile. However, when done, the comparisons between RET and biochemical testing have favored RET, so mutation analysis is generally considered to be superior for detecting risk of MEN2. In a study comparing RET to calcitonin, it was found that 4 of 91 (4.3%) patients with MEN2A had false-negative calcitonin results (that is, they were RET-positive and calcitonin-negative); similarly, it was found that 5 of 51 (9.8%) patients with FMTC had false-negative calcitonin results; all of the “calcitonin-missed” patients were young—younger than 23 years90 Similar results were reported in a second study: 8 of 74 (11%) patients with MEN2, who were mutation-positive and had MTC, had false-negative calcitonin studies; all of these individuals were in the youngest or fifth generation of the 4 five-generation families that were studied91 These two studies also found false-positive pentagastrin results in 8% and 9% of mutation-negative MEN2 patients, respectively. In the second study, all of these were thought to have a nonprogressive C-cell hyperplasia, which probably represents a new entity not related to MEN2. 92 The author's screening procedure follows: Each new case having MTC is screened for a RET mutation; these tests are now widely available. A positive result prompts screening for the mutation in all primary relatives. Those family members who are positive should have cervical exploration; those who are negative need no further evaluation unless indicated clinically at a later point. An initial negative result in the index case screened should be reassuring to family members, because it indicates a 98% chance that the tested patient and his family will not have MEN2. However, because 2% or 12% of index cases with MEN2 or FMTC, respectively, may be mutation-negative, the author advises primary relatives of mutation-negative index cases to be screened once with stimulated calcitonin tests. This reduces the risk of missing a truly affected patient to ~1%. If a mutation-negative, calcitonin-positive family is confirmed to have MTC, screening of the initially negative family members is performed again at ~1 year. If the patient is an adult and found to be negative, screening stops. If the patient is a child, screening continues at yearly intervals through age 18 years. When a mutation is found within a known affected family, confirmation of C-cell disease with pentagastrin testing probably is not necessary. However, a mutation in one person with a known associated disease (e.g., pheochromocytoma), who has no other MEN-affected family members, would cause the author to test with pentagastrin before recommending thyroidectomy. Patients who are mutation-positive–calcitonin-negative are uncommon. They should be provided with information about the inaccuracies of testing, genetic polymorphism, and other issues before undergoing thyroidectomy. At-risk patients who are mutation-negative and calcitonin-positive have been found to have an unusual form of C-cell hyperplasia92 In the familial setting, the C-cell disease is really preinvasive carcinoma or a carcinoma in situ. Outside the setting of MEN, one often finds C-cell hyperplasia in normal persons (who may have hypercalcitonemia), in older persons, and in those with hyperparathyroidism and high levels of gastrin92,93 and 94 In the author's experience, 19% of patients presumed to have sporadic disease were index cases to familial MTC,6 and 4% to 8% of patients with presumed sporadic pheochromocytoma have MTC11,95 The author would not routinely screen persons with hyperparathyroidism or a solitary nodule for MTC using calcitonin measurements because his previous limited trials were unrewarding. Nonetheless, some reports96,97 have suggested that such measurements will identify previously unsuspected cases of MTC. The author is concerned, however, that if widely applied using different immunoassays, some of which are insufficiently characterized or specific, there may be an unacceptably high rate of false positives similar to that in a case report of a patient with Hashimoto thyroiditis98 As the majority of MTC cases are sporadic, analysis for RET mutations will not be cost effective. Until further studies provide solutions to the nodule problem, the author will continue to use stimulated calcitonin tests only when the nodule workup has no suggestion of Hashimoto disease and the fine-needle aspiration results are quite atypical, unusual or strongly suggestive of MTC. Perhaps the use of RNA extracted from fine-needle aspiration biopsies as suggested by Takano, et al. will be helpful98a Finally, thyroid scintiscan should not be used as a screening device because of low sensitivity99 TREATMENT Patients with MTC should have total thyroidectomy at an early age after the exclusion, surgical excision, or medical treatment of any adrenomedullary disease100 Total thyroidectomy is recommended for several reasons: In MEN2, the tumor is bilateral; 19% of presumed sporadic cases are actually MEN2; and the report of a high calcitonin concentration after incomplete thyroidectomy mandates secondary surgery. In the hands of a skilled surgical team, the incidence of vocal cord paralysis and hypoparathyroidism should be low. The operation should include central lymph node clearance and ipsilateral functional or modified neck dissection for tumors >2.0 cm

or if palpable adenopathy is present. The lymph nodes in the central compartment of the neck should be dissected at the time of primary total thyroidectomy, because metastasis frequently occurs in this region100 Lymph nodes in the lateral compartments of the neck should be sampled and, if tumor is found, lymphadenectomy should be done. If either the jugular vein or the sternocleidomastoid muscle is involved with tumor, resection is indicated. Whether to perform a modified neck dissection or a radical neck dissection is imperfectly known100,101 and 102; a retrospective, randomized trial of this question is needed. After surgery, all patients should receive replacement with levothyroxine and be evaluated with calcitonin stimulation tests. Those patients who have undetectable or normal immunoreactive calcitonin concentrations after surgery may be followed up with annual clinical review and provocative testing at intervals of 2 to 5 years. The author continues to recommend that children have thyroidectomy by age 5 for multiple reasons. Formed MTC has been found in children aged 6, 12, and 48 months. Metastatic MTC was found at age 5.5 years during primary surgery in a MEN2B patient who died with metastases at age 9.103 An unreported MEN2A patient had metastases at age 6 years. There is no bio-chemical means to predict when metastasis will occur. When experienced endocrine surgeons perform the operation, recurrent nerve paralysis, permanent hypoparathyroidism, and anesthesia difficulties are reported to be negligible, and there is no study documenting increased complications in patients operated on at these young ages104,105 and 106 Pediatric endocrinologists have little difficulty managing thyroid hormone replacement in children. Clinicians are cautioned that avoidable delays in resection allow metastatic MTC, which causes significant morbidity and mortality. A continuously developing picture suggests that the biologic virulence of MTC differs according to its associations, with most to least aggressive behavior as follows: MEN2B, sporadic disease, and MEN2A18,71,72,107 In one large group of patients, death from MTC has occurred in 50% of those with MEN2B, 31% of those with sporadic disease, and 9.7% of those with MEN2A72 In other large studies, 60% of patients with MEN2B have died of MTC, whereas the 10-year survival rate is 55% for sporadic cases and 95% for patients with MEN2A71,108 Multivariate analysis of these data suggests a 7.7-fold greater risk of dying of sporadic disease than of MEN2A–associated MTC. Factors having a negative effect on survival have been summarized: They include incomplete primary operation, male sex, older age, large primary tumor size, extrathyroidal invasion, nodal metastasis, distant metastasis, high plasma CT, high carcinoembryonic antigen (CEA), negative tumor staining for amyloid, and the MEN2B phenotype109 In the author's continuing experience, no MEN2A patient who has had total thyroidectomy during the first decade of life has had detectable residual tumor. By contrast, the incidence of persistence progresses from 31% in those undergoing operation in the second decade to 67% in those operated on in the seventh decade17 Similarly, others have reported a 24% versus 1.5% mortality for patients with MEN2A in whom MTC was discovered when the tumor was symptomatic, compared with those in whom it was detected earlier through screening110 Furthermore, 19 of 22 patients identified by prospective screening were free of detectable disease after thyroidectomy111 POSTOPERATIVE FOLLOW-UP After operation for MTC, a stimulated-calcitonin test is done as soon as patients have stable, normal serum calcium concentrations. If the calcitonin result is negative, repeat testing is advised in 1 year and then should continue at 2- to 5-year intervals. If the initial result is positive, repeat testing is done for confirmation. Because calcitonin is more sensitive and specific than other tumor markers for MTC, generally the author does not use markers such as CEA in the follow-up of patients. Patients who have high or steadily increasing CEA, particularly values that are relatively high compared to calcitonin, have advancing metastatic disease and a very poor prognosis. What should be advised for patients who have occult MTC after primary operation (e.g., those with high immunoreactive calcitonin concentrations and, by the usual techniques, no overt disease)? Vigorous attempts should be made to eradicate persistent tumor in those having tumors with known biologic virulence such as MEN2B or sporadic MTC or those with some of the poor prognostic factors above. Three additional techniques may help determine which patients may have a poor prognosis. The first is immunostaining of primary tumors; patients with calcitonin-poor tumors have poor survivorship compared to those with calcitonin-rich tumors112 The second is measurement of the DNA content of tumors; tumors in nonsurviving patients with MTC have higher DNA contents than those in surviving patients113 The third is calculation of the doubling time of the calcitonin concentration; rapid doubling times have been reported in patients who have tumors that recur clinically within 5 years or that are associated with a reduced 3-year survival rate114 Because the long-term survival data clearly indicate that the natural history of MTC in MEN2A is biologically less aggressive, more conservative approaches are appropriate, unless the clinical course becomes virulent. In most patients with MEN2A who have occult disease and modestly elevated but relatively unchanging calcitonin concentrations during sequential follow-up, present data suggest that vigorous attempts to localize, to reoperate, or otherwise to eradicate tumor tissue are of little benefit. The sites of MTC metastasis, in order of decreasing frequency, are cervical lymph nodes, lung, mediastinum, liver, bone, adrenal, heart, and pleura. Many techniques are available for the localization of metastases; their usefulness is summarized in Table 188-5.115,116 Chest roentgenograms may show diffuse interstitial lung infiltrates, particularly in the middle and lower lobes. CT of the lungs, mediastinum, and liver may be helpful. Isotope scans may document bone metastasis, but are of little value in liver metastasis. Hepatic angiograms demonstrate vascular-pattern metastases of MTC more reliably than do isotope scans or CT. If these methods do not identify the sites of metastases, selective venous catheterization, with blood samples taken (with or without provocation) for calcitonin, locate more than half of persistent tumors. Reoperation surgery is recommended for multiple reasons: (a) when a combination of localization studies suggests local rather than disseminated metastasis allowing a curative intent, (b) when vital structures need to be protected from tumor invasion, and (c) if debulking will reduce secondary side effects (such as medullary diarrhea).

TABLE 188-5. Medullary Thyroid Carcinoma: Localization of Metastases

Initial reoperation experience was disappointing; there were no cures, some complications, and occasional operative mortality. Recent experience with extensive microdissection operations done with curative intent has been better. Up to 38% of selected patients become calcitonin-negative after surgery with acceptable morbidity and no mortality117,118 and 119 It is hoped that these results will not deteriorate or will improve with increased experience and longer follow-up. Other therapies are less successful. Single-agent or combined chemotherapy has not cured patients and generally has not produced greater than a 33% partial response rate. Adjunctive radiotherapy, such as131 iodine after thyroidectomy, does not improve survival rates or decrease the recurrence of MTC; therefore, it should not be used120 A doubled survival rate at 20 years has been shown in patients who receive radiation (either prophylactically or for extensive disease) compared with those who do not121 One group has documented eradication of microscopic disease and 30% (complete) to 45% (partial) remission rates of gross residual MTC by external radiation122 Another study showed that patients treated with external radiation after operation had survival rates similar to those of patients who underwent operation only, despite having more advanced disease123 The author has had radiated patients with similar long-term remissions. Thus, in patients with persistent, aggressive MTC, external radiotherapy seems reasonable (see also Chap. 40). ADRENOMEDULLARY DISEASE—PHEOCHROMOCYTOMA The reported prevalence of adrenomedullary disease has ranged from 12% in a young population to 42% in a study of multiple, mature kindreds followed up for a longer period; the penetrance in individual kindreds has ranged from 6% to 100%124 The pathologic spectrum extends from diffuse medullary hyperplasia to large, multilobular pheochromocytomas that occasionally involve accessory adrenal glands. MEN2 adrenal disease is uniformly bilateral and multicentric in an anatomic sense. Clinically, however, synchronously developed large adrenal masses are unusual; it is more common to encounter a larger tumor on one side and an impalpable hyperplasia or minimal tumor on the other. In an early personal series, tumors were malignant in 24% of cases36 However, the number of malignancies has been low—~10 cases. Symptomatic pheochromocytoma may dominate clinically in some families, but that is unusual125 The clinical syndrome in MEN2 is different from that in sporadic pheochromocytoma because one is looking for low-volume chromaffin-cell disease at earlier stages. Sustained hypertension is unusual. Patients may report anxiety, palpitations, and tachycardia early; headache, pallor, and perspiration are later symptoms. In a study of 17 patients, 8 were asymptomatic and 9 were symptomatic (paroxysms in 6, headache in 2, and dizziness in 1). Nine patients were normotensive, 7 were hypertensive, and 1 was hypotensive36 Episodic rather than sustained hypertension was the rule, and there was no increased incidence of orthostasis. Six of the patients (35%) died because of adrenomedullary disease (cerebral hemorrhage, hypotension, and metastasis in 2 cases each). Deaths resulting from this syndrome component should decrease with earlier case finding and improved

physician awareness. Special problems remain in ensuring compliance with screening and in pheochromocytoma occurring during pregnancy, which may be fatal126 The diagnosis of adrenomedullary disease is made by finding a high content of urinary epinephrine or a high urinary epinephrine/norepinephrine ratio127 Basal, supine plasma epinephrine also is used. Measurements of the urinary content of dopamine, metanephrine, and vanillylmandelic acid may be helpful, but are not as sensitive as the measurement of fractionated catecholamines. In normotensive patients who have normal urinary catecholamine and metabolite levels, results of stimulation tests with glucagon and histamine occasionally are useful. CT and MRI do not detect adrenomedullary hyperplasia but do localize pheochromocytomas as small as 0.5 cm.123 I- or131 I-metaiodobenzylguanidine has been used to detect adrenomedullary disease in patients with MEN2.128,129 and 130 In patients having normal or equivocal CT scan results, this radiolabeled scan has demonstrated a spectrum of adrenomedullary disease ranging from hyperplasia to bilateral pheochromocytomas (Fig. 188-4). All of these techniques for the detection of adrenomedullary disease have limitations in their sensitivity and specificity. However, they can be helpful when used for confirmation, rather than diagnosis, of adrenal disease.

FIGURE 188-4. Posterior 131iodine-metaiodobenzylguanidine (MIBG) scintiscan done at 48 hours (left) and abdominal computed tomographic (CT) scan (right) in a 14-year-old patient with multiple endocrine neoplasia type 2B (MEN2B). The scintiscan shows symmetric adrenal images (arrows). The CT scan shows normal right (arrow) and left (not shown) adrenal glands. Each adrenal gland weighed 6 g and contained macroscopic and microscopic hyperplasia. (From Valk TW, Frager MS, Gross MD, et al. Spectrum of pheochromocytoma in multiple endocrine neoplasia: a scintigraphic portrayal using 131I metaiodobenzyl-guanidine. Ann Intern Med 1981; 94:762.)

When the diagnosis of adrenomedullary disease is established, appropriate a- and b-adrenergic blockade and glucocorticoid preparation should be commenced. Bilateral total adrenalectomy should be performed, which should include exploration of all paraganglionic areas, removal of extraadrenal paragangliomas, and examination of the liver for metastasis (i.e., pheochromocytoma or MTC). The author's recommendation for initial bilateral adrenalectomy is controversial but based on several observations131 First, there is a high likelihood of bilateral involvement; in the Mayo Clinic experience, the grossly uninvolved adrenal gland always was pathologically abnormal. Second, there is mortality associated with adrenomedullary disease, attributable both to the incidence of malignant pheochromocytoma and to the tumor's endocrine function. It is unknown at what tumor stage or size hyperfunction will cause hypertensive crisis or metastasis will occur. Third, a review of the clinical results in 72 patients indicates that 88% required total adrenalectomy, and that 55% of the 20 patients who had undergone initial unilateral adrenalectomy required completion of total adrenalectomy for clinical reasons within a mean period of 4.8 years131 The presence of an adrenal gland that is likely to be diseased requires sequential monitoring, resulting in more exposure to radiation and greater expense. Although patients with adrenal insufficiency require glucocorticoid and mineralocorticoid replacement and medical follow-up, the adrenalectomized state has not constituted an undue hazard (see also Chap. 86, Chap. 87 and Chap. 89). Other groups think that only demonstrably enlarged or abnormal adrenal glands should be removed. They argue that initial unilateral adrenalectomy ameliorates the need for glucocorticoid and mineralocorticoid replacement for at least 5 to 8 years in some patients, and emphasize that death may occur due to adrenal insufficiency. The decision for unilateral or bilateral adrenalectomy is complex and should be considered on an individual basis after assessment of and discussion with each patient. PARATHYROID DISEASE Clinical or anatomic evidence of parathyroid disease has been present in 29% to 64% of patients with MEN2A1,16,17,124 In the patients with parathyroid disease summarized in the cited publications, parathyroid hyperplasia was present in 84% and parathyroid adenoma in 16%. By contrast, overt parathyroid disease is distinctly uncommon in patients with MEN2B, although some have subtle microscopic abnormalities. Patients with MEN2A seldom have symptoms of hypercalcemia. In patients who have clinical disease (recurrent nephrolithiasis, asymptomatic renal calculi, nephrocalcinosis, and chronic renal failure), there is no difference from sporadic hyperparathyroidism. The diagnosis is established for the clinical disease by finding simultaneous high serum calcium and immunoreactive parathyroid hormone concentrations. Patients with MEN2A also may have occult normocalcemic parathyroid hyperplasia. These normocalcemic patients have normal basal immunoreactive parathyroid hormone concentrations, yet, at cervical exploration, hyperplasia of one or more parathyroid glands is confirmed37 This occult disease may be predicted by finding failed suppression of parathyroid hormone during calcium infusion. The indications for parathyroidectomy and the type of operation recommended are similar for patients with MEN2A and MEN1.132 However, because the parathyroid disease more commonly is occult in MEN2A, and because the operation increases the incidence of hypoparathyroidism, the author recommends a conservative approach to parathyroidectomy in MEN2A: Grossly enlarged glands should be removed, and the remaining glands should be marked to facilitate removal if there is future recurrence, or all glands should be removed with a portion of one being autotransplanted (see Chap. 62).

OTHER MULTIPLE ENDOCRINE NEOPLASIA DISEASES There are several other diseases in which neoplasia of the endocrine glands may be present, although among the persons afflicted with the syndrome, the associated neoplasms usually are less common than the neoplasms associated with the previously discussed multiple endocrine neoplasia entities. These other diseases include Gardner syndrome, Peutz-Jeghers syndrome, and Carney complex. The clinical characteristics of these diseases are listed in Table 188-6.

TABLE 188-6. Miscellaneous Diseases in Which Neoplasia of the Endocrine Glands May Occur

CHAPTER REFERENCES 1. Steiner AL, Goodman AD, Powers SR. Study of a kindred with pheochromocytoma, medullary thyroid carcinoma, hyperparathyroidism and Cushing's disease: multiple endocrine neoplasia, type 2. Medicine (Baltimore) 1968; 47:371. 2. Erdheim J. Zur normalen und pathologischen Histologie der glandula Thyroidea, Parathyroidea und Hypophysis. Beitr Pathol Anat 1903; 33:1. 3. Underdahl LO, Woolner LB, Black BM. Multiple endocrine adenomas: report of eight cases in which parathyroids, pituitary and pancreatic islets were involved. J Clin Endocrinol Metab 1953; 13:20.

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

Wermer P. Genetic aspects of adenomatosis of endocrine glands. Am J Med 1954; 16:363. Ballard HS, Frame B, Hartsock RJ. Familial multiple endocrine adenomapeptic ulcer complex. Medicine (Baltimore) 1964; 43:481. Eberle F, Grun R. Multiple endocrine neoplasia, type I (MEN I). Ergeb Inn Med Kinderheilkd 1981; 46:76. Yamagachi K, Kameya T, Abe K. Multiple endocrine neoplasia type 1. Clin Endocrinol Metab 1980; 9:261. Majewski JT, Wilson SD. The MEN-I syndrome: an all or none phenomenon? Surgery 1979; 86:475. Harrison TS, Thompson NW. Multiple endocrine adenomatosis-I and II. Curr Probl Surg 1975; 28. Deveney CW, Deveney KS, Way LW. The Zollinger-Ellison syndrome: 23 years later. Ann Surg 1978; 188:384. Sipple JH. The association of pheochromocytoma with carcinoma of the thyroid gland. Am J Med 1961; 31:163. Cushman P Jr. Familial endocrine tumors: report of two unrelated kindred affected with pheochromocytomas, one also with multiple thyroid carcinomas. Am J Med 1962; 32:352. Manning PC Jr, Molnar GD, Black BM, et al. Pheochromocytoma, hyperpar-athyroidism and thyroid carcinoma occurring coincidentally: report of a case. N Engl J Med 1963; 268:68. Cance WG, Wells SA Jr. Multiple endocrine neoplasia type IIa. Curr Probl Surg 1985; 22:7. Melvin KEW, Miller HH, Tashjian AH Jr. Early diagnosis of medullary carcinoma of the thyroid gland by means of calcitonin assay. N Engl J Med 1971; 285:1115. Keiser HR, Beaven MA, Doppman J, Wells S Jr. Sipple's syndrome: medullary thyroid carcinoma, pheochromocytoma, and parathyroid disease. Ann Intern Med 1973; 78:561. Sizemore GW, Carney JA, Heath H III. Epidemiology of medullary carcinoma of the thyroid gland: a 5-year experience (1971–1976). Surg Clin North Am 1977; 57:633. Sizemore GW, Heath H III, Carney JA. Multiple endocrine neoplasia type 2. Clin Endocrinol Metab 1980; 9:299. Chong GC, Beahrs OH, Sizemore GW, Woolner LB. Medullary carcinoma of the thyroid gland. Cancer 1975; 35:695. Carney JA, Go VLW, Gordon H, et al. Familial pheochromocytoma and islet cell tumor of the pancreas. Am J Med 1980; 68:515. Binkovitz LA, Johnson CD, Stephens DH. Islet cell tumors in von Hippel-Lindau disease: increased prevalence and relationship to the multiple endocrine neoplasias. AJR Am J Roentgenol 1990; 155:501. Larsson C, Skogseid B, Oberg K, et al. Multiple endocrine neoplasia type 1 gene maps to chromosome 11 and is lost in insulinoma. Nature 1988; 332:85. Chandrasekharappa SC, Guru SC, Manickam P, et al. Positional cloning of the gene for multiple endocrine neoplasia type 1. Science 1997; 279:404. Agarwal SK, Kester MB, Debelenko LV, et al. Germline mutations of the MEN 1 gene in familial multiple endocrine neoplasia type 1 and related states. Hum Mol Genet 1997; 7:1169. Lemmens I, Van de Ven WJ, Kas K, et al. Identification of the multiple endocrine neoplasia type 1 (MEN1) gene. The European Consortium on MEN1. Hum Mol Genet 1997; 7:1177. Marx S, Spiegel AM, Skarulis M, et al. Multiple endocrine neoplasia type 1: clinical and genetic topics. Ann Intern Med 1998; 129:484. Mathew CGP, Chin KS, Esaton DF, et al. A linked genetic marker for multiple endocrine neoplasia type 2A on chromosome 10. Nature 1987; 328:527. Simpson NE, Kidd KK, Goodfellow PJ, et al. Assignment of multiple endocrine neoplasia type 2A to chromosome 10 by linkage. Nature 1987; 328:528. Donis-Keller H, Dou S, Chi D, et al. Mutations in the RET proto-oncogene are associated with MEN2A and FWTC. Hum Mol Genet 1993; 2:851. Mulligan LM, Kwok JBJ, Healey CS, et al. Germline mutations of the RET proto-oncogene in multiple endocrine neoplasia type 2A. Nature 1993; 363:456. Mulligan LM, Eng C, Healey CS, et al. Specific mutations of the RET proto-oncogene are related to disease phenotype in MEN 2A and FMTC. Nat Genet 1994; 6:70. Hofstra RM, Landavater RM, Ceccherini I, et al. A mutation in the RET proto-oncogene associated with multiple endocrine neoplasia type 2B and sporadic medullary thyroid carcinoma. Nature 1994; 367:375. Eng C, Smith DP, Mulligan LM, et al. Point mutation within the tyrosine kinase domain of the RET proto-oncogene in multiple endocrine neoplasia type 2B and related sporadic tumours. Hum Mol Genet 1994; 3:237.

33a. Tessitore A, Simsi AA, Pasquali D, et al. A novel case of multiple endocrine neoplasia type 2A associated with two de novo mutations of the RET pro-tooncogene. J Clin Endocrinol Metab 1999; 84:3522. 33b. Pigny P, Bauters C, Wemeau J-L, et al. A novel 9-base pair duplication in RET exon 8 in familial medullary thyroid carcinoma. J Clin Endocrinol Metab 1999; 84:1700. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.

Knudson AG Jr, Strong LC. Mutation and cancer: neuroblastomas and pheochromocytoma. Am J Hum Genet 1972; 24:514. Wolfe HJ, Melvin KEW, Cervi-Skinner SJ, et al. C-cell hyperplasia preceding medullary thyroid carcinoma. N Engl J Med 1973; 289:437. Carney JA, Sizemore GW, Sheps SG. Adrenal medullary disease in multiple endocrine neoplasia, type 2: pheochromocytoma and its precursors. Am J Clin Pathol 1976; 66:279. Heath H III, Sizemore GW, Carney JA. Preoperative diagnosis of occult parathyroid hyperplasia by calcium infusion in patients with multiple endocrine neoplasia, type 2a. J Clin Endocrinol Metab 1976; 43:428. Snyder N, Scurry M, Hughes W. Hypergastrinemia in familial multiple endocrine adenomatosis. Ann Intern Med 1974; 80:321. Friesen SR, Kimmel JR, Tomita T. Pancreatic polypeptide as screening marker for pancreatic polypeptide apudomas in multiple endocrinopathies. Am J Surg 1980; 139:61. Verner JV, Morrison AB. Islet cell tumor and a syndrome of refractory watery diarrhea and hypokalemia. Am J Med 1958; 25:374. Brown CH, Crile G Jr. Pancreatic adenoma with intractable diarrhea, hypokalemia, and hypercalcemia. JAMA 1964; 190:142. Zollinger RM, Ellison EC, O'Dorisio TM, Sparks J. Thirty years' experience with gastrinoma. World J Surg 1984; 8:427. Jensen RT, Gardner JD, Raufman JP, et al. Zollinger-Ellison syndrome: current concepts and management. Ann Intern Med 1983; 98:59. McGruigan JE, Wolfe MM. Secretin injection test in the diagnosis of gastrinoma. Gastroenterology 1980; 79:1324. Malagelada J-R, Edis AJ, Adson MA, et al. Medical and surgical options in the management of patients with gastrinoma. Gastroenterology 1983; 84:1524. Fishbeyn VA, Norton JA, Benya RV, et al. Assessment and prediction of long-term cure in patients with the Zollinger-Ellison syndrome: the best approach. Ann Intern Med 1993; 119:199. Thompson NW, Vinik AI, Eckhauser FE. Microgastrinomas of the duodenum: a cause of failed operations for the Zollinger-Ellison syndrome. Ann Surg 1989; 209:396. Rosch T, Lightdale CJ, Botet JF, et al. Localization of pancreatic endocrine tumors by endoscopic ultrasonography. N Engl J Med 1992; 326:1721. Gibril F, Reynolds JC, Doppman JL, et al. Somatostatin receptor scintigraphy: its sensitivity compared with that of other imaging methods in detecting primary and metastatic gastrinomas. A prospective study. Ann Intern Med 1996; 125:26. Bonfils S, Mignon M. Management of Zollinger-Ellison syndrome with gastric antisecretory drugs. Scand J Gastroenterol 1988; 23(Suppl 146):111. Ellison EC, O'Dorisio TM, Woltering EA, et al. Suppression of gastrin and gastric acid secretion in the Zollinger-Ellison syndrome by long-acting somatostatin (SMS 201–995). Scand J Gastroenterol 1986; 21(Suppl 119):206. Service FJ. Clinical presentation and laboratory evaluation on hypoglycemic disorders in adults. In: Service FJ, ed. Hypoglycemic disorders. Boston: GK Hall, 1983:73. Service FJ, Horwitz DL, Rubenstein AH, et al. C-peptide suppression test for insulinoma. J Lab Clin Med 1977; 90:180. Ram Z, Shawker TH, Bradford, MH, et al. Intraoperative ultrasound-directed resection of pituitary tumors. J Neurosurg 1995; 83:225. Rasbach DA, van Heerden JA, Telander RL, et al. Surgical management of hyperinsulinism in the multiple endocrine neoplasia, type 1 syndrome. Arch Surg 1985; 120:584. Demeure MJ, Klonoff DC, Karam JH, et al. Insulinomas associated with multiple endocrine neoplasia type 1: the need for a different surgical approach. Surgery 1991; 110:998. Thompson NW, Lloyd RV, Nishiyama RH, et al. MEN1 pancreas: a histological and immunohistochemical study. World J Surg 1984; 8:561. Osei K, O'Dorisio TM. The effects of a potent somatostatin analogue (SMS201-995) on serum glucose and gastro-entero-pancreatic hormones in a malignant insulinoma patient. Ann Intern Med 1985; 103:223. Croughs RJM, Hulsmans HAM, Israel DE, et al. Glucagonoma as part of the polyglandular adenoma syndrome. Am J Med 1972; 52:690. Marx SJ, Spiegel AM, Brown EM, Aurbach GD. Family studies in patients with primary parathyroid hyperplasia. Am J Med 1977; 62:698. Levine JH, Sagel J, Rosebrock G, et al. Prolactin-secreting adenoma as part of the multiple endocrine neoplasia-type 1 (MEN 1) syndrome. Cancer 1979; 43:2492. Veldhuis JD, Green JE III, Kovacs E. Prolactin-secreting pituitary adenomas: association with multiple endocrine neoplasia, type I. Am J Med 1979; 67:830. Ramsay JA, Kovacs K, Asa SL. Reversible sellar enlargement due to growth hormone-releasing hormone production by pancreatic endocrine tumors in an acromegalic patient with multiple endocrine neoplasia type 1 syndrome. Cancer 1988; 62:445. Brandl ML, Zimering MB, Marx SJ, et al. Multiple endocrine neoplasia type 1: role of a circulating growth factor in parathyroid cell hyperplasia. In: Kleerekoper M, Krane SM, eds. Clinical disorders of bone and mineral metabolism. New York: Mary Ann Liebert, 1988:323. van Heerden JA, Kent RB, Sizemore GW, et al. Primary hyperparathyroidism in patients with multiple endocrine neoplasia syndromes. Arch Surg 1983; 118:533.

65a. Wimer KK, Yanouski JA, Cutler GB Jr. Synthetic human parathyroid hormone 1–34 or calcitriol and calcium in the treatment of hypoparathyroidism. JAMA 1996; 276:631. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97.

Williams ED, Celestin LR. The association of bronchial carcinoid and pluriglandular adenomatosis. Thorax 1962; 17:120. Skogseld B, Larsson C, Lindgren P-G, et al. Clinical and genetic features of adrenocortical lesions in multiple endocrine neoplasia type 1. J Clin Endocrinol Metab 1992; 75:76. Marshall AHE, Sloper JC. Pluriglandular adenomatosis of the pituitary, parathyroid and pancreatic islet cells associated with lipomatosis. J Pathol Bacteriol 1954; 68:225. Marx SJ. Multiple endocrine neoplasia type 1. In: Vogelstein B, Kinzler KW, eds. The genetic basis of human cancer. New York: McGraw-Hill, 1998:489. Wilson SD. Wermer's syndrome: multiple endocrine adenopathy, type I. In: Friesen SR, Bolinger RE, eds. Surgical endocrinology: clinical syndromes. Philadelphia: JB Lippincott Co, 1978:265. Saad MF, Ordonez NG, Rashid RK, et al. Medullary carcinoma of the thyroid: a study of the clinical features and prognostic factors in 161 patients. Medicine (Baltimore) 1984; 63:319. Wells SA Jr, Dilley WG, Farndon JA, et al. Early diagnosis and treatment of medullary thyroid carcinoma. Arch Intern Med 1985; 145:1248. Nunziata V, Giannattasio R, Di Giovanni G, et al. Hereditary localized pruritus in affected members of a kindred with multiple endocrine neoplasia type 2A (Sipple's syndrome). Clin Endocrinol (Oxf) 1989; 30:57. Gagel RF, Levy ML, Donovan DT, et al. Multiple endocrine neoplasia type 2a associated with cutaneous lichen amyloidosis. Ann Intern Med 1989; 111:802. Verdy M, Weber AM, Roy CC, et al. Hirschsprung's disease in a family with multiple endocrine neoplasia type 2. J Pediatr Gastroenterol Nutr 1982; 1:603. Wagenmann A. Multiple Neurome des Auges und der Zunge. Zusammen-kunft Dtsch Ophthalmol Ges 1922; 43:282. Carney JA, Sizemore GW, Hayles AB. Multiple endocrine neoplasia, type 2b. Pathobiol Ann 1978; 8:105. Dyck PJ, Carney JA, Sizemore GW, et al. Multiple endocrine neoplasia, type 2b: phenotype recognition, neurologic features and their pathologic basis. Ann Neurol 1979; 6:302. Sizemore GW, Carney JA, Gharib H, Capen CC. Multiple endocrine neoplasia, type 2B: eighteen-year follow-up of a four-generation family. Henry Ford Hosp Med J 1992; 40:236. Khairi MRA, Dexter RN, Dubzynski NJ, Johnston CC Jr. Mucosal neuroma, phytochromocytoma and medullary thyroid carcinoma: multiple endocrine neoplasia, type 3. Medicine 1975; 54:89. Durbec PL, Larsson-Blomberg LB, Schuchgardt A, et al. Common origin and developmental dependence on C-RET of subsets of enteric and sympathetic neuroblasts. Development 1996; 122:349. Melvin KEW, Miller HH, Tashjian AH Jr. Early diagnosis of medullary carcinoma of the thyroid gland by means of calcitonin assay. N Engl J Med 1971; 285:1115. Sizemore GW. Medullary carcinoma of the thyroid gland and the multiple endocrine neoplasia type 2 syndrome. In: Spittel JA Jr, ed. Clinical medicine, vol 8. Philadelphia: Harper & Row, 1982:1. Hennessy JF, Gray TK, Cooper CW, Ontjes DA. Stimulation of thyrocalcitonin secretion by pentagastrin and calcium in two patients with medullary carcinoma of the thyroid. J Clin Endocrinol Metab 1973; 36:200. Wells SA Jr, Dilley WG, Farndon JA, et al. Early diagnosis and treatment of medullary thyroid carcinoma. Arch Intern Med 1985; 145:1248. McLean GW, Rabin D, Moore L, et al. Evaluation of provocative tests in suspected medullary carcinoma of the thyroid: heterogeneity of calcitonin responses to calcium and pentagastrin. Metabolism 1984; 33:790. Gharib H, Kao PC, Heath H III. Determination of silica-purified plasma calcitonin for the detection and management of medullary thyroid carcinoma: comparison of two provocative tests. Mayo Clin Proc 1987; 62:373. Eng C, Clayton D, Schuffenecker I, et al. The relationship between specific RET proto-oncogene mutations and disease phenotype in multiple endocrine neoplasia type 2. JAMA 1996; 276:1575. Easton DF, Ponder MA, Cummings T, et al. The clinical and screening age-at-onset distribution for the MEN 2 syndrome. Am J Hum Genet 1989; 44:208. Heshmati HM, Gharib H, Khosla S, et al. Genetic testing in medullary thyroid carcinoma syndromes: mutation types and clinical significance. Mayo Clin Proc 1997; 72:430. Lips CJM, Landsvater RM, Hoppener WM, et al. Clinical screening as compared with DNA analysis in families with multiple endocrine neoplasia type 2A. N Engl J Med 1994; 331:828. LiVolsi VA. C cell hyperplasia/neoplasia. (Editorial). J Clin Endocrinol Metab 1997; 82:39. Guyetant S, Rousselet M-C, Durigon M, et al. Sex-related C cell hyperplasia in the normal human thyroid: a quantitative autopsy study. J Clin Endocrinol Metab 1997; 82:42. Biddinger PW, Brennan MF, Rosen PP. Symptomatic C cell hyperplasia associated with chronic lymphocytic thyroiditis. Am J Surg Pathol 1991; 15:599. Neumann HPH, Berger DP, Sigmund G, et al. Pheochromocytomas, multiple endocrine neoplasia type 2, and von Hippel-Lindau disease. N Engl J Med 1993; 329:1531. Pacini F, Fontanelli M, Fugazzola L, et al. Routine measurement of serum calcitonin in nodular thyroid disease allows the preoperative diagnosis of unsuspected sporadic medullary thyroid carcinoma. J Clin Endocrinol Metab 1994; 78:826. Dunn J. When is a thyroid nodule a sporadic medullary carcinoma? (Editorial). J Clin Endocrinol Metab 1994; 78:624.

98. Uwaifo G, Remaley AT, Reynolds JC, et al. False hypercalcitoninemia in a patient with solitary thyroid nodule: should plasma calcitonin (CT) be used as a screening test for all thyroid nodules? Abstract 2454. The Endocrine Society 82nd Annual Meeting, Toronto, Canada: June 21–24, 2000. 98a. Takano T, Miyauchi A, Matsuzuka F, et al. Preoperative diagnosis of medullary thyroid carcinoma by RT-PCR using RNA extracted from leftover cells within a needle used for fine-needle aspiration biopsy. J Clin Endocrinol Metab 1999; 84:951. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142.

Anderson RJ, Sizemore GW, Wahner HW, Carney JA. Thyroid scintigram in familial medullary carcinoma of the thyroid gland. Clin Nucl Med 1978; 3:147. Russell CF, van Heerden JA, Sizemore GW, et al. The surgical management of medullary thyroid carcinoma. Ann Surg 1983; 197:42. Miller HH, Melvin KEW, Gibson Tashjian AH Jr. Surgical approach to early familial medullary carcinoma of the thyroid gland. Am J Surg 1972; 123:438. Brumsen C, Harm RH, Goslings BM, van de Velde CJH. Should patients with medullary thyroid carcinoma undergo extensive lymph node (re)operation to improve long-term survival? Henry Ford Hosp Med J 1992; 40:271. Bartlett RC, Myall RW, Mandelstam P. A neuropolyendocrine system. Mucosal neuromas, pheochromocytoma and medullary thyroid carcinoma. Oral Surg Oral Med Oral Pathol 1971; 31:206. Leape LL, Miller HH, Graze K, et al. Total thyroidectomy for occult familial medullary carcinoma of the thyroid in children. J Pediatr Surg 1976; 11:831. Telander RL, Zimmerman D, Sizemore GW, et al. Medullary carcinoma in children: results of early detection and surgery. Arch Surg 1989; 124:841. Decker RA, Toyama WM, O'Neal LW, et al. Evaluation of children with multiple endocrine neoplasia type IIb following thyroidectomy. J Pediatr Surg 1990; 25:939. Raue F, Kotzerke J, Reinwein D, et al, and the German Medullary Thyroid Carcinoma Study Group. Prognostic factors in medullary thyroid carcinoma: evaluation of 741 patients from the German medullary thyroid carcinoma register. Clin Invest 1993; 71:7. Gharib H, McConahey WM, Tiegs RD, et al. Medullary thyroid carcinoma: clinicopathologic features and long-term follow-up of 65 patients treated during 1946 through 1970. Mayo Clin Proc 1992; 67:934. Heshmati HM, Gharib H, Sizemore GW. Advances and controversies in the diagnosis and management of medullary thyroid carcinoma. Am J Med 1997; 103:60. Telenius-Berg M, Berg B, Hamberger B, et al. Impact of screening on prognosis in the multiple endocrine type 2 syndromes: natural history and treatment results in 105 patients. Henry Ford Hosp Med J 1984; 32:225. Gagel RF, Tashjian AH Jr, Cummings T, et al. The clinical outcome of prospective screening for multiple endocrine neoplasia type 2a: an 18-year experience. N Engl J Med 1988; 318:478. Saad MF, Ordonez NG, Guido JJ, Samaan NA. The prognostic value of calcitonin immunostaining in medullary carcinoma of the thyroid. J Clin Endocrinol Metab 1984; 59:850. Bäckdahl M, Tallroth E, Auer G, et al. Prognostic value of nuclear DNA content in medullary thyroid carcinoma. World J Surg 1985; 9:980. Miyauchi A, Onishi T, Morimoto S, et al. Relation of doubling time of plasma calcitonin levels to prognosis and recurrence of medullary thyroid carcinoma. Ann Surg 1984; 199:461. Miyauchi A, Endo K, Ohta H, et al. 99m Tc(V)-dimercaptosuccinic acid scintigraphy for medullary thyroid carcinoma. World J Surg 1986; 10:640. Clarke SEM, Lazarus CR, Wraight P, et al. Pentavalent [99m Tc]DMSA, [131 I]MIGB, and [99m Tc]MDP: an evaluation of three imaging techniques in patients with medullary carcinoma of the thyroid. J Nucl Med 1988; 29:33. Tisell L-E, Hansson G, Jansson S, Salander H. Reoperation in the treatment of asymptomatic metastasizing medullary thyroid carcinoma. Surgery 1986; 99:60. Moley JF, Dilley W, DeBenedetti MK. Improved results of cervical reoperation for medullary thyroid carcinoma. Ann Surg 1997; 225:734. Buhr HJ, Kallinowski F, Raue F, et al. Microsurgical neck dissection for occult metastasizing medullary thyroid carcinoma; three year results. Cancer 1993; 72:3685. Saad MF, Guido JJ, Samaan NA. Radioactive iodine in the treatment of med-ullary carcinoma of the thyroid. J Clin Endocrinol Metab 1983; 57:124. Tubiana M. External radiotherapy and radioactive iodine in the treatment of thyroid cancer. World J Surg 1981; 5:75. Simpson WJ, Palmer JA, Rosen IB, Mustard RA. Management of medullary carcinoma of the thyroid. Am J Surg 1982; 144:420. Rougier P, Parmentier C, Laplanche A, Calmettes C. Medullary thyroid carcinoma: prognostic factors and treatment. Int J Radiat Oncol Biol Phys 1983; 9:161. Howe JR, Norton JA, Wells SA Jr. Prevalence of pheochromocytoma and hyperparathyroidism in multiple endocrine neoplasia type 2A: results of long-term follow-up. Surgery 1993; 114:1070. Lips CJM, Veer J, Struyvenberg A, et al. Bilateral occurrence of pheochro-mocytoma in patients with the multiple endocrine neoplasia syndrome type 2a (Sipple's syndrome). Am J Med 1981; 70:1051. Moraca-Kvapilova L, Op de Coul AA, Merkus JM. Cerebral haemorrhage in a pregnant woman with a multiple endocrine neoplasia syndrome (type 2A or Sipple's syndrome). Eur J Obstet Gynecol Reprod Biol 1985; 20:257. Gagel RS, Melvin KEW, Tashjian AH Jr, et al. Natural history of the familial medullary thyroid carcinoma-pheochromocytoma syndrome and the identification of the preneoplastic stages by screening studies: a 5-year report. Trans Assoc Am Physicians 1975; 88:177. Sisson JC, Shapiro B, Bierwaltes WH. Scintigraphy with I-131 MIBG as an aid to the treatment of pheochromocytomas in patients with the multiple endocrine neoplasia type 2 syndromes. Henry Ford Hosp Med J 1984; 32:254. Valk TW, Frager MS, Gross MD, et al. Spectrum of pheochromocytoma in multiple endocrine neoplasia: a scintigraphic portrayal using 131 I metaiodo-benzyl-guanidine. Ann Intern Med 1981; 94:762. Sasaki Y, Klubo A, Kusakabe A, et al. The assessment of clinical usefulness of 131 I-MIBG scintigraphy for localization of tumors of sympathetic and adrenomedullary origin—a report of multicenter phase III clinical trials. Kaku Igaku 1992; 9:1083. van Heerden JA, Sizemore GW, Carney JA, et al. Surgical management of the adrenal glands in the multiple endocrine neoplasia 2 syndrome. World J Surg 1984; 8:612. O'Riordain DS, O'Brien T, Grant CS, et al. Surgical management of primary hyperparathyroidism in multiple endocrine neoplasia types 1 and 2. Surgery 1993; 114:1037. Traill Z, Tuson J, Woodham C. Adrenal carcinoma in a patient with Gardner's syndrome: imaging findings. AJR Am J Roentgenol 1995; 165:1460. Hizawa K, Iida M, Hoyagi K, et al. Thyroid neoplasia and familial adenomatous polyposis/Gardner syndrome. J Gastroenterol 1997; 32:196. Perrier ND, van Heerden JA, Goellner JR, et al. Thyroid cancer in patients with familial adenomatous polyposis. World J Surg 1998; 22:738. Zung A, Shoham Z, Open M, et al. Sertoli cell tumor causing precocious puberty in a girl with Peutz-Jeghers syndrome. Gynecol Oncol 1998; 70:421. Hertl MC, Wiebel J, Schafer H, et al. Feminizing Sertoli cell tumors associated with Peutz-Jeghers syndrome: an increasingly recognized cause of pre-pubertal gynecomastia. Plast Reconstr Surg 1998; 102:1151. Stratakis CA, Kirschner LS, Taymans SE, et al. Carney complex, Peutz-Jeghers syndrome, Cowden disease, and Bannayan-Zonana syndrome share cutaneous and endocrine manifestations, but not genetic loci. J Clin Endocrinol Metab 1998; 83:2972. Stratakis CA, Sarlis N, Kirschner LS, et al. Paradoxical response to dexa-methasone in the diagnosis of primary pigmented nodular adrenocortical disease. Ann Intern Med 1999; 131:585. Casey M, Mah C, Merliss AD, et al. Identification of a novel genetic locus for familial cardiac myxomas and Carney complex. Circulation 1998; 98:2560. Kirschner LS, Taymans SE, Stratakis CA. Characterization of the adrenal gland pathology of Carney complex, and molecular genetics of the disease. Endocr Res 1998; 24:863. Stratakis CA, Kirschner LS, Carney JA. Carney complex: diagnosis and management of the complex of spotty skin pigmentation, myxomas, endocrine overactivity, and schwannomas. Am J Med Genet 1998; 80:183.

CHAPTER 189 HERITABLE DISORDERS OF COLLAGEN AND FIBRILLIN Principles and Practice of Endocrinology and Metabolism

CHAPTER 189 HERITABLE DISORDERS OF COLLAGEN AND FIBRILLIN PETER H. BYERS What is Collagen, Where is it Made, and What Does it do? Osteogenesis Imperfecta Osteogenesis Imperfecta Type I Osteogenesis Imperfecta Type II Osteogenesis Imperfecta Type III Osteogenesis Imperfecta Type IV Ehlers-Danlos Syndrome Mild to Moderate Forms of Ehlers-Danlos Syndrome Ehlers-Danlos Syndrome Type IV Ehlers-Danlos Syndrome Type VI Ehlers-Danlos Syndrome Type VII Disorders Related to Other Collagen Genes Chondrodysplasias Disorders of Type II, Type IX, Type X, and Type XI Collagen Genes Alport Syndrome Dystrophic Forms of Epidermolysis Bullosa: Defects in Type VII Collagen Marfanoid Syndromes and Fibrillinopathies Marfan Syndrome Congenital Arachnodactyly Mitral Valve Prolapse Syndrome Isolated Ectopia Lentis Chapter References

Collagens are ubiquitously distributed, are present in large amounts, and have important structural and functional properties, so that disorders of their structure and biosynthesis often have generalized phenotypic effects and may be fatal. Fibrillins, although less abundant, also are ubiquitously distributed; consequently, mutations in the genes that encode them may also be generalized in their effects. This chapter discusses the collagen and fibrillin families of proteins, the genes that code for them, the intracellular and extracellular processing of the proteins, and the mutations in the collagen and fibrillin genes and in the genes for the processing enzymes that lead to identifiable inherited disorders of connective tissue. More than 19 collagen types, encoded by more than 30 genes, and two fibrillins have been identified. Mutations in several of these genes have been characterized. They give rise to characteristic phenotypes, depending on the gene in which they occur and the nature and location of the mutation. Table 189-1 lists the collagen genes and their disorders. Because physicians seek information about named disorders and disease processes, this chapter is organized around those named heritable disorders of connective tissue in which molecular defects have been identified.1,2,3,4,5,6,7 and 8

TABLE 189-1. Collagen Genes, Their Locations, and the Disorders That Result from Their Mutations

WHAT IS COLLAGEN, WHERE IS IT MADE, AND WHAT DOES IT DO? The collagens are a family of more than 19 proteins, encoded by more than 30 genes, which have a similar core amino-acid sequence, contain three chains, form triple-helical structures, and contain the two characteristic amino acids, hydroxyproline and hydroxylysine (see Table 189-1). Some collagen molecules, such as type I collagen, are heteropolymers, containing more than one type of chain, whereas others, such as types II and III, are homopolymers and contain three identical chains. Each collagen contains a long region, stretching from 600 to more than 1000 amino acids, characterized by the repeating Gly-X-Y unit, in which X and Y often are proline and hydroxyproline, respectively. In some collagens (the fibrillar collagens), this triplet domain is uninterrupted, but in most, several regions are found in which glycine is not in the third position. Each collagen has a characteristic tissue distribution, and mutations that affect the structure of one chain lead to phenotypes that reflect that distribution. For example, mutations in the chains of type I collagen frequently affect the skin, vessels, and bone; those in type II collagen alter the structure of cartilage; those in type III collagen affect skin, vessels, and bowel; and those in type VII collagen alter dermal-epidermal adhesion. Collagens play many roles in tissues, depending partly on the nature of the protein. For example, type I collagen provides tensile strength in skin, tendons, and fascia; is the chief structural component of many organ capsules; and, as the principal component of the cornea, facilitates light transmission. Type IV collagen, which is the primary component of basement membranes, separates epithelial cell layers from mesenchymal cells, an important function during tissue development. Genes for most of the collagens have been characterized, and the chromosomal locations have been determined. The prototype genes—those for the chains of type I collagen, the most abundant collagen in the vertebrate body—provide models for gene structure, from which considerable divergence is seen. The COL1A1 (a1[I]) and COL1A2 (a2[I]) genes encode proteins that contain just over 1400 amino-acid residues. Although this would require ~5000 nucleotides of coding sequence, the coding sequences are interrupted by more than 50 intervening sequences, so that the COL1A1 gene is 18 kilo–base pairs (kbp) in length, whereas the COL1A2 gene is 38 kbp. The organization of the coding domains in each is similar: The long triple-helical domain of the protein (1014 amino acids in each) is coded for by 45 different coding regions (exons) interrupted by intervening or noncoding sequences (introns). A remarkable conservation of the intronexon structure is found among several collagens, although the intron size and sequences differ widely from gene to gene. For example, almost the entire triple helical domain of the type X collagen gene (COL10A1) is contained in a single exon. In contrast, the COL7A1 gene contains more than 100 exons, yet is contained within only 30 kbp of the genome. The genes for the collagens are transcribed in the nucleus to form full-length transcripts from which the intervening sequences are removed to produce a mature mRNA molecule. Transcriptional efficiency may be controlled partly by various hormonal agents, including glucocorticoids, parathyroid hormone, and others, which may act in both a tissue- and collagen-specific manner. Along with transcriptional regulation, mechanisms are found that regulate collagen production at the level of translation of collagen mRNA, the rate of chain assembly to form molecules, and the rate of intracellular degradation of collagen molecules. The mature mRNA is transported to the cytoplasm and translated on membrane-bound ribosomes of the rough endoplasmic reticulum (ER) (Fig. 189-1). The nascent polypeptide chain (a preproa chain) is directed to the lumen of the rough ER by the presence of signal sequences at the amino terminus. During translation and passage through the membrane, most of the prolyl residues in the Y position are hydroxylated by the enzyme prolyl 4-hydroxylase; a small number of prolyl residues in the X position are hydroxylated by a different hydroxylase; many Y-position lysyl residues are hydroxylated by lysyl hydroxylase; some of the hydroxylysyl residues are glycosylated, to form glucosyl-galactosyl-hydroxylysine or the monosaccharide galactosyl-hydroxylysine; and one or more N-linked oligosaccharides are added to the precursor-specific polypeptides of most proa chains. The extent of lysyl hydroxy lation differs widely among collagen types. In types I and III col lagen, ~15% of the residues are hydroxylated; in type II collagen,~50% are hydroxylated; and in type IV collagen, almost all of the lysyl residues in the Y position are hydroxylated.

FIGURE 189-1. Schematic representation of the biosynthetic pathway of type I procollagen. The reactions depicted here can be generalized to all collagens. (nRNA, nuclear RNA; mRNA, messenger RNA; RER, rough endoplasmic reticulum.) (Modified from Byers PH, Barsh GS, Holbrook KA. Molecular pathology in inherited disorders of collagen metabolism. Hum Pathol 1982; 13:89.)

Prolyl hydroxylation is a vital step in the formation of a stable collagen molecule, because its extent determines the melting point of the triple helix of the molecule: In the absence of any prolyl hydroxylation, the melting point of type I procollagen is ~27°C, whereas with complete hydroxylation, it is 42°C. The function of lysyl hydroxylation is less clear, although it appears to be important in the formation of stable intermolecular cross-links between collagen molecules in the extracellular matrix. The completed proa chains of type I collagen are ~1400 amino acids in length (Fig. 189-2). Each chain has a propeptide-specific amino-terminal extension, which contains a globular domain and a collagen-like sequence; an amino-terminal telopeptide domain, which contains a peptidase cleavage site; a long central triple-helical domain, which is the Gly-X-Y triplet portion; a carboxyl-terminal telopeptide domain, which contains another peptidase cleavage site; and a carboxy-terminal extension of ~250 amino-acid residues, which contains a domain that governs chain selection during molecular assembly. The completed chains of type I procollagen are discharged into the lumen of the rough ER; the three chains of the molecule are assembled, stabilized by interchain disulfide bonds; and the triple helix is propagated toward the amino-terminal end of the molecule. The propagation of the triple-helical structure depends on the presence of glycine in every third position, and the structure is stabilized by hydrogen bonds that form between glycyl and hydroxyprolyl residues. The procollagen molecule is transported to the Golgi apparatus, where the N-linked oligosaccharide is converted to a high-mannose structure, and the mature molecule then is secreted into the extracellular space. After secretion, the amino-terminal and carboxy-terminal propeptide extensions are cleaved, each by a specific protease, and the resultant collagen molecule then aggregates with other molecules to form collagen fibrils.

FIGURE 189-2. Schematic representation of type I procollagen, indicating different domains of the molecule. (Redrawn from Prockop DJ, Kivirikko KI, Tuderman L, Guzman NA. The biosynthesis of collagen and its disorders [first of two parts]. N Engl J Med 1979; 301:13.)

The extent of processing differs among collagen types. The fibril structure is stabilized by the formation of intermolecular cross-links. These cross-links are formed after the oxidative deamination of lysyl residues in the amino-terminal and carboxyl-terminal telopeptide domains. These active aldehydes spontaneously condense with adjacent lysyl or hydroxylysyl residues and ultimately, in many molecules, go on to form stable cross-links derived from three lysyl or modified lysyl residues. Although collagens form highly stable structures, a constant turnover occurs, which is mediated through extracellular matrix metalloproteinases, including collagenases—enzymes capable of hydrolyzing intact collagen molecules. The stability of a tissue is a fine balance between the biosynthetic activities and degradation of the connective tissue macromolecules. In some heritable disorders of collagen biosynthesis, this balance is lost.

OSTEOGENESIS IMPERFECTA Osteogenesis imperfecta (OI) is a group of inherited generalized connective tissue disorders of which the principal manifestation is bone fragility9,8,9,10 and 11 (see Chap. 63). On the basis of the clinical findings, the mode of inheritance, and the radiologic picture, several types of OI have been described (Table 189-2). Determining which type of OI a patient has is important, because the natural history and recurrence risks in subsequent offspring differ. The discrimination of the different types of OI depends on the age of initial presentation, the family history, the presence or absence of bone deformity, and the radiographic appearance of the bones. Biochemical testing can be used to confirm the clinical impressions. Most of the forms of OI studied so far have resulted from mutations in the genes for the chains of type I collagen.11 Because these molecules are synthesized by dermal fibroblasts in culture, biopsy of a small piece of skin (a 2-mm circle usually is sufficient), culture of dermal fibroblasts, and evaluation of collagen biosynthesis and structure constitute one test to determine the type of OI in a given patient; an alternative is the determination of the sequences of both type I collagen genes. The phenotypic differences reflect the nature and location of the mutation in the type I collagen genes. The following classification is most commonly used but, for a condition that varies almost continuously from lethal to mild, any attempt to create categories faces difficulty, and many people do not fit into the molds.

TABLE 189-2. Classification, Clinical Characteristics, Mode of Inheritance, Biochemical Findings, and Ultrastructural Findings in Three Heritable Categories of Collagen Disorders

OSTEOGENESIS IMPERFECTA TYPE I OI type I is inherited in an autosomal dominant manner. Typically, affected individuals have blue sclerae and normal stature. They rarely have dentinogenesis imperfecta and may experience from a few to more than 50 fractures (usually of the long bones) before puberty. The fracture frequency usually decreases about the time of puberty, suggesting that some hormonal change improves bone strength. The fractures heal without deformity. The frequency of this form of OI has been

estimated at ~1 in 15,000, but because of the relatively mild presentation, it may go unidentified and thus may be more frequent. The diagnosis of OI type I usually is suspected on the basis of a dominant family history, the observation of blue sclerae in the patient, and bone fractures. It is confirmed by measuring the production of type I procollagen by dermal fibroblasts in culture. Ordinarily, ~85% of the collagen synthesized by these cells is type I procollagen, and most of the remainder is type III procollagen. Cells from patients with OI type I synthesize approximately half the normal amount of type I procollagen but a normal amount of other proteins. The decrease in type I procollagen production is caused by decreased synthesis of proa(I) chains that results from a decrease in the steady-state levels of cytoplasmic COL1A1 mRNA.12,13 and 14 Many different mutations in the COL1A1 gene can cause the same phenotype. Most are premature chain terminations that result from point mutations, frame shift mutations, 15,16 and 17 or splice-site mutations that lead to frame shifts. Premature termination codons dramatically destabilize the mRNA that results from the mutant allele, so little is apparent in the cell. No adequate treatment exists for OI type I that decreases the frequency of bone fracture or increases bone density. Agents that increase the production of type I collagen have therapeutic potential, but none has been effective in controlled tests. Treatment with bisphosphonates that decrease bone resorption is now being studied. The prenatal identification of affected fetuses can be accomplished by molecular diagnosis through either mutation detection or linkage. OSTEOGENESIS IMPERFECTA TYPE II The perinatal lethal form of OI, OI type II, affects ~1 in 40,000 infants and is the most severe form of bone disease.10,11,18 Affected infants have remarkably crumpled bones, a paucity of calvarial mineralization, and small thoracic cavities (Fig. 189-3). Death usually results from respiratory failure and frequently occurs during the first few hours after birth. More than 75% of these infants die during the first month, and none is known to have survived longer than a year. With the increasing use of early gestational ultrasonography, the disorder is being detected in a significant number of affected infants during the second trimester of pregnancy.

FIGURE 189-3. Osteogenesis imperfecta (OI) type II, the perinatal lethal form. Left, Infant who died in newborn period. Notice the beaked nose, the short, deformed limbs, and the position of legs. Right, Radiograph of infant with OI type II. Virtually no mineralization of calvarial bones is seen, the ribs are irregular, the bones of the limbs are all foreshortened, and the femurs appear telescoped. (From Hollister DW, Byers PH, Holbrook KA. Genetic disorders of collagen metabolism. Adv Hum Genet 1982; 12:1.)

Although OI type II was originally thought to be an autosomal recessive disorder,10,18 biochemical and molecular genetic studies indicate that in most infants this disorder is the result of new dominant mutations.19 Rarely, these mutations are large insertions or deletions (of as many as 3.5 kbp) in either the COL1A1 or COL1A2 gene. Usually they are single nucleotide changes that cause substitutions for glycyl residues within the triple-helical domain in either of the chains.11,20,21,22,23,24,25,26,27 and 28 These mutations decrease the thermal stability of the molecules into which the abnormal chains are incorporated, decrease the efficiency of secretion of those molecules, and decrease the amount of normal procollagen secreted. This form of OI must be differentiated from some of the other lethal skeletal dysplasias and from hypophosphatasia. An experienced pediatric radiologist can help with this distinction, and the diagnosis can be confirmed by examination of the collagens synthesized by cultured fibroblastic cells. Although OI type II usually results from a new dominant mutation, the observed recurrence risk is 2% to 6%. In almost all families, recurrence results from parental mosaicism for the mutation.29 An autosomal recessive form also may exist but is rare. Supportive care is appropriate, because no therapy alters the prognosis for these infants. Prenatal diagnosis should be offered in subsequent pregnancies in families to which an infant with OI type II has been born. Ultrasonographic examination of the fetus between 15 and 18 weeks of gestation reliably identifies the affected fetus. Prenatal diagnosis also can be performed by examination of the collagens synthesized by cells grown from chorionic villus biopsy specimens and by direct DNA analysis if the mutation is known. OSTEOGENESIS IMPERFECTA TYPE III The progressive deforming variety of OI, OI type III, usually is recognized at birth because of slightly shorter stature and deformities resulting from in utero fractures (Fig. 189-4). Rare autosomal recessive and, more commonly, autosomal dominant forms of this type of OI are found.30 Radiologically, the calvarium is undermineralized, the ribs are thin, the long bones are thin with evidence of fracture, and the skeleton is osteopenic. If no fractures are present at birth, they usually occur during the first year of life, when deformity becomes apparent. Angulation deformities of the tibias and the femurs reduce the efficiency of weight-bearing and increase the likelihood of fracture.

FIGURE 189-4. Osteogenesis imperfecta (OI) type III, the progressive deforming variety. Left, Marked bowing of lower legs in young man. Right, Radiograph of upper extremity in the same person. (From Hollister DW, Byers PH, Holbrook KA. Genetic disorders of collagen metabolism. Adv Hum Genet 1982; 12:1.)

Treatment is directed toward providing a more functional anatomy, and in this group of children placement of intramedullary rods in long bones appears to improve the prognosis and may facilitate walking. For some children, bone fragility makes independent ambulation difficult in all but the most restricted circumstances, and motorized wheelchairs provide the most mobility. The growth of these children is limited, and adult height between 1 and 1.5 meters is common. Because of the bone fragility and ligamentous laxity, many of these children develop significant kyphoscoliosis and eventually experience pulmonary insufficiency. Sclerae often are pale blue at birth and become nearly normal by puberty. Dentinogenesis imperfecta is common, but hearing loss is rare. No medical therapy reliably decreases fracture frequency or increases growth and bone density. Observational studies of bisphosphonate use have suggested that they may increase bone density and perhaps reduce fracture risk, but more extensive study is required. Although autosomal recessive31,32 and autosomal dominant forms of OI type III have been identified, most affected individuals have dominant mutations in the type I

collagen genes. Prenatal diagnosis is available by ultrasonography and by analysis of chorionic villus biopsy cells for abnormal proteins or the known mutations. OSTEOGENESIS IMPERFECTA TYPE IV OI type IV is characterized by normal or grayish sclerae, mild to moderate deformity with some short stature, and autosomal dominant inheritance (Fig. 189-5). Some individuals have dentinogenesis imperfecta, but hearing loss seems rare. Infants often have mild shortness of stature and femoral bowing at birth, but in some, the diagnosis is delayed until the first fracture. As with the other forms of OI, fracture frequency decreases at the time of puberty but may increase at older ages, especially in women. Because of the deformities, some individuals benefit from insertion of rods during childhood, but medical therapies do not appear to modify the relatively mild natural history of this form of OI. Again, bisphosphonate treatment is being studied.

FIGURE 189-5. Osteogenesis imperfecta type IV. Left, A 13-month-old boy. Right, Radiograph showing short bowed femurs and deformities of his humeral and tibial bones. (From Wenstrup RJ, Hunter AGW, Byers PH. Osteogenesis imperfecta type IV: evidence of abnormal triple helical structure of type I collagen. Hum Genet 1986; 74:47.)

Linkage studies indicate that in some families, this phenotype results from mutations in the a2(I) gene.33 Biochemical studies of type I collagen synthesized by cells from these patients confirm the linkage data and suggest that small deletions or single aminoacid substitutions for glycyl residues in the triple helical domains of the a1(I) and a2(I) chain cause this phenotype.11,34

EHLERS-DANLOS SYNDROME The Ehlers-Danlos syndrome (EDS) is a heterogeneous group of generalized connective tissue disorders in which the principal manifestations are skin fragility, skin hyperextensibility, and joint hypermobility4,35,36,37 and 38 (see Table 189-2). Genetic and biochemical studies had expanded the known types of EDS, but another classification has reduced some types into single categories.39 EDS types I and II are known as the classic form; EDS type III is known as the hypermobile form; EDS type IV is termed the vascular or ecchymotic form; EDS type VI is known as the ocular-scoliotic form; and EDS type VII has been divided into the arthrochalasis form and dermatosparaxis. EDS type IX has been removed from the classification and renamed the occipital horn syndrome, a disorder that results from mutations in the Menkes syndrome gene, ATP7A. Distinguishing the different types is important, because in some forms (e.g., EDS type IV and EDS type VI) complications of arterial rupture may ensue, whereas for most of the other types life expectancy usually is normal and complications are modest. Some of the literature has not differentiated the types clearly, and, often, the complications inherent in EDS type IV are cited as characteristic of the syndrome as a whole. MILD TO MODERATE FORMS OF EHLERS-DANLOS SYNDROME Several types of EDS, including types I, II, III, V, VIII, and X, have mild to moderate phenotypes and are rarely associated with lethal complications. The clinical findings in EDS types I and II may be dramatic, with markedly soft, velvety, hyperextensible skin; impressive joint hypermobility; easy bruising; and thin, atrophic, “cigarette-paper” scars (Fig. 189-6). In both forms, varicose veins are common, and in EDS type I, affected infants may be delivered prematurely. Although tissues are more friable than normal, most surgical procedures are tolerated well. Both of these conditions are inherited in an autosomal dominant fashion, and relatively little variation is found within families. Dermal collagen fibrils have a bizarre organization when viewed by electron microscopy.36 Linkage to the COL5A140,41 and COL5A2 genes has been identified, and mutations42,43,44 and 45 in both genes have been observed in some families or individuals with EDS types I and II. A significant proportion of people with EDS type I or II probably have null mutations in the COL5A1 gene.

FIGURE 189-6. Ehlers-Danlos syndrome type I. Left, Genu recurvatum in a patient. Notice abnormal scars on both knees. Right, Marked skin hyperextensibility at elbow of a patient (bottom). A control is shown at the top. (From Hollister DW, Byers PH, Holbrook KA. Genetic disorders of collagen metabolism. Adv Hum Genet 1982; 12:1.)

In EDS type III, benign familial hypermobility, findings are limited to marked joint hypermobility. Although the skin often is soft, it is not hyperextensible, and scars are not abnormal. The principal complications are recurrent dislocation and early-onset degenerative joint disease. The condition is inherited in an autosomal dominant manner and probably is common in the general population. No biochemical defects have been identified. EDS type V is distinguished from EDS type II by the X-linked recessive inheritance pattern. It appears to be rare. No biochemical defects have been identified,37 and its existence has been disputed.38 EDS type VIII is inherited in an autosomal dominant manner. It is characterized by bruising; soft, hyperextensible skin; hyper-mobile joints; and periodontal disease. Loss of teeth by the early 20s is common. 46 No biochemical defects have been identified. EDS type X is inherited in an autosomal recessive fashion. Only one family has been identified to date.47 Joint hypermobility is mild; the chief finding is easy bruising. The disorder appears to result from an alteration in fibronectin that interferes with normal platelet aggregation. EHLERS-DANLOS SYNDROME TYPE IV EDS type IV, the ecchymotic, arterial, or Sack-Barabas variety, is the most severe form of EDS and results from mutations in the COL3A1 gene.48,49 and 50 The disorder is inherited in an autosomal dominant fashion, but many of the patients are the only affected members of their families. Affected individuals have thin, translucent skin through which the venous pattern over the trunk, abdomen, and extremities is visible; minimal joint hypermobility, which may be limited to the small joints of the hands and feet; and marked bruising (Fig. 189-7). Often, the skin over the face has a parchment-like appearance, and the nose frequently is thin and beaked. Individuals with EDS type IV are at risk for arterial rupture, spontaneous rupture of the colon, and rupture of the gravid uterus.38 Recurrent abdominal pain without significant other findings is common and may result from mural hemorrhage in the small bowel. The location of arterial hemorrhage determines the presenting symptoms in some

patients (stroke, abdominal bleeding, limb compartmental syndrome), whereas bowel rupture is the first complication seen in others.

FIGURE 189-7. Ehlers-Danlos syndrome (EDS) type IV. This 26-year-old woman has a readily visible venous pattern over her trunk and abdomen (arrows). Linear markings in antecubital spaces (arrowheads) represent areas of elastosis perforans serpiginosa, a common accompaniment to EDS type IV. (From Byers PH, Holbrook KA, McGillivray B, et al. Clinical and ultrastructural heterogeneity of type IV Ehlers-Danlos syndrome. Hum Genet 1979; 47:141.)

The life span of individuals with EDS type IV is shortened, although survival into the seventh decade can be seen.50 Causes of death include exsanguination from large-artery rupture, sepsis from bowel rupture, and shock from uterine rupture. The complications of pregnancy, along with vascular rupture and uterine rupture, include tearing of the vaginal tissues during delivery. The risk of life-threatening or lethal complications during pregnancy ranges from 8% to 15%.51 Each affected woman should be aware of these potential risks. Furthermore, because of the auto-somal dominant nature of inheritance in most families, the risk of having an affected infant is 50% for each pregnancy. Prenatal diagnosis by linkage analysis (using polymorphic restriction sites in the genes for type III collagen) is useful in some families,45 and analysis of the type III collagen synthesized and secreted by cells from chorionic villus biopsy specimens may help in others. EDS type IV results from abnormalities in the structure or expression of the genes for type III collagen.49 No significant clinical variation appears to exist among families with different classes of mutations (e.g., point mutations that alter glycine residues, exon-skipping mutations, or large deletions). The suspected diagnosis can be confirmed by examining the biosynthesis of type III procollagen by cultured dermal fibroblasts. Because of the nature of complications and the risks during surgery in this disorder, confirmation of the diagnosis is essential. No medical treatment is available that increases the production of normal type III collagen. EHLERS-DANLOS SYNDROME TYPE VI EDS type VI, the first true, heritable disorder of collagen metabolism to be described, is an autosomal recessive disorder that results from decreased activity of the enzyme lysyl hydroxylase because of mutations in the lysyl hydroxylase gene.52,53,54 and 55 Some variation is found in the clinical phenotype, but a marfanoid habitus; soft, hyperextensible skin; joint hypermobility; scoliosis; and ocular fragility are the principal clinical features. Because many of the individuals identified have been children, the natural history of the disorder is not yet well understood. At least one patient appears to have died because of arterial rupture in later life. The diagnosis is established most readily by the measurement of pyridinoline cross-links in urine.56,57 These reflect the hydroxylation status of lysyl residues in collagen degradation products. The lysyl hydroxylase gene responsible for hydroxylation of helical lysyl residues has been cloned (PLOD1), and several mutations have been identified, the most common of which is a duplication of exons 9–16.58 Collagen fibrils in skin have a bizarre branching organization, similar to that seen in EDS type I. Prenatal diagnosis has been attempted by the measurement of enzyme activity in amniotic fluid cells and has correctly predicted an unaffected but heterozygous infant55; molecular diagnosis would be more specific. Although no specific therapy has been demonstrated, the use of pharmacologic doses of ascorbic acid, a cofactor for lysyl hydroxylase, has been advocated, and an increase in urinary excretion of hydroxylysine has followed ingestion of this vitamin.54 EHLERS-DANLOS SYNDROME TYPE VII EDS type VII is divided into two distinct phenotypes: arthrochalasis multiplex congenita (EDS types VIIA and VIIB) and dermatosparaxis (EDS type VIIC). EDS types VIIA and VIIB are characterized by marked joint hypermobility, multiple joint dislocations, and congenital hip dislocation. Initially, this disorder was thought to result from abnormalities in the enzyme that cleaves the amino-terminal propeptide extension from type I procollagen,59 but restudy of some of the original patients and detailed study of the collagens synthesized by cells from several new patients have demonstrated that the mutations affect the cleavage of the NH2-terminal propeptide of the substrate proa1(I) and proa2(I) chains.60,61 and 62 No effective therapy, other than surgical repair of dislocation, is known. Six children with defects in the procollagen N-protease have now been identified. They have very fragile skin, bruise easily, and have blue sclerae and very marked joint laxity.63,64 and 65 Electron micrographs of their skin reveal very bizarre collagen fibrils. Mutations in the N-proteinase gene have been identified in all affected children.65

DISORDERS RELATED TO OTHER COLLAGEN GENES CHONDRODYSPLASIAS More than 150 varieties of chondrodysplasia, disorders of bone growth and structure, have been identified.66,67,68,69 and 70 Within this relatively large group, only a few (some forms of achondrogenesis, spondyloepiphyseal dysplasia [SED], Stickler syndrome, and Kniest syndrome) result from mutations in the COL2A1 gene that encodes the chains of type II collagen. These conditions share abnormalities of the articular cartilage and of the vitreous humor of the eye. These tissues contain collagen types II and XI; type XI collagen contains three chains, one encoded by the COL2A1 gene, one by the COL11A1 gene, and one by the COL11A2 gene. This is a minor molecule in collagen. Mutations in the COL11A1 gene cause a Stickler phenotype or a Stickler-Marshall picture, whereas those in the COL11A2 gene cause a Stickler phenotype without ocular findings. Forms of metaphyseal dysplasia result from mutations in the type IX collagen genes. Mutations in the type X collagen gene result in the Schmidtype metaphyseal dysplasia. Other forms of chondrodysplasia result from mutations in other supporting matrix genes of cartilage. DISORDERS OF TYPE II, TYPE IX, TYPE X, AND TYPE XI COLLAGEN GENES STICKLER SYNDROME Stickler syndrome, hereditary arthro-ophthalmodystrophy, is an autosomal dominant disorder characterized by early degenerative joint disease in the presence of a mild skeletal epiphyseal dysplasia, and vitreal degeneration with moderate to severe myopia and retinal detachment in some individuals.71 Linkage studies in families with Stickler syndrome indicate genetic heterogeneity.72 In most families with Stickler syndrome, mutations in the COL2A1 gene have been identified that result in stop codons.49 Heterogeneity has been found, and mutations in the COL11A1 and COL11A2 genes have been characterized. Although both cause the skeletal abnormalities, those in the COL11A2 gene usually are not accompanied by ocular findings, because the COL5A2 gene product, proa2(V), substitutes for proa2(XI) in the vitreous. SPONDYLOEPIPHYSEAL DYSPLASIA SEDs are a group of chondrodysplasias in which the radiologic features include abnormal epiphyses, flattened vertebral bodies, and ocular involvement that ranges

from myopia to vitreoretinal degeneration. A range of clinical expression is found, from mild short stature to very severe short stature, pulmonary compromise, and death in infancy. Analysis of type II collagen in articular cartilage from individuals with different forms of SED have demonstrated alterations in the amount and electrophoretic mobilities of the a1(II) chains in those tissues, compatible with mutations in the COL2A1 gene.49,73 Analysis of the COL2A1 genes from several individuals with different forms of SED have identified mutations that include partial duplication of exon 48 with the resultant addition of 15 amino acids (duplication of residues 970–984 of the triple helix),74 deletion of exon 48 from one allele,75 substitution of the glycine at position 997 in the triple helix by serine,76 a splice junction mutation that results in skipping the sequences of exon 20,77 and, surprisingly, a substitution of cysteine for arginine at position 75 in the triple helical domain.78 ACHONDROGENESIS/HYPOCHONDROGENESIS The most severe end of the spectrum of chondrodysplasias in which evidence is found of involvement of the COL2A1 gene is the achondrogenesis/hypochondrogenesis group. Infants with achondrogenesis have severe short-limbed dwarfism and die in the immediate perinatal period or in utero. Electron microscopic studies of cartilage from infants or fetuses with achondrogenesis/hypochondrogenesis have demonstrated marked dilation of the rough ER with electron-dense material and a paucity of fibrils in the extracellular matrix.79 Several point mutations in the COL2A1 gene have been identified in this disorder, resulting in substitution for glycine residues.49 Too few mutations in the COL2A1 gene have been identified to permit systematic genotype/phenotype analysis. MULTIPLE EPIPHYSEAL DYSPLASIA Multiple epiphyseal dysplasia is characterized by autosomal dominant inheritance and short stature of varying severity with primary involvement of the epiphyses outside the spine. This family of disorders has been linked to mutations in genes of the type IX collagen family, and mutations have been identified in the COL9A2 gene80 and COL9A3 gene.81 METAPHYSEAL CHONDRODYSPLASIA OF SCHMID The Schmid type of metaphyseal dysplasia is characterized by dominant inheritance, and bowing of the extremities.82 Mild to moderate short stature is apparent in childhood. A variety of mutations in the COL10A1 gene have been identified, most of which interfere with chain association.83,84,85 and 86 ALPORT SYNDROME Alport syndrome is characterized by progressive hereditary nephritis and sensorineural deafness that in some families is associated with ocular abnormalities, particularly lenticonus and macular changes.87 The condition is inherited in an X-linked manner, with males likely to develop both hearing loss and renal dysfunction earlier than females in the same family. The renal basement membrane shows a characteristic array of findings that includes splitting and thinning of the glomerular basement membrane. Linkage studies showed the Alport gene88,89 and 90 to be located at Xq13; shortly thereafter, a type IV collagen gene was identified that mapped to the same location.90 In all families with X-linked nephritides, with or without significant deafness, the disorder has been linked to the COL4A5 gene locus. The mutational spectrum in the COL4A5 gene is similar to that seen in other collagen genes. It has included multiexon deletions, point mutations within the triple-helical domain that substitute for glycine residues, and mutations within the non-collagenous carboxy-propeptide domain.91 Extensive deletions involving the 5' end of the COL4A5 gene and a segment of the adjacent COL4A6 gene can result in a form of Alport syndrome in which the renal disease is accompanied by diffuse esophageal leiomyomatosis.92 DYSTROPHIC FORMS OF EPIDERMOLYSIS BULLOSA: DEFECTS IN TYPE VII COLLAGEN Epidermolysis bullosa (EB) is a highly heterogeneous group of disorders in which blistering of the skin is the common unifying theme. Blistering occurs within the epidermis (simplex forms of EB), at the dermal-epidermal junction (junctional forms of EB), or within the dermis below the basement membrane (dystrophic forms of EB).93 DYSTROPHIC FORMS OF EPIDERMOLYSIS BULLOSA Dominantly inherited forms of dystrophic EB may be generalized in their distribution (blistering occurs early and is generalized; a later involvement of the limbs is seen, and scars forming while healing occurs), or may be more localized over the extremities with a preference for sites of trauma. Generally, this is most severe during the first 5 years of life.94 Morphologic studies of the skin have indicated abnormalities in the structure or amount of anchoring fibrils or both. These structures course from the basement membrane zone into the upper portion of the papillary dermis and appear to be involved in maintaining dermal-epidermal integrity.95 Anchoring fibrils consist largely, if not exclusively, of type VII collagen.96 With the identification of polymorphic sites within the COL7A1 gene in some families with abnormal anchoring fibrils, the disorder has been shown to be linked to the COL7A1 gene. Many mutations have been identified.97,98 RECESSIVE DYSTROPHIC FORMS OF EPIDERMOLYSIS BULLOSA The severe recessive dystrophic forms of EB are characterized by generalized blistering in the newborn period, often present at birth, with scarring and progressive syndactyly from scars of the digital skin, mucosal involvement that leads to esophageal strictures, and consequent dysphagia. Protein and fluid loss through areas of chronic erosion may lead to infection and malnutrition. Survival is limited both by infection and by the consequences of involvement of the gastrointestinal tract. Early ultrastructural studies revealed abnormalities at the dermal-epidermal junction; subsequent analysis demonstrated that this is because of an absence or markedly diminished number of anchoring fibrils.99 Studies with antibodies known to react with type VII collagen have shown that staining at the epidermal-dermal junction is markedly diminished.100 In some individuals, accumulation of type VII collagen within the basal epidermal cells reinforces the idea that the type VII collagen gene is the primary gene in which mutations produce the severe recessive dystrophic phenotype.101 Studies of 25 families with more than one child with recessive dystrophic forms of EB have failed to demonstrate a single instance in which the siblings do not share the same COL7A1 genotype, a finding consistent with the view that recessive inheritance of mutations in the COL7A1 gene is the cause of the condition.102 Many mutations have now been identified in the COL7A1 gene.99

MARFANOID SYNDROMES AND FIBRILLINOPATHIES The term fibrillinopathy, however awkward, has emerged as a means to identify those disorders that result from mutations in the fibrillin genes. Two fibrillin genes, FBN1 on chromosome 15 and FBN2 on chromosome 5, have been identified and characterized.103 These genes encode two distinct proteins that form part of the microfibrillar network in skin, vessels, and most other tissues. Fibrillin 1 (the product of the FBN1 gene) makes up the large part of the microfibrils that form the zonular fibrils of the lens—one of the clues that led to the identification of this gene as the host of mutations that result in Marfan syndrome.104 The microfibrils are thought (in some tissues) to provide the scaffold on which elastin is deposited to form elastic fibers, although clearly in many regions they are devoid of associated elastin. Fibrillin 2 appears early in vessel formation and is replaced later in development by fibrillin 1.105 The marfanoid syndromes are a heterogeneous group of connective tissue disorders, most of which are inherited in an auto-somal dominant fashion, and are characterized by abnormalities of the cardiovascular system, the skeleton, and, often, the eyes.106,107 These conditions include Marfan syndrome, contractural arachnodactyly, mitral valve prolapse syndrome, isolated ectopia lentis, and homocystinuria (see Table 183-2). Only the last is inherited in an autosomal recessive fashion. MARFAN SYNDROME The classic variety of Marfan syndrome is inherited in an auto-somal dominant fashion, although 15% to 20% of affected individuals represent new mutations.107 Most affected individuals are tall and have long arms and legs, arachnodactyly, high arched palate, pectus deformities, and scoliosis; approximately half of the patients have lens dislocation.108 Virtually all individuals develop mitral valve prolapse by puberty, and most are at risk to develop aortic dilatation, aortic insufficiency, and, ultimately, aortic dissection.109,110 Cardiac complications, including both congestive heart failure resulting from valvular dysfunction and aortic dissection, usually are the chief causes of death in this disorder. When the patient is first seen, a careful family history should be taken to determine which other family members are at risk. Height, upper and lower segment, and blood pressure in both arms should be measured, a slit-lamp examination of the dilated pupils should be performed, and an echocardiogram should be taken. All affected individuals should be followed at 1- to 2-year intervals. The echocardiogram should be repeated every other year before puberty and yearly in adult life. Yearly

eye examinations should be performed to exclude glaucoma and assess lens status. The most significant complication of Marfan syndrome is aortic dissection, which can lead to early death. Other complications include progressive scoliosis (which usually occurs during the pubertal growth spurt), excessive height, dural ectasia, and joint hypermobility. With the use of prophylactic aortic surgery to treat aortic dilatation before dissection,110 life span is probably increasing, and a new range of age-related complications may appear. Treatment for the progressive scoliosis may include bracing and insertion of rods, especially if the thoracic space is compromised. Treatment of prepubertal children with androgens (for males) or estrogens (for females) to hasten epiphyseal closure can change projected growth patterns substantially. Several lines of evidence have converged to demonstrate that the FBN1 gene harbors mutations that cause Marfan syndrome. Immunofluorescence studies of cells and tissue from people with Marfan syndrome indicated a lack of fibrillin in skin and cultured cells.111 Linkage analysis with random markers identified a region on chromosome 15 that was linked.112 Analysis of fibrillin synthesis by cells cultured from affected individuals confirmed abnormalities in synthesis and secretion.113 Finally, the complementary DNA was cloned, linkage was established,114 and mutations were identified.103 More than 100 mutations have now been identified.115 The principal controversy in the care of patients with Marfan syndrome concerns the treatment of the cardiovascular complications. Most untreated individuals with this disease die from aortic and mitral insufficiency or, more commonly, aortic dissection.106 Aneurysm formation usually precedes dissection or rupture. During the last several years, surgical technique and prosthetic materials have improved to the point that many medical centers recommend replacement of the ascending aorta and aortic valve by the time the aortic root reaches double the mean value (55–60 mm).110 Still controversial is the question of whether medical therapy has a place in the treatment of aortic dilatation. The proposal has been made that all individuals with Marfan syndrome be treated with b-adrenergic blockers to decrease cardiac contractility. The results of a single therapeutic trial have been reported. These findings116 indicate that some individuals with Marfan syndrome may benefit from treatment with b-blockers. Regardless, management of the nondilated aorta in patients with Marfan syndrome often involves treatment with b-blockers. CONGENITAL ARACHNODACTYLY Congenital contractural arachnodactyly is characterized by joint contracture, rather than joint laxity, and a characteristic crumpled upper ear. Neither lens dislocation nor aortic dilatation is a feature of this condition. Linkage of the phenotype to polymorphic variants in the FBN2 gene on chromosome 5 was demonstrated when the gene was first identified,114 and several mutations have now been identified. Interestingly, they appear to cluster in the central region of the gene,117 and the phenotypes of mutations in other regions have not been convincingly demonstrated. MITRAL VALVE PROLAPSE SYNDROME Mitral valve prolapse syndrome is inherited in an autosomal dominant manner with variable expression. Women appear to manifest the valvular findings more frequently than men. Affected individuals have mitral prolapse and frequently have subtle skeletal findings that suggest that they may have Marfan syndrome. They do not appear to be at risk for aortic valvular or aortic root complications, however, and they do not have lens dislocation. Concern about the family often is raised and, in some cases, the diagnosis of Marfan syndrome is mistakenly made. ISOLATED ECTOPIA LENTIS Many of the features of Marfan syndrome, like mitral valve prolapse, may appear as isolated features. A rare example of this is isolated ectopia lentis, a dominantly inherited disorder in which aortic root dilatation and other features of Marfan syndrome are missing or exceedingly mild. The disorder results from mutations in FBN1.118 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48.

Piez KA, Reddi AH, eds. Extracellular matrix biochemistry. New York: Elsevier, 1984. Prockop DJ, Kivirikko KI. Heritable diseases of collagen. N Engl J Med 1984; 311:376. Hollister DW, Byers PH, Holbrook KA. Genetic disorders of collagen metabolism. Adv Hum Genet 1982; 12:1. McKusick VA. Heritable disorders of connective tissue, 4th ed. St. Louis: CV Mosby, 1972. Beighton P. McKusick's heritable disorders of connective tissue, 5th ed. St. Louis: CV Mosby, 1993. Royce PM, Steinmann B. Connective tissue and its heritable disorders: molecular, genetic, and medical aspects. New York: Wiley-Liss, 1993. Cheah KSE. Collagen genes and inherited connective tissue disease. Biochem J 1985; 229:287. Byers PH. Disorders of collagen biosynthesis and structure. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 7th ed. New York: McGraw-Hill, 1995: 4029. Smith R, Francis MJO, Houghton GF. Brittle bone disease: osteogenesis imperfecta. London: Butterworth, 1983. Sillence DO, Senn A, Danks DM. Genetic heterogeneity of osteogenesis imperfecta. J Med Genet 1979; 16:101. Dalgleish R. Mutations in type I and type III collagen genes. In: Humphries SE, Malcolm S, eds. From genotype to phenotype. Oxford: Bios Scientific Publishers, 1994:49. Barsh GS, David KE, Byers PH. Type I osteogenesis imperfecta: a nonfunctional allele for proa1(I) chains of type I procollagen. Proc Natl Acad Sci U S A 1982; 79:3838. Rowe DW, Shapiro JR, Poirier M, Schlesinger S. Diminished type I collagen synthesis and reduced a 1(I) collagen messenger RNA in cultured fibroblasts from patients with dominantly inherited (type I) osteogenesis imperfecta. J Clin Invest 1985; 76:604. Sykes B, Ogilvie D, Wordsworth P, et al. Osteogenesis imperfecta is linked to both type I collagen structural genes. Lancet 1986; 2:69. Willing MC, Cohn DH, Byers PH. Frameshift mutation near the 3' end of the COL1A1 gene of type I collagen predicts an elongated proa1(I) chain and results in osteogenesis imperfecta type I. J Clin Invest 1990; 85:282. Willing MC, Purchno CJ, Atkinson M, Byers PH. Osteogenesis imperfecta type I is commonly due to a COL1A1 null allele of type I collagen. Am J Hum Genet 1992; 51:508. Willing MC, Deschenes SP, Scott DA, et al. Osteogenesis imperfecta type I: molecular heterogeneity for COL1A1 null alleles of type I collagen. Am J Hum Genet 1994; 55:638. Sillence DO, Barlow KK, Barber AP, et al. Osteogenesis imperfecta type II: delineation of the phenotype with reference to genetic heterogeneity. Am J Med Genet 1984; 17:407. Byers PH, Tsipouras P, Bonadio JF, et al. Perinatal lethal osteogenesis imperfecta (OI type II): a biochemically heterogeneous disorder usually due to new mutation in the genes for type I collagen. Am J Hum Genet 1988; 42:237. Barsh GS, Byers PH. Reduced secretion of structurally abnormal type I procollagen in a form of osteogenesis imperfecta. Proc Natl Acad Sci U S A 1981; 78:5142. Williams CJ, Prockop DJ. Synthesis and processing of a type I procollagen containing shortened proa1 chains by fibroblasts from a patient with osteogenesis imperfecta. J Biol Chem 1983; 258:5915. Chu M-L, Williams CJ, Pepe G, et al. Internal deletion in a collagen gene in a perinatal lethal form of osteogenesis imperfecta. Nature 1983; 304:78. Chu M-L, Gargiulo V, Williams CJ, Ramirez F. Multiexon deletion in an osteogenesis imperfecta variant with increased type III collagen mRNA. J Biol Chem 1985; 260:691. Barsh GS, Roush C, Bonadio J, et al. Intron-mediated recombination may cause a deletion in an a1 type I collagen chain in a lethal form of osteogenesis imperfecta. Proc Natl Acad Sci USA 1985; 82:2870. DeWet WJ, Pihlajaniemi T, Myers J, et al. Synthesis of a shortened proa2(I) chain and decreased synthesis of proa2(I) chains in a proband with osteogenesis imperfecta. J Biol Chem 1983; 258:7721. Bonadio J, Byers PH. Subtle structural alterations in the chains of type I pro-collagen produce osteogenesis imperfecta type II. Nature 1985; 316:363. Steinmann B, Rao VH, Bruckner P, et al. Cysteine in the triple-helical domain of one allelic product of the a1(I) gene of type I collagen produces a lethal form of osteogenesis imperfecta. J Biol Chem 1984; 259:11129. Cohn DH, Byers PH, Steinmann B, Gelinas RE. Lethal osteogenesis imperfecta resulting from a single nucleotide change in one human pro a1(I) collagen allele. Proc Natl Acad Sci USA 1986; 83:6045. Cohn DH, Starman BJ, Blumberg B, Byers PH. Recurrence of lethal osteogenesis imperfecta due to parental mosaicism for a dominant mutation in a human type I collagen gene (COL1A1). Am J Hum Genet 1990; 46:591. Sillence DO, Barlow KK, Cole WG, et al. Osteogenesis imperfecta type III: delineation of the phenotype with reference to genetic heterogeneity. Am J Med Genet 1986; 23:821. Deak SB, Nicholls A, Pope FM, Prockop DJ. The molecular defect in a non-lethal variant of osteogenesis imperfecta: synthesis of proa2(I) chains which are not incorporated into trimers of type I procollagen. J Biol Chem 1983; 258:15192. Pihlajaniemi T, Dickson LA, Pope FM, et al. Osteogenesis imperfecta: cloning of a pro-a2(I) collagen gene with a frameshift mutation. J Biol Chem 1984; 259:12941. Tsipouras P, Myers JC, Ramirez F, Prockop DJ. Restriction fragment length polymorphism associated with the proa2(I) gene of human type I procollagen: application to a family with an autosomal dominant form of osteogenesis imperfecta. J Clin Invest 1983; 72:1261. Wenstrup RJ, Cohn DH, Cohen Y, Byers PH. Arginine for glycine substitution in the triple helical domain of the products of one a2(I) collagen allele (COL1A2) produces the osteogenesis imperfecta type IV phenotype. J Biol Chem 1988; 263:7734. Beighton P. The Ehlers-Danlos syndrome. London: William Heinemann, 1970. Vogel A, Holbrook KA, Steinmann B, et al. Abnormal collagen fibril structure in the gravis form (type I) of the Ehlers-Danlos syndrome. Lab Invest 1979; 40:201. Beighton P, Curtis D. X-linked Ehlers-Danlos syndrome type V: the next generation. Clin Genet 1985; 27:472. Steinmann B, Royce PM, Superti-Furga A. The Ehlers-Danlos syndrome. In: Royce PM, Steinmann B, eds. Connective tissue and its heritable disorders: molecular, genetic, and medical aspects. New York: Wiley-Liss, 1993. Beighton P, DePaepe A, Steinmann B, et al. Ehlers-Danlos syndrome: revisited nosology. Am J Med Genet 1998; 77:31. Loughlin J, Irven C, Hardwick LJ, et al. Linkage of the gene that encodes the alpha 1 chain of type V collagen (COL5A1) to type II Ehlers-Danlos syndrome (EDS II). Hum Mol Genet 1995; 4:1649. Burrows NP, Nicholls AC, Yates JR, et al. The gene encoding collagen alpha1(V) (COL5A1) is linked to mixed Ehlers-Danlos syndrome type I/II. J Invest Dermatol 1996; 106:1273. Toriello HV, Glover TW, Takahara K, et al. A translocation interrupts the COL5A1 gene in a patient with Ehlers-Danlos syndrome and hypomelanosis of Ito. Nat Genet 1996; 13:361. Burrows NP, Nicholls AC, Richards AJ, et al. A point mutation in an intronic branch site results in aberrant splicing of COL5A1 and in Ehlers-Danlos syndrome type II in two British families. Am J Hum Genet 1998; 63:390. Nicholls AC, Oliver JE, McCarron S, et al. An exon skipping mutation of a type V collagen gene (COL5A1) in Ehlers-Danlos syndrome. J Med Genet 1996; 33:940. Richards AJ, Martin S, Nicholls AC, et al. A single base mutation in COL5A2 causes Ehlers-Danlos syndrome type II. J Med Genet 1998; 35:846. Stewart RD, Hollister DW, Rimoin DL. A new variant of the Ehlers-Danlos syndrome: an autosomal dominant disorder of fragile skin, abnormal scarring, and generalized periodontitis. Birth Defects 1977; 13(3B):85. Arneson MA, Hammerschmidt DE, Furcht LT, King RA. A new form of Ehlers-Danlos syndrome: fibronectin corrects defective platelet function. JAMA 1980; 244:144. Pope FM, Martin GR, Lichtenstein JR, et al. Patients with Ehlers-Danlos syndrome type IV lack type III collagen. Proc Natl Acad Sci USA 1975; 72:1314.

49. Kuivaniemi H, Tromp G, Prockop DJ. Mutations in fibrillar collagens (types I, II, III, and XI), fibril-associated collagen (type IX), and network-forming collagen (type X) cause a spectrum of diseases of bone, cartilage, and blood vessels. Hum Mutat 1997; 9:300. 50. Pepin M, Schwarze V, Superti-Furga A, Byers PH. Clinical and genetic features of Ehlers-Danlos syndrome type IV, the vascular type. N Engl J Med 2000; 342:673. 51. Rudd NL, Nimrod C, Holbrook KA, Byers PH. Pregnancy complications in type IV Ehlers-Danlos syndrome. Lancet 1983; 1:50. 52. Pinnell SR, Krane SM, Kenzora JE, Glimcher MJ. A heritable disorder of connective tissue: hydroxylysine-deficient collagen disease. N Engl J Med 1972; 266:1013. 53. Steinmann B, Gitzelmann R, Vogel A, et al. Ehlers-Danlos syndrome in two siblings with deficient lysyl hydroxylase activity in cultured skin fibroblasts but only mild hydroxylysine deficient skin. Helv Paediatr Acta 1975; 30:255. 54. Elsas LJ, Miller RL, Pinnell SR. Inherited human collagen lysyl hydroxylase deficiency: ascorbic acid response. Pediatrics 1978; 92:378. 55. Dembure PP, Priest JH, Snoddy SC, Elsas LJ. Genotyping and prenatal assessment of collagen lysyl hydroxylase deficiency in a family with Ehlers-Danlos syndrome, type VI. Am J Hum Genet 1984; 36:783. 56. Pasquali M, Dembure PP, Still MJ, Elsas LJ. Urinary pyridinium cross-links: a noninvasive diagnostic test for Ehlers-Danlos syndrome type VI. N Engl J Med 1994; 331:132. 57. Steinmann B, Eyre DR, Shao P. Urinary pyridinoline cross-links in Ehlers-Danlos syndrome type VI. Am J Hum Genet 1995; 57:1505. 58. Heikkinen J, Toppinen T, Yeowell H, et al. Duplication of seven exons in the lysyl hydroxylase gene is associated with longer forms of a repetitive sequence within the gene and is a common cause for the type VI variant of Ehlers-Danlos syndrome. Am J Hum Genet 1997; 60:48. 59. Lichtenstein JR, Martin GR, Kohn L, et al. Defect in conversion of procollagen to collagen in a form of Ehlers-Danlos syndrome. Science 1973; 182:298. 60. Steinmann B, Tuderman L, Peltonen L, et al. Evidence for structural mutation of procollagen type I in a patient with the Ehlers-Danlos syndrome type VII. J Biol Chem 1980; 255:8887. 61. Cole WG, Chan D, Chambers GW, et al. Deletion of 24 amino acids from the proa1(I) chain of type I procollagen in a patient with the Ehlers-Danlos syndrome type VII. J Biol Chem 1986; 261:5496. 62. Byers PH, Duvic M, Atkinson M, et al. Ehlers-Danlos syndrome type VIIA and VIIB result from splice junction mutations or genomic deletions that involve exon 6 in the COL1A1 and COL1A2 genes of type I collagen. Am J Med Genet 1997; 72:94. 63. Smith LT, Wertelecki W, Milstone LM, et al. Human dermatosparaxis: a form of Ehlers-Danlos syndrome that results from failure to remove the amino-terminal propeptide of type I procollagen. Am J Hum Genet 1992; 51:235. 64. Nusgens BV, Verellen-Dumoulin C, Hermanns-Le T, et al. Evidence for a relationship between Ehlers-Danlos type VII C in humans and bovine dermatosparaxis. Nat Genet 1992; 1:214. 65. Colige A, Sieron AL, Li S-W, et al. Human Ehlers-Danlos type VII C and bovine dermatosparaxis are caused by mutations in the procollagen I N-proteinase gene. Am J Hum Genet 1999; 65:308. 66. Spranger JW, Langer LO, Wiedemann HR. Bone dysplasias: an atlas of constitutional disorders of skeletal development. Philadelphia: WB Saunders, 1974. 67. Rimoin DL. The chondrodystrophies. Adv Hum Genet 1975; 5:1. 68. Sillence DO, Horton WA, Rimoin DL. Morphologic studies in the skeletal dysplasia. Am J Pathol 1979; 96:813. 69. Spranger J. Radiologic nosology of bone dysplasias. Am J Med Genet 1990; 34:96. 70. Horton WA, Hecht JT. The chondrodysplasias. In: Royce PM, Steinmann B, eds. Connective tissue and its heritable disorders: molecular, genetic, and medical aspects. New York: Wiley-Liss, 1993:641. 71. Stickler GB, Belau PG, Farrell FJ, et al. Hereditary progressive arthroophthalmopathy. Proc Staff Meet Mayo Clin 1965; 40:433. 72. Snead MP, Yates JR. Clinical and molecular genetics of Stickler syndrome. J Med Genet 1999; 36:353. 73. Murray L, Bautista J, James PL, Rimoin DL. Type II collagen defects in the chondrodysplasias: I. Spondyloepiphyseal dysplasias. Am J Hum Genet 1989; 45:5. 74. Tiller GE, Rimoin DL, Murray LW, Cohn DH. Tandem duplication within a type II collagen gene (COL2A1) exon in an individual with spondyloepiphyseal dysplasia. Proc Natl Acad Sci U S A 1990; 87:3889. 75. Lee B, Vissing H, Ramirez F, et al. Identification of the molecular defect in a family with spondyloepiphyseal dysplasia. Science 1989; 244:978. 76. Chan D, Cole WG. Low basal transcription of genes for tissue-specific collagen by fibroblasts and lymphoblastoid cells: application to the characterization of a glycine 997 to serine substitution in a1(II) collagen chains of a patient with spondyloepiphyseal dysplasia. J Biol Chem 1991; 266:12487. 77. Tiller GE, Weis MA, Eyre DR, et al. An RNA splicing mutation (G+5IVS20) in the gene for type II collagen (COL2A1 produces spondyloepiphyseal dysplasia congenita (SEDC). Am J Hum Genet 1992; 51:A37. Abstract 136. 78. Reginato AJ, Passano GM, Neumann G, et al. Familial spondyloepiphyseal dysplasia tarda, brachydactyly, and precocious osteoarthritis associated with an arginine 75®cysteine mutation in the procollagen type II gene in a kindred of Chiloe Islanders. I. Clinical, radiographic, and pathologic findings. Arthritis Rheum 1994; 37:1078. 79. Godfrey M, Keene DR, Blank E, et al. Type II achondrogenesis-hypochon-drogenesis: morphologic and immunohistopathologic studies. Am J Hum Genet 1988; 43:894. 80. Holden P, Canty EG, Mortier GR, et al. Identification of novel pro-a2(IX) collagen gene mutations in two families with distinctive oligo-epiphyseal forms of multiple epiphyseal dysplasia. Am J Hum Genet 1999; 65:31. 81. Paassilta P, Lohiniva J, Annunen S, et al. COL9A3: a third locus for multiple epiphyseal dysplasia. Am J Hum Genet 1999; 64:1036. 82. Lachman RS, Rimoin DL, Spranger J. Metaphyseal chondrodysplasia, Schmid type. Clinical and radiographic delineation with a review of the literature. Pediatr Radiol 1988; 18:93. 83. Warman ML, Abbott M, Apte SS, et al. A type X collagen mutation causes Schmid metaphyseal chondrodysplasia. Nat Genet 1993; 5:79. 84. McIntosh I, Abbott MH, Warman ML, et al. Additional mutations of type X collagen confirm COL10A1 as the Schmid metaphyseal chondrodysplasia locus. Hum Mol Genet 1994; 3:303. 85. Wallis GA, Rash B, Sykes B, et al. Mutations within the gene encoding the a1(X) chain of type X collagen (COL10A1) cause metaphyseal chondrodysplasia type Schmid but not several other forms of metaphyseal chondrodysplasia. J Med Genet 1996; 33:450. 86. Wallis GA, Rash B, Sweetman WA, et al. Amino acid substitution of conserved residue in the carboxyl-terminal domain of the a1(X) chain of type X collagen occurs in two unrelated families with metaphyseal chondrodysplasia type Schmid. Am J Hum Genet 1994; 54:169. 87. Atkin CL, Gregory MC, Border WA. Alport syndrome. In: Schrier RW, Gotschalk CW, eds. Disease of the kidney, 4th ed. Boston: Little, Brown and Company, 1988:617. 88. Atkin CL, Hasstedt SJ, Menlove L, et al. Mapping of Alport syndrome gene to the long arm of the X chromosome. Am J Hum Genet 1988; 42:249. 89. Brunner H, Schroder C, Van Bennekom C, et al. Localization of the gene for X-linked Alport's syndrome. Kidney Int 1988; 34:507. 90. Flinter FA, Abbs S, Bobrow M. Localization of the gene for classic Alport syndrome. Genomics 1989; 4:335. 91. Martin P, Heiskari N, Zhou J, et al. High mutation detection rate in the COL4A5 collagen gene in suspected Alport syndrome using PCR and direct DNA sequencing. J Am Soc Nephrol 1998; 9:2291. 92. Antignac C, Zhou J, Sanak M, et al. Alport syndrome and diffuse leiomyomatosis: deletions in the 5' end of the COL4A5 collagen gene. Kidney Int 1992; 42(5):1178. 93. Fine J-D, Bauer EA, Briggaman RA, et al. Revised clinical and laboratory criteria for subtypes in inherited epidermolysis bullosa: a consensus report by the subcommittee on diagnosis and classification of the National Epidermolysis Bullosa Registry. J Am Acad Dermatol 1991; 24:119. 94. Bruckner-Tuderman L. Epidermolysis bullosa. In: Royce PM, Steinmann B, eds. Connective tissue and its heritable disorders: molecular, genetic, and medical aspects. New York: Wiley-Liss, 1993:507. 95. Palade G, Farquhar M. A special fibril of the dermis. J Cell Biol 1965; 27:215. 96. Keene DR, Sakai Y, Lunstrum GP, et al. Type VII collagen forms an extended network of anchoring fibrils. J Cell Biol 1987; 104:611. 97. Ryynanen M, Ryynanen J, Sollberg S, et al. Genetic linkage of type VII collagen (COL7A1) to dominant dystrophic epidermolysis bullosa in families with abnormal anchoring fibrils. J Clin Invest 1992; 89:974. 98. Pulkkinen L, Uitto J. Mutation analysis and molecular genetics of epidermolysis bullosa. Matrix Biol 1999; 18:29. 99. Tidman MJ, Eady RAJ. Evaluation of anchoring fibrils and other components of the dermal-epidermal junction in dystrophic epidermolysis bullosa by a quantitative technique. J Invest Dermatol 1985; 84:374. 100. Heagerty AHM, Kennedy AR, Leigh IM, et al. Identification of an epidermal basement membrane defect in recessive forms of dystrophic epidermolysis bullosa by LH7:2 monoclonal antibody: use in diagnosis. Br J Dermatol 1986; 115:125. 101. Smith LT, Sybert VP. Intra-epidermal retention of type VII collagen in a patient with recessive dystrophic epidermolysis bullosa. J Invest Dermatol 1990; 94:261. 102. Hovnanian A, Duquesnoy P, Blanchet-Bardon C, et al. Genetic linkage of recessive dystrophic epidermolysis bullosa to the type VII collagen gene. J Clin Invest 1992; 90:1032. 103. Dietz HC, Cutting GR, Pyeritz RE, et al. Marfan syndrome caused by a recurrent de novo missense mutation in the fibrillin gene. Nature 1991; 352:337. 104. Sakai LY, Keene DR, Engvall E. Fibrillin, a new 350-kDa glycoprotein, is a component of extracellular microfibrils. J Cell Biol 1986; 103:2499. 105. Zhang H, Hu W, Ramirez F. Developmental expression of fibrillin genes suggests heterogeneity of extracellular microfibrils. N Engl J Med 1999; 340(17):1307. 106. Pyeritz RE, McKusick VA. The Marfan syndrome: diagnosis and management. N Engl J Med 1979; 300:772. 107. De Paepe A, Devereux RB, Dietz HC, et al. Revised diagnostic criteria for the Marfan syndrome. Am J Med Genet 1996; 62:417. 108. Maumenee IH. The eye in the Marfan syndrome. Trans Am Ophthalmol Soc 1981; 79:684. 109. Pyeritz RE, Wappel MA. Mitral valve dysfunction in the Marfan syndrome. Am J Med 1983; 74:797. 110. Gott VL, Greene PS, Alejo DE, et al. Replacement of the aortic root in patients with Marfan's syndrome. N Engl J Med 1999; 340:1307. 111. Godfrey M, Menashe V, Weleber RG, et al. Cosegregation of elastin-associated microfibrillar abnormalities with the Marfan phenotype in families. Am J Hum Genet 1990; 46:652. 112. Kainulainen K, Pulkkinen L, Savolainen A, et al. Location on chromosome 15 of the gene defect causing Marfan syndrome. N Engl J Med 1990; 323:935. 113. Milewicz DM, Pyeritz RE, Crawford ES, Byers PH. Marfan syndrome: defective synthesis, secretion, and extracellular matrix formation of fibrillin by cultured dermal fibroblasts. J Clin Invest 1992; 89:79. 114. Lee B, Godfrey M, Vitale E, et al. Linkage of Marfan syndrome and a phenotypically related disorder to two different fibrillin genes. Nature 1991; 352(6333):330. 115. Collod-Beroud G, Beroud C, Ades L, et al. Marfan Database (third edition): new mutations and new routines for the software. Nucleic Acids Res 1998; 26:229. 116. Shores J, Berger KR, Murphy EA, Pyeritz RE. Progression of aortic dilatation and the benefit of long-term beta-adrenergic blockade in Marfan's syndrome. N Engl J Med 1994; 330:1335. 117. Park ES, Putnam EA, Chitayat D, et al. Clustering of FBN2 mutations in patients with congenital contractural arachnodactyly indicates an important role of the domains encoded by exons 24 through 34 during human development. Am J Med Genet 1998; 78:350. 118. Hayward C, Brock DJ. Fibrillin-1 mutations in Marfan syndrome and other type-1 fibrillinopathies. Hum Mutat 1997; 10:415.

CHAPTER 190 HERITABLE DISEASES OF LYSOSOMAL STORAGE Principles and Practice of Endocrinology and Metabolism

CHAPTER 190 HERITABLE DISEASES OF LYSOSOMAL STORAGE WARREN E. COHEN Spectrum of Lysosomal Storage Diseases Gangliosidoses GM1 Gangliosidoses GM2 Gangliosidoses Sphingolipidoses Gaucher Disease Niemann-Pick Disease Krabbe Disease Metachromatic Leukodystrophy Fabry Disease Mucopolysaccharidoses Hurler Syndrome Scheie Syndrome Hunter Syndrome Sanfilippo Syndrome Morquio Syndrome Maroteaux-Lamy Syndrome Other Mucopolysaccharidoses Oligosaccharidoses (Mucolipidoses) Sialidosis Mannosidosis Aspartylglycosaminuria Carbohydrate-Deficient Glycoprotein Syndrome I-Cell Disease (Mucolipidosis II) Pseudo–Hurler Polydystrophy (Mucolipidosis III) Mucolipidosis IV Diagnostic Considerations Management Issues Genetic Issues Chapter References

SPECTRUM OF LYSOSOMAL STORAGE DISEASES At least 30, and possibly as many as 50, disorders of metabolism exist that cause the abnormal accumulation of metabolites in cell lysosomes. Although individually these disorders are rare, collectively they constitute a frequent cause of chronic illness and debilitation in childhood. The clinical manifestations of these disorders are diverse but they share a common pathogenesis, an insight into which requires a basic understanding of normal lysosomal function. Lysosomes are organelles surrounded by a single membrane within which are packaged enzymes that normally degrade exogenous substances and the products of endogenous metabolism, and thus serve as the “waste disposal” system of the cell. Through this system of lysosomal enzymes, complex carbohydrates, lipids, and proteins are degraded sequentially. The enzymes of this system initially are synthesized in the endoplasmic reticulum. Multiple short oligosaccharide chains are then attached to them. They are further modified in the Golgi apparatus and eventually are packaged into lysosomes of various types, depending on the cell type and the particular metabolic pathway in which they will function. Lysosomal storage disorders arise when a genetic defect causes deficient production or function of one or more of the lysosomal enzymes, with consequent abnormal accumulation of metabolites within the lysosome. The accumulation of these metabolites distends the lysosome and produces effects that are toxic to cell constituents. This disrupts normal cell function and eventually causes cell death. The clinical manifestations of any particular storage disorder are a consequence of the specific enzyme or enzymes affected, the metabolic pathway in which they are involved, the alteration of cell function associated with abnormal lysosomal storage, and the particular tissues in which the affected enzymes normally function. All lysosomal storage diseases are progressive because the process of lysosomal storage is continuous. Because many lysosomal enzymes participate in the metabolism of neural tissues, the central and peripheral nervous systems are affected in several of the lysosomal storage diseases, and this is manifest either by progressive psychomotor retardation or by any of various progressive neurologic signs and symptoms. The diverse and multisystemic clinical manifestations observed in these disorders demonstrate that normal lysosomal enzyme function is crucial to the health of many human tissues. No simple clinical approach exists to the classification of this clinically diverse group of disorders. Thus, lysosomal storage disorders usually are classified not by their clinical manifestations, but rather according to the nature of the macromolecule that is abnormally catabolized and consequently accumulates within the lysosome. The sphingolipidoses are associated with the accumulation of complex lipids, the basic structure of which is sphingosine, a long-chain amino alcohol. The mucopolysaccharidoses (MPSs) are associated with the accumulation of complex protein-sugar macromolecules known as mucopolysaccharides or glycosaminoglycans. The oligosaccharidoses or mucolipidoses are associated with storage of complex glycoproteins. Within each of these biochemical categories is a diverse group of disorders, some with classically recognizable clinical and laboratory features and many with variant forms that are atypical in their clinical and biochemical findings1,2,3,4 and 5 (Table 190-1). Increasingly, molecular genetic techniques are demonstrating the heterogeneity of lysosomal storage diseases. Many of the specific disorders that traditionally have been described by their clinical and biochemical features are now recognized to represent more than one genetic disorder. Some of this heterogeneity is explained by the fact that the same clinical and biochemical abnormality might result from the abnormal functioning of two or more different gene loci, but much of it may be due to different mechanisms of genetic mutation that result in varying degrees of dysfunction in a specific gene.6

TABLE 190-1. Clinical Aspects of Lysosomal Storage Disorders

GANGLIOSIDOSES GM1 GANGLIOSIDOSES The GM1 gangliosidoses are characterized by a deficiency of GM1 ganglioside b-galactosidase, with the consequent accumulation of both GM1 ganglioside and a keratan sulfate–like muco-polysaccharide.7,8,9 and 10 Two classic clinical syndromes exist. INFANTILE GM1 GANGLIOSIDOSIS

In infantile GM1 gangliosidosis, symptoms usually are noted shortly after birth. Psychomotor retardation associated with hypotonia, poor suckling, lethargy, and hypoactivity are observed early. Weight gain is poor, and characteristic dysmorphic features appear early: frontal bossing, coarsened facial features, long philtrum, gingival hyperplasia, macroglossia, and, frequently, edema of the face and extremities. Visceral storage causes hepatosplenomegaly. Radiographs reveal characteristic changes in the spine and long bones. Over the first several months, the progressive psychomotor retardation becomes more apparent, and the initial hypotonia gives way to spasticity with hyperreflexia and tonic spasms. A zone of retinal degeneration surrounding the normally red macula produces a macular cherry-red spot in approximately one-half of patients, but this may not be evident until after the age of 6 months. Slight corneal clouding may be present. Progression of the disorder renders the child deaf and blind, with decerebrate rigidity usually present by the age of 12 months; death usually occurs at approximately the age of 2 years. LATE INFANTILE GM1 GANGLIOSIDOSIS In late infantile GM1 gangliosidosis, the symptoms do not manifest until the age of 1 to 2 years, when gait disturbance, ataxia, incoordination, and regression of language skills appear. As the disorder progresses, spasticity develops, and seizures frequently occur. These may become a difficult management problem. Unlike in the infantile form, the corneas are clear, no hepatosplenomegaly is present, and no significant bone changes are found on radiography. Abnormalities of cerebral cortex, white matter, and deep nuclei on magnetic resonance imaging have been reported.11 Death usually occurs between the ages of 3 and 10 years, generally during an episode of pneumonia. Several atypical variants of GM1 gangliosidosis also have been described, including several with onset in later childhood. GM2 GANGLIOSIDOSES The GM2 gangliosidoses are a group of disorders characterized by the inherited deficiency of the lysosomal enzyme hexosaminidase. The deficiency leads to an accumulation of GM2 ganglioside, other glycosphingolipids, and various other metabolites.10,12,13,14,15 and 16 The biochemical genetics of hexosaminidase function are complex because the activity of this enzyme involves at least two subunits, three isoenzymes, three gene loci, and at least one activator protein.13 More than 20 forms of hexosaminidase deficiency are known. TAY-SACHS DISEASE Previously, the most common form of hexosaminidase deficiency was Tay-Sachs disease, a disorder that has become uncommon because of a widespread carrier detection program targeted at a well-defined population at risk for the disorder. Approximately 1 in 30 Jews of Eastern European (Ashkenazi) heritage are carriers of the autosomal recessive trait. Affected children with this disorder are deficient in hexosaminidase A and hexosaminidase S. They appear healthy during the first 4 to 6 months of life, after which signs of psychomotor deterioration appear. The earliest signs usually are hypotonia and inability to sit, followed by more obvious signs of regression of motor skills. An abnormal motor response to stimulation by sound and sometimes by light is characteristic of the disorder, appearing as an exaggerated infantile startle response. As the disorder progresses, weakness increases, and spasticity and hyperreflexia develop. A macular cherry-red spot is seen in 90% of cases, and by 12 to 18 months, severe visual impairment is present, with eventual blindness. During the second year, the intraneural accumulation of storage material leads to marked enlargement of the brain, reflected by markedly accelerated growth of the head circumference. Seizures occur; they may be generalized, minor motor, or myoclonic in character. By the age of 3 years, the child usually appears debilitated, with decerebrate rigidity, dementia, marked macrocephaly, blindness, and seizures. Most affected children die between the ages of 3 and 5 years; a few have survived longer. SANDHOFF DISEASE Children with Sandhoff disease are deficient in hexosaminidase A and hexosaminidase B. The clinical picture is almost identical to that of infantile Tay-Sachs disease, but many children with Sandhoff disease also manifest moderate hepatosplenomegaly—a sign absent in Tay-Sachs disease. This disorder is rarer than Tay-Sachs disease and has no ethnic predilection. OTHER VARIETIES Other variants of Tay-Sachs disease include disorders with onset in infancy, childhood, and adulthood.14 Patients have been described with various symptom complexes, including a juvenile-onset disorder with manifestations similar to infantile Tay-Sachs, with blindness, seizures, and progressive psychomotor retardation. Molecular studies have demonstrated that the gene responsible for the juvenile disorder is allelic to the gene responsible for infantile Tay-Sachs disease.15 Other such disorders include a disorder appearing as spino-cerebellar degeneration, a disorder with a clinical picture resembling that of isolated lower motor neuron disease (similar to Kugelberg-Welander disease), and a disorder with signs of both upper and lower motor neuron dysfunction that resembles amyotrophic lateral sclerosis. Some individuals with hexosaminidase deficiency have no symptoms. Because of the wide variation of clinical manifestations in these disorders, a reasonable approach is to consider the possibility of hexosaminidase deficiency in virtually any individual with an unexplained degenerative neurologic disorder.

SPHINGOLIPIDOSES GAUCHER DISEASE Gaucher disease is a group of sphingolipidoses characterized by a deficiency of glucocerebrosidase with the consequent accumulation of glucocerebroside in body tissues.17,18 The disorder is multisystemic and affects tissue macrophages of the reticuloendothelial system, primarily in the spleen, liver, bone, and lung. Macrophages in these tissues become laden with lipid, giving them a characteristic light-microscopic appearance (“Gaucher cells”). TYPE 1 GAUCHER DISEASE Type 1 Gaucher disease is by far the most common of the lysosomal storage disorders. Previously it was designated the “adult” form of Gaucher disease, but the onset of symptoms may be in childhood or adulthood, and many patients have no symptoms. Splenomegaly is the usual initial symptom; it may be associated with various degrees of anemia, thrombocytopenia, and leukopenia. Hepatomegaly usually is not as prominent and may appear later in the disease course. Any hepatic dysfunction usually is mild. Skeletal involvement produces various bone complications. Many patients experience episodes of severe bone pain similar to the crises of sickle cell disease. These episodes may be associated with fever and local swelling, erythema, and tenderness, and may be mistaken for osteomyelitis. Nonspecific chronic, mild joint pain is a common complaint. Other bone complications include a propensity to develop multiple fractures, vertebral collapse, and aseptic necrosis of the femoral head. Radiographs reveal osteopenia, cortical thinning, endosteal scalloping, cyst-like cavities, and characteristic Erlen-meyer-flask deformities of the distal femurs. Pulmonary complications are much less common and result from interstitial pulmonary infiltration. In those few patients who manifest severe liver dysfunction, associated pulmonary complications with intrapulmonary shunting occur. Patients with type 1 Gaucher disease exhibit no clinical signs of neurologic dysfunction. The clinical course of type 1 Gaucher disease is extremely variable. Most patients are affected only mildly; they may lead normal, productive lives, with minimal to mild disability and with normal longevity. Others may be severely debilitated, and death may occur in childhood. An increased incidence of multiple myeloma, leukemia, and Hodgkin disease is found among individuals with this disorder. Intravenous enzyme replacement therapy, using a modified form of the deficient enzyme that has been engineered to maximize cellular uptake, has been successful in treating the hematopoietic, hepatic, and skeletal abnormalities associated with type 1 Gaucher disease.19 TYPE 2 GAUCHER DISEASE Type 2 Gaucher disease has a characteristic clinical course. It was previously designated the “infantile” form of the disorder, and the age of onset usually is between 3 and 6 months. Hepatosplenomegaly is noted initially, followed by signs of neurologic dysfunction. The classic clinical triad of symptoms consists of trismus, strabismus, and head retroflexion. Spastic quadriparesis progresses rapidly, with associated hyperreflexia, cranial nerve dysfunction, and dysphagia. Seizures may occur but are not a common complication. As the disorder progresses, the child becomes hypotonic, lethargic, and apathetic. Most children die by the age of 1 year, and virtually none survive beyond the age of 2 years. TYPE 3 GAUCHER DISEASE Type 3 Gaucher disease has characteristics intermediate between those of types 1 and 2, but the clinical presentation can be variable. Systemic complications are similar to those of type 1 Gaucher disease but in general may be more severe. Neurologic complications are of later onset and progress more slowly. They include

ataxia, incoordination, myoclonus, seizures, a characteristic oculomotor apraxia (supranuclear eye movement disorder), and slowly progressive dementia. NIEMANN-PICK DISEASE Niemann-Pick disease is a group of disorders characterized by the abnormal accumulation of sphingomyelin.20,21 and 22 In two of the forms of the disorder, a deficiency of the lysosomal enzyme sphingomyelinase is found. In other variant forms, the nature of the biochemical abnormality is unclear. TYPE A NIEMANN-PICK DISEASE Types A and B Niemann-Pick disease are associated with deficient sphingomyelinase activity. Type A disease is characterized by hepatosplenomegaly, which usually develops by the age of 6 months. It frequently is associated with failure to thrive. The enlargement of the liver precedes and is more prominent than enlargement of the spleen. Progressive neurologic deterioration ensues, with a loss of motor and cognitive skills, axial hypotonia coexistent with spasticity of the extremities, dysphagia, and blindness. A cherry-red macular spot is seen in approximately one-half the cases. As the disorder progresses, the child becomes increasingly cachectic, rigid, and opisthotonic. Foam cells (vacuolated histiocytes) are found in the bone marrow, peripheral blood, and other tissues, and are of diagnostic importance. Death usually occurs by the age of 3 years and frequently ensues after fulminant hepatic failure. OTHER TYPES OF NIEMANN-PICK DISEASE In type B Niemann-Pick disease, hepatosplenomegaly becomes apparent in infancy or childhood. The nervous system is not involved, and no neurologic symptoms are present. Other clinical forms that have been categorized as variants of Niemann-Pick disease are poorly understood. Type C is characterized by onset of mild hepatosplenomegaly in infancy or childhood associated with slowly progressive ataxia, dementia, and seizures. The disorder can be associated with childhood narcolepsy.23 Foam cells are found in bone and liver biopsy specimens. Most affected patients die by the age of 20 years. A similar disorder, which is limited geographically to a coastal area in western Nova Scotia, has been designated type D disease. A few reported patients with visceral storage and foam cells have been designated as having type E Niemann-Pick disease. The biochemical defect in type C is caused by defective intracellular transport of exogenous cholesterol, resulting in the accumulation of unesterified cholesterol in the lysosome.24 KRABBE DISEASE Krabbe disease, or globoid cell leukodystrophy, is characterized by a deficiency of galactocerebroside b-galactosidase, which causes extensive demyelination of the central and peripheral nervous systems.25,26 and 27 Neurologic symptoms usually are apparent by the age of 6 months, and some patients have symptoms at birth. In the classic disorder, the initial symptoms include hyperirritability and inexplicable crying, with episodes of tonic stiffening. Progressive development of spasticity is seen, and seizures frequently occur. Hyperreflexia, optic atrophy, opisthotonic posturing, and an arrest of head growth are noted, and deafness may occur. Tendon reflexes are diminished early in the course secondary to the peripheral neuropathy; later, the reflexes become hyperreactive. Nerve conduction velocities are delayed, and the cerebrospinal fluid protein level is elevated. No signs of visceral storage are seen. Late in the course, the child appears decerebrate, blind, and unresponsive to stimuli. Most affected children die by the age of 1 year, and virtually all die by the age of 2 years. Several atypical clinical presentations have been reported, including later-onset forms featuring dementia, spasticity, and optic atrophy, as well as a late-onset disorder with spinocerebellar degeneration. METACHROMATIC LEUKODYSTROPHY Metachromatic leukodystrophy is another disorder of myelin metabolism. It is caused by a deficiency of the lysosomal enzyme arylsulfatase A and is associated with the accumulation of cere-broside sulfate.28,29,30 and 31 This disorder probably is the most common of the inherited leukodystrophies, which include disorders of lysosomal metabolism, disorders of peroxisomal metabolism, and disorders of uncertain etiology. In the classic disorder, affected children usually appear normal until the age of 12 to 18 months, at which time difficulty of gait becomes apparent. The initial signs and symptoms may differ depending on the balance of the effects of pyramidal tract dysfunction and of peripheral neuropathy. Some children have flaccid paraparesis and hypore-flexia, others have spasticity and hyporeflexia, and the remainder have spasticity and hyperreflexia. The onset of spastic weakness may be abrupt and asymmetric. As the disorder progresses, spastic quadriparesis and dementia ensue, and dysarthria and dysphagia are frequent problems. Visual impairment may occur, and various ophthalmologic findings may be present. Approximately one-third of patients have optic atrophy. Some individuals have a characteristic grayish discoloration of the macula, whereas others have a macular cherry-red spot. At the end stages of the disease, the child appears decerebrate, often with sensory-induced tonic spasms, pseudobulbar signs, and unresponsiveness. Death ensues after a course of several years. Bone marrow transplantation appears to slow the progression of the disease, but often the risks of the procedure outweigh the potential benefits by the time the diagnosis is made. Several variant forms of metachromatic leukodystrophy have been described on clinical and biochemical grounds. In juvenile metachromatic leukodystrophy, symptoms appear between the ages of 3 and 16 years. In this disorder, school-age children frequently have behavioral abnormalities followed by incontinence and gait difficulty. The disorder progresses rapidly and renders the children debilitated, with spastic paraparesis, dementia, pseudobulbar signs, and tonic spasms. Most of these children die during the teenage years. In adult metachromatic leukodystrophy, progressive neurologic dysfunction often is heralded by behavioral and emotional abnormalities that frequently are mistaken for psychiatric illness. This form of the disease also is relentlessly progressive, ending in death after a variable period of neurologic deterioration. Two biochemical variants of metachromatic leukodystrophy are important to note. In the rare disorder of multiple sulfatase deficiency, nine lysosomal sulfatases, including arylsulfatase A, are deficient. The clinical manifestations include features of late infantile metachromatic leukodystrophy and also of the muco-polysaccharide disorders, with coarse facial features, skeletal abnormalities, and hepatosplenomegaly. Finally, in a few patients with clinical and histopathologic features of metachromatic leukodystrophy, no deficiency of arylsulfatase A can be demonstrated. These patients lack an activator protein, saposin B, which is required for that enzyme to function. FABRY DISEASE Fabry disease is a rare disorder characterized by a deficiency of a-galactocerebrosidase A, with the consequent accumulation of ceramide trihexoside in tissues.32,33 and 34 This multisystemic disorder is transmitted as an X-linked recessive trait, but symptoms frequently occur in carrier females. A characteristic initial symptom is peculiar painful acrodysesthesias. Patients complain of severe pain of the hands and feet, often triggered during febrile illnesses or by exercise or exposure to heat. The pain can be debilitating and, because of its peculiar nature, often is attributed to a psychiatric disorder. The pain frequently responds to therapy with phenytoin sodium (Dilantin) or carbamazepine (Tegretol). Other clinical findings include angiokeratomas and a characteristic corneal dystrophy that is evident on slit-lamp examination. Vascular changes in the conjunctiva and retina also occur. The more serious complications of the disorder include angina and myocardial infarction, and progressive renal failure. Hemodialysis and renal transplantation have become lifesaving procedures for affected individuals. Neurologic complications do not arise secondary to intraneuronal storage, as seen in other lysosomal disorders, but rather due to cerebrovascular disease, which causes recurrent strokes. Death usually ensues at approximately the age of 40 years in affected males, but longer survival has been reported.

MUCOPOLYSACCHARIDOSES The MPSs are a heterogeneous group of disorders characterized by a deficiency of the enzymes involved in the catabolism of complex carbohydrate polymers called glycosaminoglycans.35,36,37,38,39,40,41,42 and 43 These substances occur in large amounts in connective tissue of cartilage, bone, blood vessels, heart valves, skin, tendons, and cornea, and in lesser amounts in liver and brain. Thus, in storage disorders in which glycosaminoglycans accumulate, various multisystemic manifestations occur. HURLER SYNDROME Hurler syndrome (MPS IH), the prototype of the MPS disorders, is characterized by a deficiency of a-L -iduronidase, which leads to the accumulation of dermatan sulfate and heparan sulfate. It is the most severe and most rapidly progressive of this group of disorders, usually resulting in death by the age of 10 years. After an initial period of normal development in infancy, affected children begin to display signs of progressive dementia, with the loss of previously acquired cognitive skills. Characteristic physical features develop with time and include coarse facies, macroscaphocephaly, sagittal sutural ridging, corneal clouding, prominent abdomen, hepatosplenomegaly, umbilical hernias, and thickened skin (Fig. 190-1). Skeletal involvement produces a characteristic clinical and radiographic picture that has been termed dysostosis multiplex. Clinical features of dysostosis include short stature, lumbar lordosis, thoracolumbar gibbus deformity, and joint stiffness, with claw-hand

deformity (Fig. 190-2) and flexion contractures at the hips. Radiographic findings include hypoplastic vertebrae at the gibbus deformity with a broad-hook appearance, flaring of the iliac wings of the pelvis, coxa valga deformities, oarshaped widened ribs, an absence of diaphyseal constriction of metacarpals, and widened shafts of the long bones. Neurologic complications include progressive dementia, seizures, spasticity, entrapment neuropathies (especially carpal tunnel syndrome), arachnoid cysts, sensorineural deafness, optic atrophy, and retinitis pigmentosa. The deposition of storage material in the meninges overlying the cerebral hemispheres frequently causes hydrocephalus, and a similar deposition in the meninges around the cervical spine causes the syndrome of pachymeningitis cervicalis circumscripta, with resultant spinal cord compression or nerve root involvement.

FIGURE 190-1. Hurler syndrome (mucopolysaccharidosis I) in a 17-month-old girl, who shows typical coarsening of facial features characteristic of several of the disorders of mucopolysaccharide metabolism. Although the classic “gargoyle” facial appearance often depicted in photographs of individuals with Hurler syndrome is usually more striking, this child's features are typical for a toddler with the disorder.

FIGURE 190-2. Claw-hand deformity of woman with Scheie syndrome (mucopolysaccharidosis IS). This progressive hand deformity is typical of that seen in several of the mucopolysaccharidoses and is extremely disabling.

SCHEIE SYNDROME Scheie syndrome (MPS IS) is associated with the same enzyme deficiency as that in Hurler syndrome (a-L -iduronidase). MPS IS probably represents a different mutation at the same structural enzyme locus. Affected individuals have short stature, with coarse facies and signs of dysostosis multiplex. Visual impairment occurs because of severe corneal clouding, retinal pigmentary degeneration, and glaucoma. Cardiac involvement causes aortic stenosis, aortic regurgitation, or both. Neurologic complications include entrapment neuropathies, sensorineural deafness, and pachymeningitis cervicalis circumscripta. Intelligence is unaffected. A variant of the a-iduronidase deficiency states is the Hurler-Scheie syndrome (MPS IH/IS). Children with this disorder have clinical manifestations intermediate between those of the Hurler and Scheie syndromes. HUNTER SYNDROME Hunter syndrome is associated with deficiency of the lysosomal enzyme iduronidate sulfatase, with accumulation of dermatan sulfate and heparan sulfate in body tissues. A broad range of severity is seen among children with this disorder, and affected individuals are loosely categorized as having either the “severe” or the “mild” form of the disorder. Features are similar to those seen in Hurler syndrome but generally are less severe, and survival is longer. Clinical features include dementia that usually is milder and less rapidly progressive than in Hurler syndrome, short stature, a characteristic pebbled appearance of the skin, coarse facies, hepatosplenomegaly, umbilical hernias, and dysostosis multiplex. SANFILIPPO SYNDROME Sanfilippo syndrome (MPS III) is a clinically homogeneous group of disorders that are caused by four distinct deficiency states involving the following enzymes: heparan-sulfate sulfatase (type A); N-acetyl-a-D-glucosaminidase (type B); N-acetyl-CoA acetyltransferase (type C); and N-acetylglucosamine-6-sulfate sulfatase (type D). The clinical course begins with a variable period of normal psychomotor development (usually 18 to 24 months) followed by a slowing, and then a plateau, of developmental progress. Normal development may be preserved for as long as 5 to 9 years. After the plateau of developmental progress is reached, a rapid mental deterioration ensues, usually with an onset between the ages of 3 and 5 years. Mental deterioration often is heralded by behavioral difficulties and sleep abnormalities and is rapidly progressive, with a rapid loss of language abilities. Physical signs, which usually are not striking, may include mild coarsening of the features, coarse hair, hirsutism, and macrocephaly. Mild dwarfing of stature and some radiographic signs of dysostosis multiplex are present. Seizures occur commonly. MORQUIO SYNDROME Morquio syndrome (MPS IV) consists of two similar disorders characterized by a deficiency of galactosamine-6-sulfate sulfatase (type A) or b-galactosidase (type B). These disorders are manifest almost exclusively in the skeleton and through the secondary effects of these skeletal changes on the spinal cord. Intelligence usually is normal. Marked dwarfing of stature occurs, mild corneal clouding is present, and progressive deafness is almost invariable. The skeleton is affected in a characteristic manner. Restricted joint movement and claw-hand deformities do not occur; several joints (especially the wrists) are excessively loose secondary to ligamentous laxity. The odontoid process is characteristically absent or markedly hypoplastic, and this, combined with laxity and redundancy of the paraspinal ligaments, causes cervical myelopathy secondary to atlantoaxial subluxation. This serious complication can be prevented by cervical spine fusion, which should be considered in all patients with the disorder. MAROTEAUX-LAMY SYNDROME Maroteaux-Lamy syndrome (MPS VI) is characterized by a deficiency of arylsulfatase B and manifests clinically, in the severe or classic form, with coarse facial features, corneal clouding, short stature, and dysostosis multiplex with claw-hand deformity (Fig. 190-3). The course may be complicated by heart valve involvement, atlantoaxial subluxation, or hydrocephalus. Mental retardation is not associated with this disorder except as a complication of untreated hydrocephalus. Several patients have been reported with only mild features of the disorder.

FIGURE 190-3. Eight-year-old boy with Maroteaux-Lamy syndrome (mucopolysaccharidosis VI). Note the coarsened facial features, extremely short stature, and claw-hand deformities. Mental retardation is not a feature of this disorder.

OTHER MUCOPOLYSACCHARIDOSES Two other MPSs, Sly syndrome (MPS VII) and DiFerrante syndrome (MPS VIII), are so rare that they are of no practical clinical importance.

OLIGOSACCHARIDOSES (MUCOLIPIDOSES) SIALIDOSIS Sialidosis constitutes a rare group of disorders associated with a deficiency of N-acetylneuraminidase (sialidase), with the excretion of oligosaccharides containing sialic acid in the urine. The clinical manifestations are variable,44,45 and two classic syndromes have been delineated. In sialidosis type I, also known as the cherry-red spot/myoclonus syndrome, symptoms begin in adolescence, with progressive impairment of vision, a macular cherry-red spot, and I severe myoclonus.46 In sialidosis type II, symptoms begin in early or late childhood, with coarsened features suggestive of Hurler syndrome, dysostosis multiplex, hepatomegaly, and intellectual impairment. MANNOSIDOSIS Mannosidosis is a rare disorder associated with a deficiency of a-mannosidase, with the excretion of oligosaccharides containing mannose in the urine.45,47 The clinical features include mental retardation, coarsened facial features, dysostosis multiplex, hepatosplenomegaly, deafness, and recurrent infections. FUCOSIDOSIS Fucosidosis is a heterogeneous group of rare disorders associated with a deficiency of a-fucosidase.45,47 In fucosidosis type I, symptoms begin in infancy, with progressive psychomotor retardation, coarse facial features, hepatosplenomegaly, dysostosis multiplex, and seizures. In fucosidosis type II, the onset is in late childhood or adolescence, with progressive dementia, coarsened features, dysostosis multiplex, and skin changes; angiokeratomas similar to those seen in Fabry disease are present (Fig. 190-4).

FIGURE 190-4. Typical angiokeratoma skin lesions in a boy with fucosidosis. The lesions are similar to those seen in Fabry disease.

ASPARTYLGLYCOSAMINURIA Aspartylglycosaminuria is a rare disorder associated with a deficiency of N-aspartyl-b-glucosaminidase.45 The clinical features include mental retardation, behavioral difficulties, coarse facies, thickened and sagging skin, crystalline lens opacities, and a slow course, with survival well into adulthood. The gene is common among the people of Finland, where this otherwise extremely rare disorder represents the third most common genetic cause of mental retardation.48 CARBOHYDRATE-DEFICIENT GLYCOPROTEIN SYNDROME Carbohydrate-deficient glycoprotein syndrome is a rare (though probably underdiagnosed) disorder for which the exact biochemical and molecular defects are not yet known. Clinical features include progressive psychomotor retardation, lipocutaneous abnormalities, liver dysfunction, retinal degeneration, episodic stupor or stroke-like events, and cerebellar hypoplasia.49 I-CELL DISEASE (MUCOLIPIDOSIS II) I-cell disease is associated with a deficiency of N-acetylglu-cosamine 1-phosphotransferase, an enzyme normally involved in the phosphorylation of mannose to newly synthesized lysosomal enzymes.47,49,51,52 and 53 The result is that several lysosomal enzymes are deficient functionally because they have not undergone normal posttranslational modification. The incompletely modified enzymes are secreted rather than incorporated into lysosomes, and this causes a state of deficiency of multiple lysosomal enzymes. Symptoms arise during the first year of life, and the infant shows coarse features, dysostosis multiplex, joint contractures, gingival hyperplasia, and progressive psychomotor retardation (Fig. 190-5). Death usually occurs by the age of 5 years. The disorder is named for the characteristic inclusions seen on phase-contrast microscopy of cultured fibroblasts (inclusion cells).

FIGURE 190-5. Eleven-month-old girl with I-cell disease (mucolipidosis II). Note the periorbital fullness and coarsened facial features, which, unlike those seen in the mucopolysaccharidoses, are evident during early infancy.

PSEUDO–HURLER POLYDYSTROPHY (MUCOLIPIDOSIS III) The biochemical abnormality in pseudo-Hurler polydystrophy (mucolipidosis III) is similar to that in I-cell disease, but the clinical findings are different.52 Affected individuals develop joint stiffness by the age of 4 to 5 years, with dysostosis multiplex, coarsened facial features, corneal clouding, and aortic or mitral valve disease. Intelligence may be normal or mildly impaired. MUCOLIPIDOSIS IV The precise nature of the biochemical abnormality in mucolipidosis IV is unclear.54 Affected children have corneal clouding and progressive psychomotor retardation. The diagnosis can be made by identification of the characteristic cytoplasmic inclusions in biopsy samples of the conjunctiva or the skin and also in histiocytes of aspirated bone marrow.

DIAGNOSTIC CONSIDERATIONS There is no simple, practical approach to the diagnostic evaluation of children in whom disorders of lysosomal metabolism are suspected. The laboratory evaluation of lysosomal enzyme function is tedious, expensive, and not widely available.55 Therefore, a rational approach to these disorders involves the judicious use of diagnostic tests.5 The tests frequently used in the diagnosis of lysosomal storage disorders are skeletal radiography (dysostosis multiplex), nerve conduction velocity testing (peripheral neuropathy), ophthalmologic evaluation (corneal clouding, macular changes), neuropsychological testing (psycho-motor retardation), urine mucopolysaccharide measurement56 (screening examination for several of the MPS disorders), urine oligosaccharide measurement (screening examination for several of the mucolipidoses), electron microscopy of biopsy samples (of skin, conjunctiva, rectal mucosa, liver), and specific enzyme assays (of leukocytes or skin fibroblasts). The diagnostic evaluation must begin with a careful history and physical examination by an examiner experienced with these disorders. Using the historical and physical findings, a differential diagnosis usually can be attained based on the major features of the disorder, including the age of onset of symptoms and the cardinal clinical features (Table 190-2). Frequently, and especially in children with progressive neurologic deterioration, nonlysosomal disorders also must be considered in the differential diagnosis. Screening tests then can be used to narrow the scope of diagnostic possibilities (e.g., urine screening tests for mucopolysaccharide excretion), and other procedures, such as a thorough ophthalmologic examination and radiographic evaluation, may be indicated. A particularly useful screening test is electron microscopic examination of biopsy material from skin, conjunctiva, or rectal mucosa for evidence of intralysosomal storage.

TABLE 190-2. Diagnostic Considerations According to Symptom Complex

A definitive diagnosis rests on the laboratory demonstration of deficient activity of the specific lysosomal enzyme. This type of study should be performed only in qualified laboratories, where experience and careful controls enhance the validity of the findings. The diagnostician always must keep in mind that patients with these disorders often have unique clinical and biochemical findings. Finally, the use of molecular genetic analysis is becoming increasingly available to identify specific genetic mutations in many of the lysosomal disorders, and this can be a useful clinical tool, particularly when specific mutations correlate with clinical symptoms.

MANAGEMENT ISSUES A common misconception is that patients with “incurable” disorders such as the lysosomal storage disorders require no therapy. These patients present management problems that are among the most challenging in all of medical practice. Providing the parents of these affected children with the appropriate guidance, prognostic information, and emotional support to help them through what frequently is a lengthy and difficult struggle is a tremendous undertaking. Issues that are crucial to the care of patients with these disorders include control and management of seizures, feeding difficulties and nutritional disorders, hydrocephalus, joint contractures, pneumonia, urinary tract infections, hematologic abnormalities, orthopedic complications, spinal cord compression, peripheral nerve entrapment, chronic pain, visual and auditory impairment, sleep apnea, cardiovascular disease, cerebrovascular disease, renal failure, and hepatic failure, just to cite a few. The accomplishment of effective enzyme-replacement therapy for Gaucher disease is encouraging and may be a model for the future development of similar therapeutic interventions for the other lysosomal disorders.57,58 The greatest challenge in the development of such interventions is how to deliver replacement enzyme across the blood–brain barrier to the cells of the brain, where most of the deleterious effects of enzyme deficiency occur. Bone marrow transplantation remains an alternative mode of therapy for many of these disorders, but the morbidity and mortality associated with this procedure make patient selection and gous carrier state is possible for some of the disorders, usually by the demonstration of specific enzyme activity intermediate timing difficult.59 Ultimately, the ideal mode of therapy for lysosomal storage disease will come with the advent of successful human gene transplantation.60 GENETIC ISSUES The genetic aspects of the lysosomal storage disorders are extremely complex. The molecular defects are being elucidated by molecular genetic technology, which promises potentially curative approaches. With the exceptions of Hunter disease and Fabry disease (which are X-linked recessive), all of these disorders are heritable autosomal recessive conditions, and many have a characteristic ethnic predilection.61 Clinically, these facts have several important implications. First, accurate diagnosis is crucial for accurate genetic counseling, which usually entails informing the parents of affected children that their risk for recurrence of the disorder is 25% with each subsequent pregnancy (see Chap. 187). Second, prenatal diagnosis is available for almost all of these disorders,62 usually by specific enzyme assay of cultured amniocytes from amniocentesis or from chorionic villus biopsy specimens (Table 190-3). Third, detection of the heterozy-between the normal and the affected homozygote ranges. This often is useful for family members at risk for these disorders to determine who should undergo prenatal diagnostic studies, and has been used for mass screening programs for populations at risk, such as screening of the Ashkenazi Jewish population for Tay-Sachs disease. Finally, the new molecular genetic technology soon should allow the transplantation of normal genes into affected individuals.

TABLE 190-3. Genetic Aspects of Lysosomal Storage Disorders

CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62.

Rosenberg RN, Prusiner SB, DiMauro S, et al., eds. The molecular and genetic basis of neurological disease. Boston: Butterworth-Heinemann, 1993. Adams R, Lyon G, Kolodny EH. Neurology of inherited metabolic diseases of children. New York: McGraw-Hill, 1996. Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Kolodny EH, Cable WL. Inborn errors of metabolism. Ann Neurol 1982; 11:221. Lowden JA. Approaches to the diagnosis and management of infants and children with lysosomal storage disease. In: Kaback MM, ed. Genetic issues in pediatric and obstetric practice. Chicago: Year Book Medical Publishers, 1981:267. Beaudet AL, Scriver CR, Sly WS, et al. Genetics, biochemistry, and molecular basis of variant human phenotypes. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Suzuki K. Gangliosides and disease: a review. Adv Exp Med Biol 1984; 174:407. O'Brien JS, Ho MW, Veath ML, et al. Juvenile GM1-gangliosidosis: clinical, pathological, chemical, and enzymatic studies. Clin Genet 1972; 3:411. Okada S, O'Brien JS. Generalized gangliosidosis: beta galactosidase deficiency. Science 1968; 160:1002. Suzuki Y, Sakuraba H, Oshima A. b-Galactosidase deficiency: GM1 gangliosidosis and Morquio B disease. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Chen CY. Neuroimaging findings in late infantile GM1 gangliosidosis. Am J Neuroradiol 1998; 19(9):1628. Gravel RA, Clarke JTR, Kaback MM, et al. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Sandhoff K, Kolter T. Biochemistry of glycosphingolipid degradation. Clin Chim Acta 1997; 266:51. Johnson WG. The clinical spectrum of hexosaminidase deficiency states. Neurology 1981; 31:1453. Petroulakis E, Cao Z, Clarke JTR, et al. W474C amino acid substitution affects early processing of the alpha-subunit of beta-hexosaminidase A and is associated with subacute G(M2) gangliosidosis. Hum Mutat 1998; 11:432. Johnson WG, Wigger HJ, Karp HR, et al. Juvenile spinal muscular atrophy: a new hexosaminidase deficiency phenotype. Ann Neurol 1982; 11:11. Winkelman MD, Banker BQ, Victor M, Moser HW. Non-infantile Gaucher's disease: a clinicopathologic study. Neurology 1983; 33:994. Beutler E, Grabowski GA. Gaucher disease. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Barton NW, Brady RO, Dambrosia JM, et al. Replacement therapy for inherited enzyme deficiency—macrophage-targeted glucocerebrosidase for Gaucher's disease. N Engl J Med 1991; 324:1464. Crocker AC, Farber S. Niemann-Pick disease: a review of eighteen patients. Medicine (Baltimore) 1958; 37:1. Yan-Go FL, Yanagihara T, Pierre RV, Goldstein NP. A progressive neurologic disorder with supranuclear vertical gaze paresis and distinctive bone marrow cells. Mayo Clin Proc 1984; 59:404. Schuchman EH, Desnick RJ. Niemann-Pick disease types A and B: acid sphingomyelinase deficiencies. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Challamel MJ, Mazzola ME, Nevsimalova S, et al. Narcolepsy in children. Sleep 1994; 17:17. Pentchev PG, Vanier MT, Suzuki K, Patternson MC. Niemann-Pick disease type C: a cellular cholesterol lipidosis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Farrell DF, Swedberg K. Clinical and biochemical heterogeneity of globoid cell leukodystrophy. Ann Neurol 1981; 10:364. Suzuki Y, Suzuki K. Krabbe's globoid cell leukodystrophy: deficiency of galactocerebrosidase in serum, leukocytes, and fibroblasts. Science 1971; 171:73. Suzuki K, Suzuki Y. Galactosylceramide lipidosis: globoid cell leukodystrophy (Krabbe's disease). In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Kihara H. Genetic heterogeneity in metachromatic leukodystrophy. Am J Hum Genet 1982; 34:171. Farooqui AA, Horrocks LA. Biochemical aspects of globoid and metachromatic leukodystrophies. Neurochem Pathol 1984; 2:189. Hahn AF, Gordon BA, Hinton GG, Gilbert JJ. A variant form of metachromatic leukodystrophy without arylsulfatase deficiency. Ann Neurol 1982; 12:33. Kolodny EH. Metachromatic leukodystrophy and multiple sulfatase deficiency: sulfatide lipidosis. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Desnick RJ, Ioannou YA, Eng CM. a-Galactosidase A deficiency: Fabry disease. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Menkes DL. Images in neurology. The cutaneous stigmata of Fabry disease: an X-linked phakomatosis associated with central and peripheral nervous system dysfunction. Arch Neurol 1999; 56(4):487. Donati D, Novario R, Gastaldi L. Natural history and treatment of uremia secondary to Fabry's disease: an European experience. Nephron 1987; 46:353. Dorfman A, Matalon R. The mucopolysaccharidoses: a review. Proc Natl Acad Sci U S A 1976; 73:630. Kelly TE. The mucopolysaccharidoses and mucolipidoses. Clin Orthop 1976; 114:116. McKusick VA, Pyeritz RE. Genetic heterogeneity and allelic variation in the mucopolysaccharidoses. Johns Hopkins Med J 1980; 146:71. Neufeld EF. The biochemical basis of mucopolysaccharidoses and mucolipidoses. Prog Med Genet 1974; 10:81. Neufeld EF, Muenzer J. The mucopolysaccharidoses. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Van Schrojensteinde Valk HM, van der Kamp JJ. Follow-up on seven adult patients with mild Sanfilippo B-disease. Prenat Diagn 1987; 7:603. Wraith JE, Danks DM, Rogers JG. Mild Sanfilippo syndrome: a further cause of hyperactivity and behavioral disturbance. Med J Aust 1987; 147:450. Colville GA. Sleep problems in children with Sanfilippo syndrome. Dev Med Child Neurol 1996; 38(6):538. Sataloff RT, Schiebel BR, Spiegel JR. Morquio's syndrome. Am J Otol 1987; 8:443. O'Brien JS, Warner TG. Sialidosis: delineation of subtypes by neuraminidase assay. Clin Genet 1980; 17:35. Thomas GH, Beaudet AL. Disorders of glycoprotein degradation: In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Rapin I, Goldfischer S, Katzman R, et al. The cherry-red-spot–myoclonus syndrome. Ann Neurol 1978; 3:234. Warner TG, O'Brien JS. Genetic defects in glycoprotein metabolism. Annu Rev Genet 1983; 17:395. Mononen T, Mononen I, Matilainen R, Airaksinen E. High prevalence of aspartylglycosaminuria among school-age children in eastern Finland. Hum Genet 1991; 87:266. Jaeken J, Hagberg B, Stromme P. Clinical presentation and natural course of the carbohydrate-deficient glycoprotein syndrome. Acta Paediatr Scand Suppl 1991; 375:6. Whelan DT, Chang PL, Cockshott PW. Mucolipidosis II: the clinical, radiological and biochemical features in three cases. Clin Genet 1983; 24:90. Leroy LG, Spranger JW, Feingold M, et al. I-cell disease: a clinical picture. J Pediatr 1971; 79:360. Kornfeld S, Sly WS. I-cell disease and pseudo-Hurler polydystrophy: disorders of lysosomal enzyme phosphorylation and localization. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995. Ben-Yoseph Y, Mitchell DA, Nadler HL. First trimester prenatal evaluation for I-cell disease by N-acetyl-glucosamine 1-phosphotransferase assay. Clin Genet 1988; 33:38. Crandall BF, Phillipart M, Brown WJ, Bluestone DA. Mucolipidosis IV. Am J Med Genet 1982; 12:301. Takahashi Y, Orii T. Diagnosis of subtypes of GM1 gangliosidosis in vitro and in vivo—using urinary oligosaccharides as substrates. Clin Chim Acta 1989; 179:219. Whitley CB, Ridnour MD, Draper KA, et al. Diagnostic test for muco-polysaccharidosis. I. Direct method for quantifying excessive urinary glycosaminoglycan excretion. Clin Chem 1989; 35:374. Brady RO, Barton NW. Enzyme replacement therapy for type 1 Gaucher disease. In: Desnick, ed. Treatment of genetic disease. New York: Churchill Livingstone, 1991:153. Barton NW, Brady RO, Dambrosia JM, et al. Replacement therapy for inherited enzyme deficiency—macrophage-targeted glucocerebrosidase for Gaucher's disease. N Engl J Med 1991; 324:1464. Barranger JA. Marrow transplantation in genetic disease. N Engl J Med 1984; 311:1629. Hoogerbrugge PM, Valerio D. Bone marrow transplantation and gene therapy for lysosomal storage diseases. Bone Marrow Transplant 1998; 21 (Suppl 2):S34. Zlotogora J, Zeigler M, Bach G. Selection in favor of lysosomal storage disorders? Am J Hum Genet 1988; 42:271. D'Alton ME, Deherney AH. Prenatal diagnosis. N Engl J Med 1993; 328:114.

CHAPTER 191 HERITABLE DISEASES OF AMINO-ACID METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 191 HERITABLE DISEASES OF AMINO-ACID METABOLISM HARVEY J. STERN AND JAMES D. FINKELSTEIN Enzyme Defects General Considerations Clinical Sequelae Therapeutic Strategies Specific Disorders Hyperphenylalaninemia Tyrosinemia Alkaptonuria Albinism Homocystinurias Other Disorders of Sulfur Amino Acid Metabolism Disorders of Glutathione Metabolism Disorders of the Urea Cycle Disorders of Branched-Chain Amino-Acid Metabolism Short-Chain Organic Acidemias Transport Defects Storage Diseases Diagnosis of Inherited Disorders of Amino-Acid Metabolism Chapter References

The disorders of amino-acid metabolism are the most classic of the inherited metabolic disorders and have provided great insight into the pathogenesis, biochemistry, and treatment of human genetic disease. Sir Archibald Garrod, at the turn of the century, studied patients and families with alkaptonuria, cystinuria, pentosuria, and albinism.1 From his studies of these amino-acid disorders, Garrod developed the conceptual framework for what he called “inborn errors of metabolism.” Based on these observations, the one gene–one enzyme concept was proposed in the 1940s,2 and shortly thereafter, the first enzyme defect in humans was described (methemoglobinemia). In 1949 Pauling and coworkers3 demonstrated that mutations produced an alteration in the primary amino-acid sequence of a protein. This represented the beginning of the molecular age of human genetic research, which has evolved into the powerful and sophisticated methodologies of molecular biology and DNA-based diagnosis of human genetic diseases. With these discoveries has come an increasing appreciation of the molecular heterogeneity of these disorders, and the way in which various alterations of the amino-acid sequence of enzymatic or nonenzymatic proteins lead to major differences in the biochemical and clinical expression of disease in the patient.4 The elucidation and understanding of amino-acid disorders have paralleled the development of new laboratory methodologies to study the amino-acid composition of complex biologic fluids. Paper chromatography of urine followed by ninhydrin staining was the first technique to be widely used in the identification of patients with abnormal patterns of amino-acid excretion. Early column chromatography methods coupled to quantitative photometric ninhydrin analysis have been replaced by automated ion-exchange chromatography with computer-assisted data acquisition and calculation of results. Commercially available amino-acid analyzers can perform a complete physiologic analysis of plasma, urine, or cerebrospinal fluid in 120 µmol/L and the urinary excretion of metabolites of this amino acid. The enzymatic impairment appears to be most complete in type I, which is classic phenylketonuria (PKU).8,9 This disorder occurs in ~1 of every 10,000 births, and 2% of the population may be asymptomatic carriers. The dietary restriction of phenylalanine during the neonatal period is essential to prevent mental retardation in affected infants. The success of this therapy mandates the screening of all newborn infants, but testing can be useful only after the initiation of protein intake, for only then does the defect become manifest. The dietary restriction should be continued as long as possible, although maintenance of a protein-restricted diet becomes more difficult as the child enters adolescence. A prudent course is to maintain strict control through the first decade and then to periodically monitor the blood phenyl-alanine concentration as dietary intake is increased with the introduction of new foods. Note should be made, however, that multiple studies have shown a decline in school performance for older patients with PKU who go off a restricted diet. The successful management of children with PKU has led to a new problem: the management of pregnancies of phenylketonuric women. During pregnancy, phenylalanine restriction is necessary to prevent mental retardation, congenital heart disease, and other birth defects in the offspring.10 TYROSINEMIA Two defects cause the accumulation of tyrosine. A primary abnormality in tyrosine aminotransferase occurs in tyrosinemia type II (Richner-Hanhart syndrome). The clinical features include corneal ulcers, palmar and plantar keratoses, and possibly mental retardation, and are thought to derive from the intracellular crystallization of the amino acid. The primary defect in tyrosinemia type I (tyrosinosis) is a deficiency of fumarylacetoacetate hydrolase (type Ia) or maleylacetoacetate hydrolase (type Ib). Succiny-lacetone, a metabolite that accumulates, inhibits both tyrosine aminotransferase and porphobilinogen synthase. Moreover, succinylacetone impairs renal tubular function and methionine metabolism, resulting in a clinical syndrome that includes tyrosinemia, hypermethioninemia, liver cirrhosis, porphyria, and renal Fanconi syndrome. Primary hepatocellular carcinoma frequently occurs in survivors. Dietary restriction of tyrosine and phenylalanine, which is effective in type II tyrosinemia,11 may not be sufficient therapy for tyrosinosis. Treatment with an inhibitor of 4-hydroxyphenylpyruvate dioxygenase (which leads to decreased production of succinylacetone) offers promising results.12 Liver transplantation can be curative if performed before the development of hepatocellular carcinoma. ALKAPTONURIA Although alkaptonuria is extremely rare, it is the disorder that led Garrod in 1902 to conceptualize the inborn errors of metabolism. The primary defect, deficiency of homogentisic acid oxidase, leads to the accumulation of that substrate and its derivatives in connective tissue. The resultant blue discoloration and arthritis characterize the condition, which is called ochronosis.13 Intervertebral disc calcification is typical (Fig. 191-2). The oxidation of excreted homogentisic acid may cause a black discoloration of the urine, which is intensified by either exposure to air or alkalinization. No effective therapy for ochronosis is known. The gene for homogentisate 1,2-dioxygenase has been cloned and is located on human chromosome band 3q21-q2312. This raises the possibility of using gene therapy in the future to replace this missing enzymatic function.14

FIGURE 191-2. Alkaptonuria. Degeneration of intervertebral discs of lumbar spine is apparent; the joint spaces are narrowed, and the discs are calcified. Osteophyte formation is only moderate, and intervertebral ligaments are not calcified.

ALBINISM The group of genetic diseases making up albinism includes the ten or more variants of oculocutaneous albinism (OCA) as well as the several forms of ocular albinism. The common characteristic is a derangement in melanin metabolism that leads to abnormal pigmentation of the eyes (in ocular albinism), or of the eyes, skin, and hair (in OCA). Although previously the classification of various forms of oculocutaneous albinism had been based on clinical findings, information regarding the fundamental molecular defects in the production and transport of melanin is now beginning to emerge.15 The aggregate frequency of occurrence for this group of disorders approximates 1 in 10,000 among blacks and 1 in 20,000 among whites. The absence of retinal pigmentation causes loss of acuity as well as nystagmus and photophobia. Most individuals with albinism lack binocular vision secondary to a failure of development of the normal decussation pattern of the optic nerve. In affected individuals, the fibers from the temporal retina cross at the optic chiasm instead of continuing on the same side as the eye of origin. The variants of OCA differ in the degree of hypopigmentation of the hair and skin. Some of the syndromes have additional, unique abnormalities. The Hermansky-Pudlak form of OCA includes a mild coagulopathy secondary to platelet abnormalities,16 the accumulation of storage material in the reticuloendothelial system and gastrointestinal mucosa, and the development of restrictive lung disease. In Chédiak-Higashi syndrome, another variant of OCA, leukocytes contain giant peroxidase-positive granules that seem to impair chemotaxis. This finding may explain the extreme vulnerability of these patients to severe, often lethal, infections as well as the subsequent development of lymphoreticular neoplasms. HOMOCYSTINURIAS Homocysteine is a sulfhydryl amino acid formed by the demethylation of methionine. It is metabolized by either remethylation to methionine (with betaine or methyltetrahydrofolate as the methyl donor) or by the irreversible synthesis of cystathionine (transsulfuration).17 Impairment of either pathway leads to the accumulation of homocysteine in tissues and blood, together with excretion of the amino acid in the urine.18 Normal plasma contains homocysteine (the free thiol) and its disulfide deriva-tives—homocystine, homocysteine-cysteine mixed disulfide, and protein-homocysteine mixed disulfide. The latter is the predominant form.19 The sum of all forms is termed tHcy, and the upper limit of normal ranges from 12 to 15 µmol/L—depending on the methodology and the laboratory. In addition to homocysteine accumulation, a defect in cystathionine-b-synthase (classic homocystinuria) is associated with the augmented resynthesis of methionine. Conversely, the impairment of methyltetra-hydrofolate-homocysteine methyltransferase results in decreased levels of methionine and variable increases in the concentration of cystathionine. These differences, which are apparent on amino acid analysis of plasma, allow for the discrimination between the two classes of homocystinuria.18 The classic clinical expression of cystathionine synthase deficiency includes a marfanoid habitus (dolichostenomelia with pectus deformities), subluxation of the ocular lenses, osteoporosis, and thromboembolic episodes that often involve the major vessels.18 Mental retardation is variable but may be constant within a pedigree. Response to therapy defines at least two major variants of this disorder. Large doses of pyridoxine normalize the amino acid concentrations and prevent the medical complications in approximately one-half of patients. The remaining patients are treated by dietary manipulations that reduce the accumulation of homocysteine. The program consists of methionine restriction and cyst(e)ine supplementation. Folic acid and betaine, the methyl donors for the two homocysteine methyltransferases, often are provided to augment homocysteine methylation. Both the success of these therapies and the fact that the untreated disease may become symptomatic after childhood make cystathionine synthase deficiency a clinical concern for adult medicine.18,20 Failure of homocysteine methylation results from impairment of the methyltetrahydrofolate-homocysteine methyltransferase reaction. This reaction is vulnerable to several genetic defects. Usually, the disorder derives from a defect in the synthesis of the cosubstrate, methyltetrahydrofolate,21 or the coenzyme, methyl-cobalamin.22 Patients with severe methylenetetrahydrofolate reductase deficiency may have vascular pathology and a variable spectrum of neuropsychiatric dysfunction. The clinical chemical findings in patients with defects in cobalamin absorption or metabolism are more complex, because the synthesis of adenosyl-cobalamin also may be limited. Thus, methylmalonic aciduria may accompany the homocystinuria. Furthermore, megaloblastic anemia with pancytopenia may occur in some patients. The aim of therapy in the remethylation defects is to reduce the concentration of homocysteine while increasing that of methionine. Some patients with deranged cobalamin metabolism respond to large doses of vitamin B12; however, the effectiveness of folic acid in methylenetetrahydrofolate reductase deficiency is unproved. Betaine appears helpful in both conditions. Due to the clinical observation that patients with classical homocystinuria (with plasma tHcy >100 µmol/L) have early onset thrombovascular disease, clinical studies were initiated to evaluate the clinical consequences of “hyperhomocysteinemia”—a more moderate elevation in plasma homocysteine concentration (usually in the range of 15–30 µmol/L). Although many reports identify increased plasma tHcy as an independent risk factor for vascular disease, the issue of causality remains unresolved.23a,23b Until that relationship is clarified, there can be no consensus regarding the routine laboratory determination of plasma tHcy.24a Likewise, there is no agreement concerning the value of increasing folate intake, either by food fortification or dietary supplementation, in order to prevent thrombovascular disease, although such treatment does decrease plasma tHcy in most subjects. Lastly, the identification of the causes of hyperho-mocysteinemia remains incomplete. Nutritional deficiencies, particularly of folate and cobalamin, are the basis for a large number of cases. The search for genetic factors has been less rewarding. Heterozygosity for cystathionine synthase deficiency does not appear to be a significant etiology. In contrast, homozygosity for an abnormal allele of methylenetetrahydro-folate reductase results in a thermolabile enzyme form that has increased vulnerability to folate deficiency. OTHER DISORDERS OF SULFUR AMINO ACID METABOLISM Both cystathioninuria and hypermethioninemia may be benign metabolic defects.18 The former derives from a deficiency in cystathionase. The accumulated cystathionine appears to be without toxicity, and the only consequence may be impairment in the conversion of methionine to cysteine. Dietary supplements of the latter amino acid may be appropriate, particularly during the neonatal period. Cystathioninuria is one of the pyridoxine-responsive defects, and treatment with this vitamin often reverses the biochemical abnormalities. Hypermethioninemia results from a deficiency in the high-Km isoenzyme of methionine adenosyltransferase (MAT I/III), the enzyme that catalyzes the synthesis of S-adenosylmethionine (AdoMet) from methionine. Because the affected isoenzyme occurs primarily in the liver, the synthesis of AdoMet in other tissues should be unimpaired. However, the capacity for the catabolism of excessive quantities of methionine is limited. The absence of significant pathologic sequelae despite the accumulation of methionine suggests that methionine toxicity requires AdoMet synthesis. Several animal studies support this hypothesis. However, a report of CNS demyelination in a limited number of patients requires explanation.24b DISORDERS OF GLUTATHIONE METABOLISM Glutathione, the tripeptide g-glutamyl-cysteinyl-glycine, is the principal intracellular thiol and appears to be essential for the maintenance of other sulfhydryl compounds. This compound participates in the detoxification of peroxides and other chemical agents, and studies indicate another role for glutathione and g-glutamylcysteine in the membrane transport of amino acids. These physiologic functions explain the clinical consequences of the disorders of glutathione metabolism.25 Defects that cause glutathione deficiency are associated with erythrocyte hemolysis and central nervous system dysfunction. These defects include both glutathione synthase deficiency and g-glutamyl-cysteine synthase deficiency. The latter, which impairs the synthesis of both transport glutamyl peptides, leads to a generalized aminoaciduria. The 5-oxoproline (pyroglutamic acid) excretion in glutathione synthase deficiency results from an overproduction of 5-oxoproline resulting from the failure of feedback inhibition. The effect of the accumulation of glutathione is uncertain; possibly the mental retardation reported in patients with g-glutamyltranspeptidase deficiency is attributable to an ascertainment bias. DISORDERS OF THE UREA CYCLE The function of the urea cycle is to dispose of waste nitrogen and synthesize arginine. The early diagnosis of the diseases resulting from disordered urea synthesis is essential to prevent the potentially irreversible changes in the central nervous system that can result from recurrent or chronic hyperammonemia. Patients may present in the neonatal period with lethargy, poor feeding, respiratory alkalosis, and vomiting, and rapidly progress to hyperammonemic coma, which is often fatal. Survivors are usually left with significant cognitive and motor deficits. In other cases, depending on the degree of enzyme deficiency and nitrogen load, clinical symptoms may not develop until adulthood. Women who carry the X-linked disorder ornithine transcarba-moylase deficiency may develop significant hyperammonemia during periods of

metabolic stress, including childbirth. The differential diagnosis of urea cycle disorders includes transient hyperammonemia of the newborn, a relatively benign, transient, neonatal hyperammonemia; acquired hepatic diseases; and the other inherited disorders listed in Table 191-1. Treatment by protein restriction may cause a deficiency of essential amino acids. Thus, a controlled diet containing the ketoacid analogs of the essential amino acids is preferable. Encouraging results have been reported for treatment with benzoic acid, sodium phenyl-butyrate, or sodium phenylacetate.26 Benzoate is conjugated with glycine to form hippurate, which is excreted by the kidney, and phenylacetate combines with glutamine. This process removes excess amino groups and reduces the hyperammonemia. Hemodialysis is usually required in the acutely ill newborn. Prenatal diagnosis of urea cycle disorders is now performed by DNA analysis of the affected enzyme. Biochemical analysis is not possible because urea cycle enzymes are not expressed in amniocytes or chorionic villous tissue. DISORDERS OF BRANCHED-CHAIN AMINO-ACID METABOLISM The three neutral branched-chain amino acids—leucine, isoleucine, and valine—share a common pathway for initial catabolism. The sequence is transamination, oxidative decarboxylation of the ketoacid derivative, and dehydrogenation of the resultant acylcoenzyme A (CoA) metabolites. As shown in Table 191-1, the reported biochemical defects in these pathways may be specific to one of the branched-chain amino acids or common to all three (maple syrup urine disease). Ketoacidosis and hypoglycemia are characteristic of several of these diseases. Only defects in the transaminases and ketoacid decarboxylases cause the accumulation of the corresponding amino acids. This is not true of the more distal enzymatic lesions of branched-chain aminoacid metabolism (i.e., isovaleric acidemia). The diagnosis of these disorders is made by organic-acid analysis of the urine using gas chromatography/mass spectrometry rather than amino-acid analyses. Treatment involves consumption of diets deficient in branched-chain amino acids. SHORT-CHAIN ORGANIC ACIDEMIAS Propionyl-CoA is part of the final pathway common to the catabolism of multiple amino acids, sterols, and odd-chain fatty acids. The propionyl-CoA is carboxylated to form D-methylmalonyl-CoA, which, after racemization, is converted to succinyl-CoA. Recurrent metabolic acidosis, pancytopenia, growth failure, and mental retardation occur in patients with disorders of this pathway. Defects in either the synthesis of the apoenzyme or the metabolism of the biotin cofactor may impair propionyl-CoA carboxylase and produce propionic acidemia. In multiple carboxylase deficiency, the biotin cofactor deficiency is shared by other mitochondrial carboxylases, and abnormalities of pyruvate and of methylcrotonyl metabolism are apparent. Similarly, the defect in methylmalonyl-CoA mutase in methylmalonic acidemia may be secondary to faulty synthesis of either the apoenzyme or the cobalamin coenzyme. If the cobalamin defect involves the synthesis of both adenosylcobalamin, the coenzyme for methylmalonyl-CoA mutase, and methylcobalamin, homocystinuria accompanies the methylmalonic aciduria.

TRANSPORT DEFECTS Transport enzymatic defects result in an “overflow” type of aminoaciduria. The increased concentration of the amino acid in the glomerular filtrate exceeds the capacity for tubular reabsorption, resulting in aminoaciduria. Moreover, the urine may contain other amino acids that share and compete for the same tubular transport mechanism. In contrast, syndromes also are found that derive from primary defects in epithelial transport. Although multiple organs may be affected, impairment of renal and small intestine absorption are the most relevant to the clinician. The characteristic biochemical abnormality is an aminoaciduria in the presence of a normal concentration of the substrate in the blood. Intestinal malabsorption of the same substrate may occur. Because epithelial transport systems may be specific for one amino acid or common to a defined group of amino acids, the disorders may involve one or more compounds. Of the disorders listed in Table 191-2, only methionine malabsorption and tryptophan malabsorption are limited to the gastrointestinal tract; each of the others involves the kidney and has variants with intestinal involvement.

TABLE 191-2. Genetic Disorders of Specific Amino-Acid Transport Mechanisms

The nature and the magnitude of the defect determine the clinical consequences. Intestinal malabsorption may lead to nutrient deficiency, as in the pellagra seen in Hartnup disease. Alternatively, metabolism of the nonabsorbed substrate by the intestinal flora may generate absorbable toxins. The aminoaciduria is rarely sufficient to cause plasma deficiency. In cystinuria, however, the urinary concentration of cystine may exceed the solubility of the amino acid, and nephrolithiasis is the consequence (see Chap. 69). Cystinuria is a relatively common defect with an estimated incidence of 1 in 7000 live births. Three variants can be defined on the basis of the degree of intestinal involvement (limited in type III) and the presence of detectable aminoaciduria in obligate heterozygotes (absent in type I). Nephrolithiasis is the only significant pathology and often appears in the second or third decade of life. Because the intestinal absorption of cysteine and cysteine-containing peptides is not impaired, dietary restriction is not effective. Therapy relies on increasing both the volume and the alkalinity of the urine.27 Penicillamine, a thiol that forms more soluble disulfides with cysteine, may be tried in refractory cases. However, this drug can cause hepatic and renal dysfunction. Tiopronin, another thiol compound, appears to be less toxic.28

STORAGE DISEASES Cystinosis is the only known storage disorder of amino-acid metabolism.29 The estimated incidence of this autosomal recessive disease approximates 1 in 200,000 live births. The fundamental defect is a failure of the transport of cystine out of the lysosomes. Consequently, cystine accumulates in these organelles and impairs the function of multiple tissues. Nephropathic cystinosis is the most common and is the classic form of the disease. During the first decade of life, affected children show a marked retardation of growth without intellectual impairment. This effect may derive from the severe and progressive impairment of renal function. The initial abnormality is the failure of the tubular reabsorption of multiple solutes (glucose, amino acids, phosphate), together with both renal tubular acidosis and nephrogenic diabetes insipidus. Cystinosis is the most common identifiable cause of this Fanconi syndrome in children. Subsequently, glomerular damage begins to dominate the clinical picture, and renal failure usually is present by the end of the first decade, mandating transplantation or long-term dialysis to maintain life. The grafted kidney remains unaffected. Prolongation of life by renal transplantation has unmasked the more slowly occurring involvement of other organs. Cystine crystals in the cornea, conjunctivae, and irides (Fig. 191-3) are associated with severe photophobia and other ophthalmic symptoms. Hypothyroidism also is common. Hepatomegaly and splenomegaly occur in almost half of the patients.

FIGURE 191-3. Cystinosis. Slit-lamp examination demonstrates extensive crystal deposits in the cornea (C).

Two other variants of cystinosis are both rare. A late-onset or adolescent form becomes apparent at a later age but then follows the same course.30 Benign (adult) cystinosis spares the kidney, although cystine crystals occur in the cornea, conjunctivae, and bone marrow. In the past, therapy for cystinosis had consisted of the meticulous management of the consequences of the progressive renal dysfunction. The finding that exogenous cysteamine mobilizes intralysosomal cystine, however, has provided the basis for more specific therapy. Unfortunately, this treatment does not restore lost functions; therefore, cysteamine therapy may be of maximal value in the prevention of the progression of cystinosis. Thus, early diagnosis remains essential. Specific evaluations of the younger siblings of known patients and of children with renal Fanconi syndrome are mandatory. Plasma cystine levels are normal; the increased urinary cystine excretion is part of a generalized aminoaciduria. Conversely, both excessive leukocyte cystine and corneal cystine crystals (observed by slit-lamp evaluation) are pathognomonic. Chorionic villus sampling allows prenatal detection; however, the advisability of maternal cysteamine therapy remains unclear. Linkage analysis in families with cystinosis has localized the gene to the short arm of chromosome 17. Based on this information, a gene (CTNS) was identified of which the product (cystinosin) has the properties of a membrane transport protein. Mutation analysis of 108 nephropathic cystinosis patients found that 44% were homozygous for a 65-kb deletion of the cystinosis gene.31,31a

DIAGNOSIS OF INHERITED DISORDERS OF AMINO-ACID METABOLISM Many patients with inherited disorders of amino-acid metabolism are identified by newborn metabolic screening procedures, which are performed on blood spots obtained at the birth hospital. Currently, only a few states repeat the screen again at 2 to 3 weeks of age. The infant must have consumed a protein meal, and preferably at least two to three feedings, before the sample is obtained to ensure that elevated concentrations of amino acids will be detected. The trend of early discharge from the birth hospital, often within 24 hours of delivery, has led to the misdiagnosis of some affected newborns. For this reason, in cases of early hospital discharge, the recommendation is that newborn metabolic screening be repeated at the first pediatric visit to ensure identification of infants with amino-acid disorders. Most state newborn screening programs focus on diseases that have an initial asymptomatic period and for which an effective treatment exists to prevent irreversible pathology. PKU is the prototype disorder for which newborn screening was instituted, although most states also screen for tyrosinemia, maple syrup urine disease, biotinidase deficiency, and homocystinuria. A sibling of a patient with a known aminoacidopathy, in addition to undergoing newborn screening, should be evaluated by quantitative plasma amino-acid analysis as soon as protein-containing feedings are instituted. In addition to identification by newborn screening programs, many patients with inborn errors of amino-acid metabolism are diagnosed when the astute clinician suspects a biochemical disorder. Pediatricians are most frequently involved, because they see the young patient with unexplained ketoacidosis, failure to thrive, formula intolerance, or developmental delay. Because of the organ-specific conditions associated with many of these disorders (Table 191-3), however, diagnosis by other physicians is not uncommon. The ophthalmologist may be the first to evaluate a patient with homocystinuria or gyrate atrophy of the choroid and retina, whereas the dermatologist may see the patient with alkaptonuria, albinism, or Hartnup disorder. Hematologists must be aware of the metabolic syndromes associated with severe anemia, pancytopenia, and other hematologic abnormalities. The nephrologist must consider cystinuria in young patients with nephrolithiasis, and unexplained thromboembolism or premature arteriosclerosis should alert the cardiologist to the possibility of homocystinuria. The endocrinologist, in particular, must be at least superficially familiar with these models of disordered metabolism, as well as their clinical manifestations.

TABLE 191-3. Organ-Specific Involvement in Inherited Disorders of Amino-Acid Metabolism

Prenatal diagnosis is now possible for many disorders of amino-acid metabolism. Amniotic fluid may be directly analyzed for abnormal metabolites; alternatively, fetal tissue obtained by amniocentesis (at 15–17 weeks' gestation) or by chorionic villus sampling (at 10–12 weeks' gestation) can be used as source material for enzyme assays. The limited tissue distribution of many of the enzymes often restricts the applicability of enzymatic diagnosis. For example, phenylalanine hydroxylase (which is deficient in PKU) is not expressed in amniocytes or chorionic villus cells and is present only in liver. In some cases, diagnosis by fetal tissue sampling is theoretically possible but is associated with a significant risk of fetal loss. Many of these difficulties in prenatal testing are being overcome by the use of DNA-based diagnosis, because all cells with a nucleus share the same genetic characteristics. Another experimental approach involves obtaining and analyzing nucleated fetal red blood cells present in the maternal circulation. Laboratory services used to diagnose and monitor patients with inborn errors of amino-acid metabolism are available through many commercial laboratories and in academic medical centers. The development of dedicated amino-acid analyzers allows rapid and accurate determination of amino-acid concentrations of biologic fluids. Although many laboratories still use paper chromatography of urine samples as a screening procedure, the speed and accuracy of quantitative plasma amino-acid analysis makes it the procedure of choice, and such analysis is certainly indicated in any patient for whom there is a high suspicion of an aminoacidopathy. To avoid false-positive elevations due to the ingestion of protein, blood specimens should be obtained after a 12-hour fast in adults or after at least a 4- to 6-hour fast in young children. Quantitative amino-acid analysis of urine can also be performed, but the results are much more difficult to interpret than are those from plasma, owing to the much wider variation of normal urine amino-acid concentrations. Analysis of the cerebrospinal fluid glycine concentration is particularly useful in the diagnosis of nonketotic hyperglycinemia. The urine metabolic screen (metabolic urinalysis) consists of a series of rapid screening tests that often provide the first clue to the presence of a metabolic disorder (Table 191-4). The ferric chloride test is the best known of these and is useful in the diagnosis of PKU, tyrosinemia, histidinemia, and alkaptonuria. Several drugs (phenothiazines, salicylates), as well as the presence of conjugated bilirubin, produce a positive result with the ferric chloride reagent. The 2,4-dinitrophenylhydrazine test detects 2-oxoacids and is useful in the diagnosis of maple syrup urine disease. Nitrosonaphthol reacts with tyrosine and its metabolites (tyrosinemia), and cyanide-nitroprusside detects sulfur amino acids (cystinuria, homocystinuria).

TABLE 191-4. Urine Screening Tests for Amino-Acid Disorders

Many amino-acid disorders also produce disturbances in organic-acid excretion, because an early step in the catabolism of many amino acids is the removal of the amino group, which produces the corresponding organic acid. Analysis of urine organic-acid excretion by gas chromatography/mass spectrometry is an extremely sensitive and powerful tool that is widely available and, when used with plasma amino-acid analysis and urine metabolic screening, provides a comprehensive evaluation of the patient suspected of having an inborn error of metabolism. In some cases, the diagnosis of an amino-acid disorder can be confirmed by specific enzyme analysis in leukocytes, fibroblasts, or liver biopsy tissue. The limited tissue distribution and instability of some enzymes, however, and the accuracy of plasma amino-acid analysis in most cases make enzyme-based diagnosis unnecessary. New tests based on DNA technology may overcome the obstacles to enzyme-based analysis. Identification of persons with amino-acid disorders by a finding of abnormalities in routine laboratory studies such as a chemistry panel is uncommon. Diagnosis, therefore, requires suspicion on the part of the clinician as well as the expertise of special laboratories that have experience in performing these complex studies. Clear communication between the referring physician and laboratory director is essential to ensure that appropriate studies are performed. This avoids a “shotgun” approach to metabolic evaluation, which, because of the labor-intensive nature of these studies, could result in a high expense for unnecessary testing. Referral to a regional clinic for metabolic disease (often located at academic medical centers) is recommended for patients with unusual or complex disorders. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.

Garrod AE. Croonian Lecture I: inborn errors of metabolism. Lancet 1908; 2:1. Beadle GW. Genetic control of biochemical reactions. Harvey Lect 1945; 40:179. Pauling L, Itano HA, Singer SJ, Wells IC. Sickle cell anemia: a molecular disease. Science 1949; 11:543. Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 7th ed. New York: McGraw-Hill, 1995. Rashed MS, Ozand PT, Bucknall MP, Little D. Diagnosis of inborn errors of metabolism from blood spots by acylcarnitine and amino acid profiling using automated electrospray tandem mass spectrometry. Pediatr Res 1995; 38:324. Finkelstein JD. Methionine metabolism in mammals: the biochemical basis of homocystinuria. Metabolism 1974; 23:387. Scriver CR. Vitamin responsive inborn errors of metabolism. Metabolism 1973; 22:1319. Scriver CR, Kaufman S, Eisensmith JL, Woo SL. The hyperphenylalaninemias. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 7th ed. New York: McGraw-Hill, 1995:1015. Svensson E, Isellius L, Hagenfeldt L. Severity of mutation in the phenylalanine hydroxylase gene influences phenylalanine metabolism in phenylketonuria and hyperphenylalaninemia heterozygotes. J Inherit Metab Dis 1994; 17:215. Lenke RR, Levy HL. Maternal phenylketonuria and hyperphenylalaninemia. N Engl J Med 1980; 303:1202. Sammartino A, Carbella R, Cecio A, et al. The effect of diet on the ophthalmologic, clinical and biochemical aspects of Richner-Hanhart syndrome: a morphological ultrastructural study of the cornea and the conjunctiva. Int Ophthalmol 1987; 10:203. Lindstedt S, Holme E, Lock EA, et al. Treatment of hereditary tyrosinaemia type I by inhibition of 4-hydroxyphenylpyruvate dioxygenase. Lancet 1992; 340:813. Gaines JJ Jr. The pathology of alkaptonuric ochronosis. Hum Pathol 1989; 20:40. Fernandez-Canon JM, Granadino B, Beltran-Valero de Bernabe D, et al. The molecular basis of alkaptonuria. Nat Genet 1996; 14:19. Oetting WS, King RA. Molecular basis of albinism: mutations and polymorphisms of pigmentation genes associated with albinism. Hum Mutat 1999; 13:99. Witkop CJ, Kromwiede M, Sedano H, White JG. Reliability of absent platelet dense bodies as a diagnostic criterion for Hermansky-Pudlak syndrome. Am J Hematol 1987; 26:305. Finkelstein JD. Pathways and regulation of homocysteine metabolism in mammals. Semin Thromb Hemost 2000; 26:219. Mudd SH, Levy HL, Skovby F. Disorders of transsulfuration. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995:1279. Mudd SH, Finkelstein JD, Refsum H, et al. Homocysteine and its disulfide derivatives. A suggested consensus nomenclature. Arterioscler Throm Vasc Biol 2000; 20:1704. Yap S, Naughten ER, Wilcken B, et al. Vascular complications of severe hyperhomocysteinemia in patients with homocystinuria due to cystathionine beta-synthase deficiency: effects of homocysteine-lowering therapy. Semin Thromb Hemost 2000; 26:335. Rosenblatt DS. Inherited disorders of folate transport and metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995:3111. Fenton WA, Rosenberg LE. Inherited disorders of cobalamin transport and metabolism. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular bases of inherited disease. New York: McGraw-Hill, 1995:3129.

23a.Refsum H, Ueland PM, Nygard O, Vollset SE. Homocysteine and cardiovascular disease. Annu Rev Medicine 1998; 49:31. 23b.Brattstrom L, Wilcken DEL. Homocysteine and cardiovascular disease: cause or effect. Am J Clin Nutr 2000; 72:315. 24a.Malinow MR, Bostom AG, Krauss RM. Homocyst(e)ine, diet and cardiovascular diseases. A statement for healthcare professionals from the Nutrition Committee, American Heart Association. Circulation 1999; 99:178. 24b.Chamberlin ME, Ubagai T, Mudd SH, et al. Demyelination of the brain is associated with methionine adenosyltransferase I/III deficiency. J Clin Invest 1996; 98:1021. 25. 26. 27. 28. 29.

Meister A, Anderson ME. Glutathione. Annu Rev Biochem 1983; 52:711. Brusilow SA, Horwich AL. Urea cycle enzymes. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 7th ed. New York: McGraw-Hill, 1995:1187. Scriver CR. Cystinuria. N Engl J Med 1986; 315:1155. Tiopronin for cystinuria. Med Lett Drugs Ther 1989; 31(Jan 27):7. Gahl WA, Renlund M, Thoene JG. Lysosomal transport disorders: cystinosis and sialic acid storage disorders. In: Scriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic basis of inherited disease, 6th ed. New York: McGraw-Hill, 1989:2619. 30. Gahl WA, Schneider JA, Aula PP. Lysosomal transport disorders: cystinosis and sialic acid storage disorders. In Seriver CR, Beaudet AL, Sly WS, Valle D, eds. The metabolic and molecular basis of inherited disease. 7th edition. McGraw Hill, NY: 1995, 3763. 31. Shotelersuk V, Larson D, Anikster Y, et al. CTNS mutations in an American-based population of cystinosis patients. Am J Hum Genet 1998; 63:1352. 31a.Touchman JW, Anikster Y, Dietrich NL, et al. The genomic region encompassing the nephropathic cystinosis gene (CTNS). Genome Res 2000; 10:165.

CHAPTER 192 HERITABLE DISEASES OF PURINE METABOLISM Principles and Practice of Endocrinology and Metabolism

CHAPTER 192 HERITABLE DISEASES OF PURINE METABOLISM EDWARD W. HOLMES AND DAVID J. NASHEL General Considerations Signs and Symptoms Classification Disorders of Purine Synthesis Consequences of Purine Overproduction Control of Purine Synthesis Phosphoribosylpyrophosphate Synthetase Overactivity Hypoxanthine-Guanine Phosphoribosyltransferase Deficiency Management of Purine Overproduction Disorders of Purine Salvage Hypoxanthine-Guanine Phosphoribosyltransferase Deficiency Adenine Phosphoribosyl Transferase Deficiency Disorders of Purine Catabolism Adenosine Deaminase Deficiency Purine Nucleoside Phosphorylase Deficiency Xanthine Oxidase Deficiency Glucose-6-Phosphatase Deficiency Disorders of Purine Nucleotide Interconversion Adenosine Monophosphate Deaminase Deficiency Adenylosuccinate Lyase Deficiency Disorders of Uric Acid Excretion Hypouricemia Chapter References

GENERAL CONSIDERATIONS Disorders of purine metabolism, which may be inherited or acquired, affect many organ systems, and clinical symptoms differ depending on the disorder and the particular organ systems affected. This chapter focuses on inherited defects, but, where appropriate, acquired defects also are cited for comparison. For some disorders, such as gout, it is difficult to separate inherited abnormalities completely from environmental influences, and the interplay between genetics and environment must be considered. SIGNS AND SYMPTOMS Arthritis secondary to the deposition of monosodium urate crystals in synovial fluid is a prominent clinical manifestation of several inherited defects: hypoxanthine-guanine phosphoribosyltransferase (HPRT) deficiency, phosphoribosylpyrophosphate synthetase (PRPPS) overactivity, and glucose-6-phosphatase deficiency. Nephrolithiasis secondary to uric acid crystal formation in renal collecting structures also is a clinical feature of disorders such as HPRT deficiency and PRPPS overactivity, where uric acid is overproduced and excreted in the urine. Two other inherited defects, xanthine oxidase (XO) deficiency and adenine phosphoribosyl-transferase (APRT) deficiency, lead to excessive excretion of the relatively insoluble purine bases xanthine and 2,8-dihydroxyadenine, respectively. Renal colic also develops in patients having these enzyme defects. Neurologic symptoms, including spasticity, choreoathetosis, mental retardation, and self-mutilation, ensue from high-grade deficiency of HPRT (Lesch-Nyhan syndrome), and autism may result from adenylosuccinate (S-AMP) lyase deficiency. Immunodeficiency states resulting from combined T- and B-cell defects or isolated T-cell defects are consequences of deficiencies of adenosine deaminase (ADA) and purine nucleoside phosphorylase (PNP), respectively. Patients with these enzyme deficiencies are subject to recurrent infections from an early age. Metabolic myopathy symptoms, including easy fatigability, muscle cramps, and myalgias, result from a deficiency of adenylate deaminase. In most of these disorders, the enzyme defect is manifest in all tissues, and selective involvement of organ systems relates to the accumulation of metabolic precursors or end products of purine metabolism in particular organs. However, in at least one condition, adenylate deaminase deficiency, the enzyme deficiency is localized to a specific tissue, skeletal muscle. CLASSIFICATION Disorders of purine metabolism have been organized into categories of defects in purine synthesis, catabolism, interconversion, or excretion. Inspection of Figure 192-1 reveals numerous interconnections between the various components of the purine pathway. It follows that a primary defect in one component, such as purine synthesis, may lead to secondary changes in other components, such as purine catabolism and excretion. The clinical presentation is then influenced by the combined derangements in purine metabolism. For example, a person with a primary defect in purine synthesis develops hyperuricemia because of increased purine nucleotide production, with subsequent catabolism to uric acid. Hyperuricosuria is the result of increased uric acid production and excretion. Thus, gouty arthritis, a consequence of hyperuricemia, and renal calculi, a consequence of hyperuricosuria, reflect the secondary changes more than the primary derangement in purine metabolism. For other disorders, clinical manifestations may arise from a secondary effect on an entirely different metabolic system, such as a derangement in energy metabolism or DNA synthesis. Where known, interrelations between defects in purine metabolism and other pathways are pointed out; and to the extent possible, these derangements in metabolism are related to the symptoms and signs patients exhibit.

FIGURE 192-1. Disorders of purine metabolism. Defects are broken down into derangements in purine synthesis, interconversion, catabolism, and excretion. (PRPP, phosphoribosylpyrophosphate; IMP, inosine monophosphate; AMP, adenosine monophosphate; S-AMP, succinyl-AMP; GMP, guanosine monophosphate; Ado and dAdo, adenosine and deoxyadenosine, respectively; Guo and dGuo, guanosine and deoxyguanosine, respectively; Ino and dIno, inosine and deoxyinosine, respectively; Hx, hypoxanthine; Xan, xanthine; Ade, adenine; HPRT, hypoxanthine-guanine phosphoribosyltransferase; APRT, adenine phosphoribosyltransferase; PNP, purine nucleoside phosphorylase; lyase, adenylosuccinate lyase; AMP-D, AMP deaminase; ADA, adenosine deaminase; XO, xanthine oxidase; GTP, guanosine triphosphate; dGTP, deoxyguanosine triphosphate; ATP, adenosine triphosphate; dATP, deoxyadenosine triphosphate.)

DISORDERS OF PURINE SYNTHESIS CONSEQUENCES OF PURINE OVERPRODUCTION Excessive rates of purine synthesis lead to excessive rates of uric acid production because urate is normally the end product of purine metabolism in humans.1,2 Other mammals, except some species of New World apes, are able to oxidize urate to allantoin and other degradation products that are more soluble than urate and uric

acid. It is this poor solubility of urate and uric acid that leads to clinical problems in patients who overproduce purines. The precipitation of urate, predominantly as the sodium salt, in cartilage, synovial membranes, and synovial fluid is responsible for initiating an inflammatory response leading to the acute attacks of arthritis that are characteristic of gout (Fig. 192-2). The arthritis typically is monoarticular in the early stages of gout but later may become polyarticular. Any joint lined with synovium may be affected, but distal joints are affected more often, and joints of the lower extremities are affected more often than joints of the upper extremities. After an acute attack of gouty arthritis, symptoms usually resolve completely, and the patient enters an intercritical period in which there are no symptoms. This symptom-free interval may last for years.1,2

FIGURE 192-2. A, Acute attack of podagra (involvement of the big toe). Note the swollen area (arrows) and the shiny skin, which was warm to the touch. B, Chronic tophaceous gout affecting the first and fifth metatarsophalangeal joints. The bandage covers a draining tophaceous area. Initially, the patient had had recurrent attacks of pain of the metatarsophalangeal joint of the great toe (podagra).

Urate salt deposition is not limited to joints and periarticular structures, but also occurs in soft tissues throughout the body. These deposits, which usually are not inflammatory, are referred to as tophi3 (see Fig. 192-2). Subcutaneous tophi may ulcerate and drain urate crystals and, rarely, become secondarily infected. Tophi in the interstitium of the kidney may lead to fibrosis and mild renal insufficiency, a condition referred to as urate nephropathy. Urate salts are unlikely to precipitate from extracellular fluids unless their concentration exceeds the solubility limit. For sodium urate, the predominant urate salt in body fluids, the solubility limit is ~7 mg/dL in serum. The distinction between a normal and an elevated serum urate concentration is somewhat arbitrary, because the statistical definition of hyperuricemia depends on the method by which urate is quantitated, the ethnic background of the population used to determine the statistical limits, and the environment in which the test population lives. For example, enzymatic methods for determining serum urate give results that are ~1 mg/dL lower than the colorimetric methods used in many automated clinical laboratories. The statistical definition of a normal urate concentration is confounded by the nonnormal distribution of urate concentrations in most populations, by differences in urate concentration among different ethnic groups, and by the effects of diet and other environmental factors on urate metabolism. In the authors' opinion, the most reasonable definition of hyperuricemia is that value at which urate concentration exceeds its solubility limit in body fluids, possibly leading to urate deposition and clinical symptoms. As mentioned earlier, when the urate concentration is >7 mg/dL, as determined by the enzymatic method, serum is supersaturated, and the patient is at risk for urate precipitation. If serum urate is determined by a colorimetric method that yields a higher value, the serum concentration that is saturating for urate may be 1 mg/dL higher. This chemical definition of hyperuricemia provides a uniform benchmark for monitoring patients of diverse ethnic backgrounds in different environmental situations. Its clinical utility is borne out by its widespread use. Uric acid is much less soluble than its salt, but the pK of this acid (5.75) precludes its formation anywhere in the body except the urine. Patients who overproduce purines overexcrete uric acid in the urine; consequently, they are predisposed to the formation of uric acid calculi. Stone formation is exacerbated in patients who excrete a persistently acid urine, thereby favoring the formation of uric acid over that of urate. Patients who excrete low urine volumes, thereby increasing the concentration of uric acid, also are predisposed to stone formation. Rarely, uric acid excretion reaches extraordinary levels in patients with inherited defects in purine synthesis or in those with malignancies where accelerated cell turnover and nucleic acid catabolism lead to large increases in uric acid excretion. In these unusual situations, uric acid crystallizes in the collecting tubules, leading to acute renal failure. The latter condition is referred to as acute uric acid nephropathy to distinguish it from the urate nephropathy described earlier.4 Hyperuricosuria, or excessive uric acid excretion, is more difficult to define than hyperuricemia, because it depends on many factors, one of the most important being the purine content of the diet. On a purine-free diet, urate excretion normally is 1 g per day) with uric acid nephrolithiasis. The activity of this enzyme, which can be documented by assay of erythrocyte lysates, is matic defects, including Km mutants and feedback-resistant mutants, account for overactivity of this enzyme, and all share the property of increasing the intracellular levels of PRPP. This increase in PRPP concentration activates the first enzyme in the de novo pathway of purine synthesis (see Fig. 192-3), with an increase in purine nucleotide formation. Subsequent catabolism of the purine nucleotides increases uric acid formation and excretion (see Fig. 192-1). Increased PRPP synthetase activity is a good example of a primary defect in purine synthesis “pushing” the catabolic and excretory components of the pathway. HYPOXANTHINE-GUANINE PHOSPHORIBOSYLTRANSFERASE DEFICIENCY HPRT, another enzyme encoded by a gene on the X chromosome, catalyzes a significant reaction in the salvage pathway in which the purine bases, hypoxanthine and guanine, are phosphoribosylated to form purine nucleotides (see Fig. 192-1). This is one of the best studied of the human enzymatic deficiencies, including comprehensive metabolic studies in patients, biochemical analyses of cultured cells, and documentation of amino-acid changes in the HPRT protein and base substitutions in the HPRT gene.8 Two clinical syndromes are associated with this enzyme defect, which is inherited as an X-linked recessive disorder. High-grade deficiencies (0.1% residual activity) of HPRT are associated with early (teenage)-onset gout and uric acid crystal formation (infancy), and, occasionally, acute renal insufficiency.10 Neurologic symptoms either are absent or mild. Rarely, mothers of HPRT-deficient male children exhibit mild degrees of purine overproduction and late-onset gout, indicating that this disorder can be inherited as an incomplete X-linked dominant trait in unusual situations.

FIGURE 192-4. Lesch-Nyhan syndrome. Self-destructive behavior has resulted in injury to chin and bridge of nose and destruction of upper and lower lips.

FIGURE 192-5. Lesch-Nyhan syndrome. Choreoathetosis of upper extremities and scissoring of lower extremities are typical of the spasticity observed in this disorder.

HPRT deficiency can be diagnosed readily by assay of erythrocyte lysates. A deficiency of this enzyme has two important consequences for purine synthesis: First, PRPP accumulates because it is not used in this salvage reaction, and second, purine nucleotide formation through salvage mechanisms is reduced (see Fig. 192-1). Consequently, there is increased activity of the first enzyme in the de novo pathway of purine synthesis (see Fig. 192-3). (HPRT deficiency is also discussed in the section Disorders of Purine Salvage.) MANAGEMENT OF PURINE OVERPRODUCTION The treatment of purine overproduction states in patients with a primary derangement in purine synthesis, as in PRPP synthetase overactivity or HPRT deficiency, or in patients with a secondary increase in purine synthesis due to a primary defect in purine catabolism, is aimed at reducing uric acid formation. This is accomplished through the use of allopurinol, a drug that inhibits XO activity (Fig. 192-6; see Fig. 192-1). In the setting of Lesch-Nyhan syndrome, the use of allopurinol results in increased concentrations of hypoxanthine and xanthine in the urine. As a consequence, there is heightened risk of xanthine calculi formation. Sonography has been successfully used to detect these xanthine-containing calculi.11 In patients with normal HPRT activity, the accumulation of hypoxanthine as a result of XO inhibition leads to increased purine base salvage and decreased de novo purine synthesis. This effect reduces the total amount of purine (hypoxanthine, xanthine, and uric acid) produced each day. XO inhibition also is beneficial because it replaces uric acid as the end product of purine metabolism with the more soluble purine base, hypoxanthine. The latter effect reduces urate and uric acid concentrations in serum and urine, respectively, thereby decreasing crystal formation in all patients with gout, even those with HPRT deficiency. In patients with normal HPRT activity, there is the added effect of decreased purine production. Although there still is no satisfactory treatment for the neurologic symptoms associated with HPRT deficiency, the production of a mouse model for Lesch-Nyhan syndrome12 and the prospect of germline gene modification13 raise the potential for somatic gene therapy and possibly even future treatment and prevention of this devastating disorder.

FIGURE 192-6. Substrates and inhibitors of xanthine oxidase.

DISORDERS OF PURINE SALVAGE HYPOXANTHINE-GUANINE PHOSPHORIBOSYLTRANSFERASE DEFICIENCY HPRT deficiency is a disorder of both purine base salvage and purine synthesis. Failure to reuse hypoxanthine (and guanine) contributes to the excessive rate of uric acid production observed in patients with this enzyme defect. In normal persons, a large portion of the hypoxanthine produced each day through purine nucleotide catabolism is salvaged by the HPRT reaction. Loss of this salvage mechanism contributes to the excess production of uric acid characteristic of this disorder. Several mutations have been described.13a ADENINE PHOSPHORIBOSYL TRANSFERASE DEFICIENCY Reutilization of the purine base, adenine, is catalyzed by APRT (see Fig. 192-1). This enzyme is the product of a gene different from that for HPRT. The salvage of adenine is quantitatively much less significant than that of hypoxanthine because human cells have a limited capacity to produce adenine. Thus, APRT deficiency has no discernible effect on purine nucleotide synthesis, and patients with this enzyme defect do not exhibit increased rates of purine synthesis and uric acid production. When adenine is not phosphoribosylated by APRT, this purine base becomes available for oxidation by XO (see Fig. 192-1). The affinity of APRT for adenine is considerably greater than that of XO, and in normal individuals, essentially all of the adenine is metabolized through the salvage pathway. Symptoms in patients with APRT deficiency are the consequence of failure to “scavenge” or reuse the small amount of adenine produced or ingested each day.14 Oxidation of adenine to 2,8-dihydroxyadenine leads to the formation of an extremely insoluble purine base, and crystals of 2,8-dihydroxyadenine form in the urine, causing renal calculi.15 Complete deficiency of APRT is inherited as an autosomal recessive trait and can be diagnosed readily by assay of erythrocyte lysates. Recurrent radiolucent renal calculi may form in children and young adults with this enzyme defect. Renal insufficiency has developed in some patients. Occasionally, the stones have been misdiagnosed as uric acid calculi because of the similar physical properties of 2,8-dihydroxyadenine and uric acid. The treatment of this disorder depends on early recognition of the type of stone being formed and institution of appropriate dietary (low purine) and drug therapy (allopurinol to inhibit XO). Partial deficiency of APRT appears to be a common genetic polymorphism. Generally it is of no clinical significance but has been associated with urolithiasis.15 Using gene-targeting, a mouse model of APRT deficiency has been generated. The kidney disease that ensues resembles that seen in humans and offers new avenues of research into this disorder.16

DISORDERS OF PURINE CATABOLISM Disorders in one portion of the purine pathway frequently lead to secondary changes in other parts of this pathway, as well as producing derangements in other metabolic pathways. This is best illustrated by the group of disorders categorized as defects in purine catabolism. Inherited defects in other pathways may alter purine catabolism, secondarily, and the resultant derangements in purine metabolism contribute to the symptoms observed in these patients. ADENOSINE DEAMINASE DEFICIENCY Adenosine deaminase deaminates both adenosine and deoxyadenosine to form inosine and deoxyinosine, respectively (see Fig. 192-1). Deficiency of this enzyme activity, which is inherited as an autosomal recessive disorder, leads to profound lymphopenia.17 Patients having this enzyme defect have reduced numbers of T and B cells, decreased immunoglobulin levels, and diminution in cellular and humoral immune responses. ADA deficiency accounts for a significant number of patients with severe combined immunodeficiency (SCID). These patients experience recurrent infections, and without therapy, they often die within the first few years of life. More recently, ADA deficiency has been recognized as a cause of lymphopenia and immunodeficiency in adults.18 ADA deficiency can be diagnosed by assay of any easily obtained tissue, such as erythrocytes. The deficiency leads to the accumulation of both adenosine and deoxyadenosine. The buildup of these naturally occurring purine nucleosides, particularly deoxyadenosine, may alter lymphocyte growth and function through derangements in purine, nucleic acid, or transmethylation reactions. The accumulation of deoxy-ATP may inhibit ribonucleotide reductase, leading to a decrease in DNA synthesis and lymphocyte replication. An accumulation of S-adenosylhomocysteine may inhibit transmethylation reactions. It is unclear why a generalized deficiency of ADA has such profound effects on lymphocytes while most other organs function normally, but the susceptibility of lymphocytes to ADA deficiency may be explained partly by other features of purine metabolism in lymphocytes that lead to the accumulation of deoxy-ATP or other purine intermediates in these cells. The removal of deoxyadenosine and adenosine by replacement of ADA activity through red blood cell transfusion or bone marrow transplantation has restored immune function in some of these patients. Several SCID patients have been treated with injections of polyethylene glycol-modified bovine ADA, resulting in significant clinical improvement.19 The successful cloning of the gene responsible for ADA deficiency has opened new avenues of therapeutic intervention for this disorder. In one of the first uses of gene transfer therapy, patients with SCID have been given infusions of genetically corrected T cells, resulting in improved immune status.20 PURINE NUCLEOSIDE PHOSPHORYLASE DEFICIENCY Purine PNP catalyzes the reaction in which inosine, deoxyinosine, guanosine, and deoxyguanosine undergo phosphorolysis to purine bases (see Fig. 192-1). PNP deficiency is characterized by recurrent infections with nonbacterial organisms, reflecting the primary defect in cellular immunity.21 Laboratory tests reveal lymphopenia and a diminished number of T cells. Although there are normal numbers of B cells, their function may be significantly impaired. A confirmation of the diagnosis is obtained by the assay of erythrocyte lysates or other cell extracts. The deficiency of PNP, which is inherited as an autosomal recessive trait, leads to the accumulation of several purine nucleosides, the most important of which is deoxyguanosine. Thus, deoxyguanosine triphosphate concentrations become elevated in T lymphocytes, leading to the inhibition of ribonucleotide reductase, a reduction in DNA synthesis, and the death of T cells. The prognosis in PNP deficiency usually is better than in ADA deficiency; the replacement of PNP activity by bone marrow transplantation or red blood cell infusion has been less successful. XANTHINE OXIDASE DEFICIENCY XO catalyzes the last reactions in the purine catabolic pathway, that is, oxidation of hypoxanthine to xanthine and of xanthine to uric acid (see Fig. 192-1). This disorder is suspected in individuals with persistently low (50 years of age),26 whereas up to 40% of elderly women have some degree of thyroidal lymphocytic infiltration, usually not recognized clinically.2 The condition is probably slowly progressive in all, but only reaches overt hypothyroidism in a small percentage at the top of the disease “pyramid.”27 Approximately two-thirds of goiters in euthyroid adolescents are due to lymphocytic thyroiditis. GD is also common, occurring in ~1% of the population.2 In both GD and HT, there are several aspects that suggest the participation of an autoimmune process (Table 197-3). The overlap between GD and HT has long been recognized. Indeed, GD and HT frequently aggregate in the same families. There are several reports of identical twins, one with GD and the other with HT. In fact, both GD and HT can cohabit the same thyroid gland; the clinical expression will depend on which condition predominates.2 As there remain several genetic, immunologic, laboratory, and clinical elements that differ between these maladies, they should be considered separate (albeit closely related) entities.

TABLE 197-3. Immune Stigmata Associated with Graves Disease and Hashimoto Thyroiditis

HUMORAL IMMUNITY IN AUTOIMMUNE THYROID DISEASE The initial observations included thyroid autoantibodies in the serum of patients with HT,28 the induction of experimental thyroiditis in rabbits,29 and the presence of an abnormal thyroid stimulator in the serum of some patients with GD, capable of stimulating the guinea pig thyroid gland.30 This was termed long-acting thyroid stimulator (LATS) and proved to be an immunoglobulin G (IgG). It is now termed TSAb.20 The various antibodies that may be detected in AITD and their possible functions are documented in Table 197-4 and Figure 197-1. These are largely produced by intrathyroidal B lymphocytes, although help is required from T helper cells. TSAb has received the most attention because it acts to stimulate thyroid cells, resulting in hyperthyroidism; it is an antibody directed against epitope(s) on the extracellular domain of the TSH receptor and is an agonist of TSH; however, it stimulates the thyrocyte for several hours (compared to the short duration of stimulation by TSH).2

TABLE 197-4. Antigen-Antibody Systems Involved in Humoral Responses of Thyroid Autoimmune Disease

FIGURE 197-1. Composite diagram of a thyroid follicle, showing possible immune effector mechanisms in autoimmune thyroid disease. Cytotoxic mechanisms include direct cytotoxicity by sensitized effector (Te) cells, antibody-dependent cytotoxicity by killer (K) cells armed with thyroid autoantibody, and cell lysis by complement-fixing thyroid autoantibody. The various types of thyrotropin receptor (TSH-R) antibodies (TRAb) have different mechanisms of action. One type binds to the TSH-R and blocks thyroid-stimulating hormone (TSH) from binding to or stimulating the receptor. Another type— thyroid-stimulating antibody (TSAb)—binds to and stimulates the TSH receptor, causing excess thyroid hormone production. Thyroid growth-promoting and -inhibiting antibodies remain controversial. (T4, thyroxine; ATP, adenosine triphosphate; cAMP, cyclic adenosine monophosphate.) (Reproduced from Strakosch CR, Wenzel BE, Row VV, Volpé R. Immunology of autoimmune thyroid diseases. N Engl J Med 1982; 307:1499.)

The original LATS assay was an in vivo guinea pig and, later, mouse assay.31 Many other assays for the TSH-receptor (TSH-R) antibodies found in GD have now been described.2 Since some of these antibodies are not stimulatory, the term TSAb should be used to refer only to those antibodies that cause increased cyclic adenosine monophosphate (cAMP) in thyroid cells. The radioligand assay that measures binding of IgG to the TSH receptor (by inhibition of binding of labeled TSH) is generally referred to as thyrotropin-binding–inhibitory immunoglobulin (TBII). In patients with GD specifically, there is a close correlation between TBII and TSAb, which may be useful in the diagnosis of that condition.2 However, while TSAb is a TSH-R antibody demonstrable by the TBII assay, some IgGs that are positive in the TBII assay will not stimulate the thyroid cells, and some even inhibit TSH activity in vivo and in vitro (thyroid-stimulating–blocking antibody, TSBAb), and, thus, are capable of causing or contributing to hypothyroidism.2 This most typically is associated with atrophic thyroiditis.2 TSBAb has also been associated with transient neonatal hypothyroidism, due to passive placental transfer to the fetus.2 While, presently, only functional assays (i.e., stimulation vs. inhibition) differentiate between the two antibodies, evidence suggests that they bind to different epitopes on the TSH receptor.32 By use of assays that measure the generation of cAMP in the human or in Fischer rat thyroid line-5 thyrocytes, several laboratories have demonstrated TSAb in the sera of ~95% of patients with untreated GD.2 TSAb in pregnant GD patients may rise in the first trimester, but tends to decline in the third trimester, and may sometimes temporarily disappear. However, in some cases, while it may have become reduced from very high levels, it may remain markedly elevated throughout pregnancy. In such instances, there is a very real possibility of fetal and neonatal hyperthyroidism, which can be a very serious illness. Since this is due to placental passive transfer of the antibody, it gradually declines over several weeks in the newborn; thus, treatment is necessary for only this length of time.2 Following delivery, maternal TSAb may rebound to higher levels once again, sometimes leading to postpartum GD. TSAb tends to decline in many patients with long-term antithyroid drug therapy; while this, along with regression of the goiter, might suggest that the patient is in remission, sometimes the latter is very short-lived, and the disease recurs.2 The continued presence of TSAb is associated with a high relapse rate, following discontinuance of medication.2 TSAb will become further elevated after iodine-131 (131I) therapy for several months as a result of the radiation-induced release of thyroid antigens, including TSH-R, and will then decline.2 In all of these circumstances, TBII results will in general parallel those of TSAb, aside from the exceptions noted. TSAb may also transiently appear with liberation of membrane antigens in subacute or silent thyroiditis,13 or due to molecular mimicry following acute yersiniosis33; in the latter patients, however, there was no evidence of thyroid dysfunction.33 Thus, the presence of TSAb does not invariably signify a diagnosis of GD. Other thyroid antibodies that are useful clinically are the thyroglobulin and thyroperoxidase (microsomal) antibodies. A high titer of thyroglobulin antibody, a non–complement-fixing antibody, is found in ~55% of patients with HT and 25% of patients with GD. It occurs less often in patients with thyroid carcinomas, and other (nonthyroidal) autoimmune diseases. Thyroperoxidase antibodies (TPO Ab; complement fixing) are found much more commonly in AITD than in thyroglobulin Ab, and have a much closer relationship with thyroid dysfunction and abnormal histology.1,2 They are observed in ~95% of patients with HT, 90% of those with “idiopathic myxedema,” and 80% of patients with GD; 72% of patients positive for TPO Ab manifest some degree of thyroid dysfunction (at least a state of compensated hypothyroidism, with high TSH values; patients with such antibodies in a state of compensated hypothyroidism will go on to overt hypothyroidism at the rate of 5–10% per year).2 In terms of case-finding and cost reductions, performance of thyroglobulin Ab is hardly necessary; one study has shown that TPO Ab was the only positive test in 64% of all patients with positive tests, while thyroglobulin Ab was the only positive test in 1%.34 The widespread practice of performing both tests increases the costs without an offsetting diagnostic gain.34 Even low titers of TPO Ab correlate with thyroid lymphocytic infiltration.2 High titers are highly suggestive of AITD, and very high titers are virtually diagnostic of AITD; nevertheless, a minority of patients with these disorders have only low titers or no detectable thyroid antibodies. The expression of recombinant TPO by the baculovirus system has allowed screening by an enzyme-linked immunosorbent assay (ELISA) for screening for AITD.35 Low titers of TPO Ab are also observed in some cases of papillary thyroid carcinoma, nontoxic goiter, and subacute thyroiditis, as well as in many patients with no clinical evidence of thyroid disease.2 When such antibodies are detected in women before or in early pregnancy, they tend to predict the later development of postpartum thyroiditis; moreover, these antibodies transiently increase in parallel with the functional abnormalities that occur with postpartum thyroiditis.2 Thyroid antibodies in HT with elevated TSH values will generally decline with suppressive therapy of TSH by thyroxine (T4).2 They also often decline in patients with GD treated with antithyroid drug therapy.2 Antibodies to the thyroid hormones themselves are of clinical importance only when the thyroid gland is incapable of responding to excess TSH. In such instances, antibodies to T4 and triiodothyronine (T3) will reduce the free moiety; the addition of exogenous thyroid hormone may not yield the expected clinical improvement until the hormone-binding sites on the antibody are saturated. Moreover, in the presence of these antibodies, determination of thyroid hormone concentrations may

misleadingly appear very high or low (depending on the assay system used) and conflict with the clinical presentation.1,2 Antinuclear antibodies are also found in AITD, as are antibodies to cell membranes and to a colloid component other than thyroglobulin. Whether there are thyroid growth-promoting, or growth-inhibiting, antibodies separate from TSH-R antibodies remains controversial.2 Autoantibodies to a few other organs, such as gastric antigens, islet cell antigens, and others, are found more commonly in patients with AITD by virtue of the inheritance of more than one disease susceptibility gene situated in close proximity on chromosome 6. Antibodies against certain bacteria, such as Yersinia enterocolitica, are encountered commonly in AITD. This appears to arise from an artifact of homology between thyroid and bacterial antigens, and does not signify the presence or relevance of actual bacterial infection.2,13 THE T LYMPHOCYTES IN AUTOIMMUNE THYROID DISEASE Most intrathyroidal T lymphocytes are “primed” or activated cells, and in HT are TH1-type cells; TH2 cells are seldom found.36 The proportion of intrathyroidal CD8+ clones versus CD4+ clones is greater than in the periphery. Studies in GD thyroids have had variable results. “Homing” of autologous lymphocytes to the thyroid has been demonstrated in AITD.37 Restricted heterogeneity of TCR Vbb gene expression in both GD and HT has been reported.38 Studies of antigen-specific regulatory T-lymphocyte function have yielded results of interest.2,20,39 First, the idea that there could be an antigen-specific disorder in regulatory T lymphocytes seems more rational than that of a generalized disturbance in regulatory T-lymphocyte function, since the latter would result in multiple clinical disorders of immunoregulation, making it discordant with genetic observations in which AITD appears much more frequently in families in whom the propositus also has AITD (rather than other autoimmune disorders). There are several studies indicating the presence of an organ-specific defect in regulatory T lymphocytes in AITD, in which the observation does not relate to the thyroid function of the patient (i.e., whether the patient is hyperthyroid, euthyroid, or hypothyroid).14,20,39,40,41,42,43,44,45,46,47,48,49,50,51,52,53 and 54 These studies have taken the form of measuring the impact of these cells specifically on antibodies directed against the target cell, or of reducing the sensitization of (helper) T cells against their specific antigen(s), without having a generalized effect as previously described. (Of course, under some conditions [e.g., hyperthyroidism] both effects may coexist.) Thyroid-specific antigens will not sufficiently activate CD8+ CD11b+ T lymphocytes (“pure” suppressor cells) from patients who have AITD, as well as irrelevant antigen, whereas in normal persons, CD8+CD11b+ T lymphocytes respond equally well to both relevant and irrelevant antigen.53 Moreover, TSH-R will not activate CD8 T cells from patients with GD as well as control CD8 cells (from normal persons, and patients with HT, nontoxic goiter, and type 1 diabetes). Conversely, glutamic acid decarboxylase (GAD-65), the putative antigen of the pancreatic B cells, will not activate CD8 cells in patients with type 1 diabetes as well as CD8 cells from patients with GD or HT and normal persons.54 In animal models of AITD, there is also increasing evidence for a role for regulatory T lymphocytes in many studies.14 The antigen-specific regulatory cell defect, which is partial or relative,50,51,52,53 and 54 may be due to an abnormality of specific antigen presentation, resulting in reduced specific regulatory cell activation (Fig. 197-2). Whether antigen can be presented directly from thyroid follicular cells that lack some costimulators, or require professional APCs, has not been definitively settled.20

FIGURE 197-2. Hypothesis for the pathogenesis of autoimmune thyroid disease. The fundamental defect may relate to reduced activation by specific antigen (Ag; by means of an abnormality of specific antigen presentation by histocompatibility genes) of regulatory (suppressor) T lymphocytes (Ts). Precipitating environmental factors (e.g., stress, infection, trauma, drugs, smoking, and aging) may cause a reduction in nonspecific Ts function, thus adding to the antigen-specific Ts dysfunction. The result is to reduce suppression of thyroid-directed helper T lymphocytes (TH [TH1 and TH2]), allowing them to become activated in the presence of antigen-presenting cells (APCs; e.g., monocytes) and the thyroid antigens (which must “drive” the process, activating the helper T lymphocytes). Helper T lymphocytes may then act directly on the thyrocyte through production of cytokines, which are essential at every stage of the process, or may cooperate with cytotoxic cells to produce thyroid damage. The activated helper T lymphocytes will produce interferon-g (IFN-g) at close proximity to the thyrocytes, causing the latter to express class II antigens, which may initially promote anergy, but may later allow the thyrocyte to present antigen. Helper T lymphocytes will also “help” specific B lymphocytes to produce thyroid autoantibodies, which may add to the pathologic process. Certainly, thyroid-stimulating hormone (TSH) receptor autoantibodies play a major role in causing either thyroid stimulation (thyroid-stimulating antibody in Graves disease) or thyroid inhibition. Thyroid-stimulating antibody or TSH will enhance IFN-g –induced thyrocyte HLA–DR expression and also increase thyroid antigen presentation, thus further stimulating helper T lymphocytes. Excess thyroid hormone further reduces nonspecific suppressor T-lymphocyte function and numbers, thus allowing further helper T-lymphocyte activation. These secondary processes thus tend to self-perpetuate the disease. (Reproduced from Volpé R. Immunology of autoimmune thyroid disease. In Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:217.)

THE ROLE OF THE ANTIGEN There is no evidence for an alteration in the thyroid antigen.2,13 AITD appears to be primarily a disorder of immunoregulation, with the organ dysfunction resulting from an antigen-specific attack mounted by inadequately suppressed lymphocytes directed toward these specific cellular targets. Certainly, the antigen must be available and presented to the T lymphocytes for this assault to occur. This requires expression of HLA-DR antigens on the APC55; this includes not only macrophages and dendritic cells, but thyroid cells can also express class I and class II antigens, although this may not be sufficient to present antigen directly2,20; however, it is now evident that thyrocyte HLA-DR expression is a secondary event, and that macrophages, lymphocytes, and the production of interferon-g are all necessary for the thyroid cells to express HLA-DR after the initiation of the immune assault.13 It has been suggested that the thyroid cell might be damaged by some external stimulus, then express HLA-DR as a consequence, and, thus, precipitate AITD,12 but this notion does not withstand close scrutiny.13 Rather, the evidence is consistent with the idea that the thyroid cell is initially normal, and expresses HLA-DR (as well as heat shock protein-72 and intercellular adhesion molecule-1, further amplifying the immune disorder56) as consequences of the immune disturbance. The notion that the primary event might be infective11,12 is entirely speculative; the author has argued that the thyroid cell is a passive captive to immunologic events in AITD13 and has postulated that the condition results from a disorder in immuno-regulation, with environmental factors precipitating the disease by virtue of nonspecific effects on the immune system,4 adding to the genetically induced specific partial defect. GENETIC CONTROL OF THE IMMUNE RESPONSE5,6,10 In mice, genes that map in the region that corresponds to the HLA-D locus in humans are responsible for controlling the response of T lymphocytes to a given antigen. In humans, GD is associated with HLA-DR3 (whites), while goitrous HT is associated with HLA-DR5, and atrophic thyroiditis with HLA-DR3. However, persons bearing HLA-DR3 have only a moderate increase of relative risk (three-fold); thus, this clearly does not represent the “disease susceptibility gene.” The frequency of subjects positive for HLA-DQAl*0501 is significantly increased among white GD patients, markedly increasing the relative risk.57 Patients with an autoimmune response to the TSH receptor (involving either stimulating or blocking antibodies) are genetically different in terms of HLA from patients whose thyroid autoimmune response does not involve TSH-R antibodies.58 Nonetheless, other genes are undoubtedly involved in the pathogenesis of these disorders, but these are not well defined, and their role is not understood.10,58 NATURE OF THE REMISSION OF GRAVES DISEASE More than one form of clinical remission occurs in GD.2 Surgical or 131I destruction of sufficient tissue may prevent recurrence. Conversely, continuous immunologic thyroid destruction may bring about clinical remission, or even hypothyroidism. The latter may also result from a change in the nature of a TSH-R antibody from a stimulating to a blocking antibody.2 Another important form of remission is one in which all immunologic stigmata of the disease disappear, including thyroid antibodies, TSAb, and evidence of sensitization of T lymphocytes.59 This form of remission may occur only in patients with a less severe defect in immunoregulation; in such patients, hyperthyroidism is initiated by some environmental insult acting on the immune system, converting an occult specific regulatory T-lymphocyte defect to an overt one. This is reversible when that circumstance is

overcome. The restoration of an euthyroid state by whatever means (antithyroid drugs,131I, or surgery) should further relieve the situation because the effects of hyperthyroidism on the immune system would be reversed under these circumstances. Moreover, rest, the passage of time, the clearing of infection, the use of sedation, and other nonspecific measures each will serve to allow the partially defective immunoregulatory system to be restored to its previous functional capacity.2,20 Those persons with a presumed severe defect would not be expected to enter an immunologic remission, no matter how long their antithyroid drugs were continued. Only those remissions associated with spontaneous or iatrogenic thyroid destruction would occur in this group. IMPLICATIONS FOR THERAPY FOR GRAVES DISEASE The understanding of the immune nature of GD has led to few changes in its management. Since patients with very large goiters rarely achieve immunologic remission, some selection for long-term antithyroid drug therapy can be made. It has been claimed that the antithyroid drugs are, themselves, immunosuppressive.60,61 It is difficult, however, to reconcile this proposal with the fact that many patients continue to manifest immunologic activity throughout the course of treatment, no matter what dosage of antithyroid drug is used, and no matter how well the hyperthyroidism is controlled. It is also difficult to comprehend, because of the short duration of action of the drugs, how a long-term remission after cessation of therapy would persist. Moreover, the normalization of thyroid function is attended by normalization of the suppressor/helper T-lymphocyte ratio.2 Thus, it seems more likely that the action of the antithyroid drugs on thyroid cells, normalizing all thyroid cell functions and restoring euthyroidism, is more decisive in bringing about remission than any direct immunosuppressive effect.62 Indeed, evidence indicates that antithyroid drugs bring about remissions by a direct effect on thyrocytes, reducing thyrocyteimmunocyte signaling.63,64 The use of 131I therapy de novo also is associated with immunologic perturbations, namely, a transient rise in TSAb and other thyroid autoantibodies, followed by an ultimate decline.2 This may be due to the liberation of thyroid antigens, stimulating the already disturbed immune system. Finally, subtotal thyroidectomy is often associated with a decline in TSAb activity, perhaps because most of the offending thyroid-committed lymphocytes are removed with the gland.2 Recurrences after surgery would have to be associated with (a) sufficient remaining thyroid parenchyma to respond to TSAb and (b) sufficient remaining thyroid-committed lymphocytes to mount the immune attack. The ophthalmopathy of GD is not discussed in these pages at length, since its pathophysiology is poorly understood, and there are no immune assays that are considered of general credence for use in diagnosis and management (see Chap. 43). However, the consensus is that the target cell is the retroorbital fibroblast, and the TSH-R is the chief candidate for an antigen that would cross-react with that cell.1

THE IMMUNOLOGY OF TYPE 1 DIABETES GENETICS It is evident that type 1 diabetes tends to aggregate in certain families and, thus, is genetically induced (see Chap. 136). The concordance rate in identical twins is ~50%; the disparity of the age at onset in the concordant twins suggests that the disease seems to occur at random in those predisposed to develop the disease.1,2,65,66 and 67 This implies that there is, first, a genetic factor and, second, a nongenetic factor. This has prompted the search for environmental factors in the development of this disease, such as viruses, drugs, nutritional factors, and so forth. The most compelling evidence that this disease in humans could be caused by a virus was the finding of a Coxsackie virus B4, isolated from a child who had died with diabetic ketoacidosis and overwhelming viral infection.68 However, a pathologic examination of the islets of Langerhans of this child demonstrated evidence of previous chronic B-cell damage that had preceded the acute viral infection.69 (See also ref. 69a.) Although there are experimental models that indicate that viruses may induce diabetes mellitus in vulnerable animals by infecting the islets, the evidence in humans is not compelling.1,2,65,66 and 67 It is of considerable interest, however, that ~20% of children with congenital rubella will develop diabetes mellitus in later life.65,66 and 67,70 This diabetes is somewhat atypical, in that it is often non–insulin dependent, although it usually does occur in patients who are HLA-DR3 and/or DR4 positive, many of whom express islet cell antibodies. In these cases, the diabetes does not occur at the time of the infection, but years later, and the same children have a high incidence of thyroid autoantibodies. It is likely that this is not because of viral-induced target cell damage, but is due to an effect of viruses on the immune system in genetically predisposed individuals. Another putative environmental factor may be a bovine albumin peptide found in cow's milk that cross-reacts with islet cell antigen, and has been proposed as a precipitating factor in inducing type 1 diabetes in susceptible individuals.71 This hypothesis has been refuted.72 The genetic predisposition relates to genes in the region of the HLA-DR alleles residing on chromosome 6. Type 1 diabetes is associated with the HLA alleles DR3 and DR4.5,6,73,74 Approximately 95% of persons with type 1 diabetes are positive for these alleles, as opposed to 40% of normal persons. Moreover, when siblings are identical for the HLA genes, the risk of diabetes is increased 90-fold, whereas a sibling having only one of the HLA loci in common with the diabetic sibling possesses a 37-fold increased risk.73,74 On the other hand, an HLA-nonidentical sibling has a risk for this disease similar to that of the general population. As with AITD, however, the presence of these genes does not ensure the development of type 1 diabetes, because 40% of the population have these same genes and only a small proportion of them develop diabetes, unless they happen to be members of families in whom diabetes is already present. Moreover, there are patients with this disease who have neither DR3 nor DR4 alleles. Thus, it is believed that the diabetogenic genes reside close to these genes in linkage disequilibrium with them. In most whites with type 1 diabetes, HLA-DQbeta sequences tend to share the common characteristic of not encoding aspartic acid at the 57-amino-acid position of the HLA-DQbeta peptide chain.73 However, the HLA-DQ 3.2 (DQbeta 1*0302) gene has been found to be the most prevalent susceptibility gene in white patients with this illness,74 with a relative risk of 8. It, thus, is a susceptibility gene, but not sufficient by itself to precipitate the disease. Other genetic and environmental factors are required to add to this susceptibility, conspiring to induce type 1 diabetes. Patients with this illness often manifest autoantibodies to other organs, with an increased incidence of certain other organ-specific autoimmune diseases, most commonly AITD; these are generally related to HLA-DR3 alleles. IMMUNE PHENOMENA In most patients with type 1 diabetes, the appearance of islet cell antibodies precedes overt diabetes, sometimes by many years, and occult changes in glucose metabolism may precede the overt expression of this condition.65,66 and 67 Various anti-islet cell antibodies (ICA) have been described, with various methods of detection75,76 (Table 197-5). Several groups have reported that up to 80% of patients have these antibodies at or before the onset of the disease.77,78 Indeed, the presence of ICA combined with a decrease in the first phase of insulin secretion (40 JDF u) is the best predictor of forthcoming type 1 diabetes, particularly when combined with decreased insulin secretion.83,84,85 and 86 The predictive value of ICA for development of type 1 diabetes within 10 years in first-degree relatives of patients with this illness increases from 40% at low levels of ICA to 100% at high levels, whereas the sensitivity is 88% at low levels and 31% at high levels.85 Thus, the risk of this disease in relatives of propositi increases with the titer of ICA, is greater in multiplex families, and is increased in those 15 years discordant) to their diabetic twinmates.100 No immunosuppressive therapy was provided because the recipients were monozygotic twins. At first the diabetic state cleared, but within 4 months, lymphocytic infiltration appeared in the grafts, with B-cell destruction. Thus, without any apparent antigenic stimulation other than the presence of the normal islets, the immune system still was abnormal and capable of reestablishing the disease. IMMUNOTHERAPY Interestingly, a fourth twin in the foregoing study, who also had been transplanted with a pancreatic segment, was treated prophylactically with azathioprine after the transplantation, and the diabetic state did not recur.99 This raises the issue of immunotherapy in the treatment of type 1 diabetes, before total destruction of the B cells has occurred. Because B-cell destruction does proceed at a subclinical level for years before the onset of overt diabetes, it clearly would be of importance to predict which person is going to develop diabetes and to commence therapy before B-cell destruction is complete. This may soon be possible with new means of detecting antibodies or T-lymphocyte response to islet cell antigens, such as GAD-65.92 It is well known that cyclosporine, an immunosuppressive agent, when given to newly diagnosed diabetic patients, will suppress the diabetic process, despite the considerable B-cell loss that has already occurred.101 Azathioprine has also proved similarly useful.102,103 Indeed, insulin administration itself, given before the development of frank type 1 diabetes, might prevent or delay the onset by “resting” the B cells and reducing the presentation of their antigens.102,104 Immunotherapy with other models (e.g., T-cell vaccination,105,106 and oral vaccination with myelin basic protein107) is being investigated. Vaccination with GAD-65 to susceptible, but not yet diabetic, mice has prevented the development of type 1 diabetes.108,109 This is an exciting development in the quest for a means to prevent this very serious malady. INSULIN RESISTANCE DUE TO INSULIN-RECEPTOR ANTIBODIES Brief mention should be made of this rare entity, which is not genetically related to type 1 diabetes or the other organ-specific endocrinopathies. It is termed type B insulin resistance and may be associated with profound hyperglycemia and acanthosis nigricans, although occasionally hypoglycemia may be noted when the insulin-receptor antibodies manifest an agonist, rather than antagonist, effect on insulin action.110 However, insulin-receptor antibodies may be seen in occasional type 1 diabetic patients.111

AUTOIMMUNE DISEASES OF ADRENALS, GONADS, PARATHYROIDS, AND PITUITARY ADDISON DISEASE The original description by Addison112 of 11 examples of this disorder included cases now recognized as idiopathic (now known to be autoimmune) adrenal atrophy, as well as tuberculosis of the adrenal gland and metastatic carcinoma. Although tuberculosis once accounted for most patients with Addison disease (see Chap. 213), autoimmune adrenalitis has become the most common form of this condition in Western countries. There is ample evidence supporting an autoimmune basis for this disease.1,2,18,113 The evidence derives from the histology of the disorder, the finding of autoantibodies against the adrenal cortex in many patients with this condition, the association with other organ-specific autoimmune disease, study of HLA antigens, genetic studies, and experimental observations.1,2,18,113,113a In the human disease, both adrenal glands are found to be very small, and difficult to locate at autopsy. The capsule is generally thickened and the cortex is usually completely destroyed. The remaining adrenocortical cells may be single or in small clusters. A mononuclear cell infiltration is invariable, with lymphocytes, plasma cells, macrophages, and, occasionally, germinal centers. The few remaining parenchymal cells are surrounded by the heaviest infiltration of lymphocytes, and a variable amount of fibrosis is evident.1,2,18,113 HUMORAL IMMUNITY Antiadrenal antibodies are detectable in approximately two-thirds of patients with autoimmune Addison disease.1,2,18,113 The means of detection have included the complement fixation test and immunofluorescence, but with identification of the actual antigens (21-hydroxylase in the adult disease and 17-hydroxylase in the type I

childhood Addison disease),18,113 Western blotting has been utilized. The adrenal antibodies tend to be more common in those patients with a short duration of disease, and in those who develop the disorder at an early age. The titers of adrenal antibodies are much lower than for thyroid or gastric antibodies in patients with AITD or pernicious anemia, respectively, but they may persist for many years after adequate medical therapy. Such antibodies are found very rarely in the control population and are also quite rare in first-degree relatives of patients with Addison disease (providing that these relatives do not have idiopathic hypoparathyroidism). In “idiopathic” hypoparathyroidism, adrenal antibodies occur in 25% to 30% of patients. In patients with Addison disease who have antiadrenal antibodies, there also may be antibodies that react with ovary, testis, and steroid-producing cells in the placenta. This cross-reacting antibody may be associated with primary ovarian failure. Although these are IgG antibodies and, therefore, can cross the placenta, there is no evidence that they cause damage to the fetal adrenal glands. In patients with autoimmune Addison disease, there is a high prevalence of antibodies to other organ antigens, including not only steroid-producing cells, but also parathyroid, thyroid, islet cell antigens, or gastric antigens. There is also a higher prevalence of other overt organ-specific autoimmune diseases associated with these same antibodies (Table 197-6). Thus, in patients with autoimmune adrenal disease, careful consideration should be given to the probability that there will be other organ-specific autoimmune diseases, either in an overt or occult (serologic) form.

TABLE 197-6. Associated Organ-Specific Autoimmune Diseases in Patients with Autoimmune Adrenalitis (Addison Disease)

GENETIC STUDIES Autoimmune Addison disease tends to be familial. It generally is considered to be an autosomal recessive characteristic, although the inheritance has not been completely settled. There is an increased incidence of HLA-B8 and DR3 in whites, similar to that seen with GD. This is true, however, only with patients who do not have generalized candidiasis and hypoparathyroidism. OVARIAN FAILURE Approximately 25% of women with autoimmune Addison disease have premature menopause or amenorrhea.1,2,113 Most of these have circulating antibodies against steroid-secreting cells. Such antibodies are almost never detected in patients with amenorrhea that is not associated with Addison disease. The question arises whether autoimmune gonadal failure is a closely associated, but separate, organ-specific autoimmune disease, or whether it results from cross-reactive antigens shared by gonads and adrenals. Certainly, some steroid cell antibodies are cross-reactive between adrenal, gonadal, and placental antigens. This would also explain the finding of antiovarian antibodies in some males with autoimmune adrenal failure. Sensitized T lymphocytes may be similarly cross-reactive.1,2 In some instances, however, premature menopause, which is only occasionally of proven autoimmune etiology, may not be related to Addison disease. Histologic features of the ovaries of patients with amenorrhea associated with autoimmune Addison disease will show lymphocytic infiltration and fibrous tissue, similar to that seen in AITD. Autoimmune testicular failure associated with Addison disease is uncommon; it generally is associated with polyendocrine autoimmune failure related to candidiasis and hypoparathyroidism. HYPOPARATHYROIDISM Autoimmune hypoparathyroidism occurs mostly in children and adolescents, and is often associated with mucocutaneous candidiasis (type I polyendocrine autoimmune disease). Thus, it frequently is associated with Addison disease and other organ-specific autoimmune (see Chap. 60 and Chap. 70) diseases.1 The pathologic picture is characterized by lymphocytic infiltration and atrophy. Antiparathyroid antibodies and evidence for cell-mediated immunity have been demonstrated.1 HYPOPHYSITIS There are an increasing number of cases of autoimmune hypophysitis being reported, all in women between the third and eighth decade.2,114 In many of these patients, the diagnosis was made at autopsy. A conspicuous feature of this condition has been its association with pregnancy and the postpartum state. In many of the reported cases, the disease was detected after delivery, with the longest interval after gestation being 14 months. Possibly, cases that have been diagnosed as postpartum Sheehan syndrome may instead be examples of postpartum autoimmune pituitary disease, developing insidiously during and after pregnancy. Another prominent feature of autoimmune lymphocytic hypophysitis has been its association with other organ-specific autoimmune disorders, such as HT, adrenalitis, and pernicious anemia (see Chap. 11 and Chap. 17).

AUTOIMMUNE POLYENDOCRINE DISEASE Theoretically, any patient with one expressed autoimmune endocrine disease showing serologic reactivity with another organ should be considered as potentially belonging to the polyendocrinopathies.115,115a Although many of these target organs are indeed endocrine glands, other nonendocrine organ-specific autoimmune diseases that are associated in increased frequency include pernicious anemia, myasthenia gravis, Sjögren disease, vitiligo, alopecia areata, chronic active hepatitis, idiopathic thrombocytopenic purpura, and rheumatoid arthritis.116,117,118 and 119 Those patients with overt autoimmune diseases of these organs would belong to the categories depicted in Table 197-7.115

TABLE 197-7. Classification of Polyendocrine Autoimmune Disease

Of the categories listed in Table 197-7, only categories II and III are associated with definite HLA genes.115 Category I, which seems to be the most severe form of

autoimmune polyglandular endocrine failure, does not have any particular HLA type and generally occurs in children. Indeed, the inheritance has now been assigned to chromosome 21q22.3 (i.e., not part of the HLA system).120 Possibly, in this condition, the putative suppressor T-lymphocyte defect is more severe and more nonspecific than that seen in the other entities. There are various other, rarer, polyendocrine autoimmune syndromes, such as central diabetes insipidus, autoimmune enteropathy, autoimmunity to gut hormone-secreting cells, and autoimmunity directed against specific prolactin cells in the anterior pituitary.115 The observation that all of the disorders listed in Table 197-1 have a close association with one another and often are associated with specific HLA genes suggests a very similar pathogenesis for all these autoimmune entities. It may well be that each disease has separate genes, and each a separate organ-specific defect in immunoregulation. Probably, the defect is an organ-specific abnormality in regulatory T-lymphocyte function, which is specific for each disease. The fact that some persons develop two or more diseases may relate to the inheritance of more than one closely related gene, or sets of genes. Nonetheless, much remains to be learned before these disorders are fully understood.117,119 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.

Volpé R. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:1. Volpé R. Autoimmune diseases of the endocrine system. Boca Raton, FL: CRC Press, 1990:1. Davies TF. Autoimmune endocrine disease. New York: John Wiley and Sons, 1983:1. Plotnikoff N, Murgo A, Faith R, Wybran J. Stress and immunity. Boca Raton, FL: CRC Press, 1991:1. Nepom GT, Erlich H. MHC class II molecules and autoimmunity. Ann Rev Immunol 1990; 90:493. Nagataki S, Yamashita S, Tamai H. Immunogenetics of autoimmune endocrine disease. In: Volpé R, ed. Autoimmune diseases of the endocrine system. Boca Raton, FL: CRC Press, 1990:51. Wachtel SS, Koo GC, Boyce EA. Evolutionary conservation of HY male antigen. Nature 1975; 254:270. Kovacs W, Olsen N. Sex and autoimmune disease. In: Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:183. Svejgaard A, Platz P, Ryder LR. HLA and endocrine disease. In: Volpé R, ed. Autoimmunity in endocrine disease. New York: Marcel Dekker Inc, 1985:93. Tomer Y, Steinberg D, Davies TF. Immunogenetics of autoimmune endocrine diseases. In: Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:57. Tomer Y, Davies TF. Infection, thyroid disease and autoimmunity. Endocr Rev 1993; 14:107. Martin A, Davies TF. T cells in human autoimmune thyroid disease: emerging data show lack of need to invoke suppressor T cell problems. Thyroid 1992; 2:247. Volpé R. A perspective on human autoimmune thyroid disease: is there an abnormality of the target cell which predisposes to the disorder? Autoimmunity 1992; 12:3. Volpé R. Suppressor T lymphocyte dysfunction is important in the pathogenesis of autoimmune thyroid disease. Thyroid 1993; 3:345. Volpé R. Immunoregulation in autoimmune thyroid disease. Thyroid 1994; 4:157. Iwatani Y. Normal mechanisms for tolerance. In: Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:1. Ishimura H, Kuchroo V, Abramson-Leeman S, Dorf ME. Comparison between helper and suppressor cell induction. Immunol Rev 1988; 106:93. Drexhage HA. Autoimmune adrenocortical failure. In: Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:309. Murakami M, Miyashita K, Monden T, et al. Evidence that a soluble form of TSH receptor is present in the peripheral blood of patients with Graves' disease. In: Nagataki S, Mori T, Torizuka K, eds. 80 Years of Hashimoto disease. Amsterdam: Elsevier, 1993:683. Volpé R. Immunology of autoimmune thyroid disease. In: Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:217. Weetman AP, McGregor AM. Autoimmune thyroid disease: further developments in our understanding. Endocr Rev 1994; 15:788. Burman KD, Baker JR. Immune mechanisms in Graves' disease. Endocr Rev 1985; 6:183. Hashimoto H. Zue Kenntnis der lymphomatosen Veranderung der Schild-druse (Struma lymphomatosa). Arch Klin Chir 1912; 97:219. McGregor AM, Weetman AP, Ratanachaiyavong S, et al. Iodine: an influence on the development of autoimmune thyroid disease? In: Hall R, Kob-berling J, eds. Thyroid disorders associated with iodine deficiency and excess. New York: Raven Press, 1985:209. Sundick RS, Bagchi N, Brown TR. Mechanisms by which iodine induces autoimmunity. In: Drexhage HA, de Vijlder JJM, Wiersinga WM, eds. The thyroid gland, environment and autoimmunity. Amsterdam: Elsevier Science, 1990:13. Helfand M, Redfern CC. Screening for thyroid disease: an update. Ann Intern Med 1998; 129:144. Dayan CM. The natural history of autoimmune thyroiditis: how normal is autoimmunity? Proc R Coll Physicians Edinb 1996; 26:419. Roitt IM, Doniach D, Campbell RN, Hudson RV. Autoantibodies in Hashimoto's disease (lymphadenoid goitre). Lancet 1956; 2:820. Rose NR, Witebsky E. Studies on organ specificity: V. Changes in the thyroid glands of rabbits following active immunization with rabbit thyroid extracts. J Immunol 1956; 76:417. Adams DD, Purves HD. Abnormal responses in the assay of thyrotrophin. Univ Otago Med School Proc 1956; 34:11. McKenzie JM. Humoral factors in the pathogenesis of Graves' disease. Physiol Rev 1968; 48:252. Nagayama Y, Wadsworth HL, Russo D, et al. Binding domains of stimulatory and inhibitory thyrotropin (TSH) receptor autoantibodies determined with chimeric TSH-lutropin/chorionic gonadotropic receptors. J Clin Invest 1991; 88:336. Wolf M, Misaki T, Bech K, et al. Immunoglobulins of patients recovering from Yersinia enterocolitica infections exhibit Graves'-like activity in human thyroid membranes. Thyroid 1991; 1:315. Nordyke RA, Gilbert FI, Miyamoto LA, Fleury KA. The superiority of anti-microsomal over antithyroglobulin antibodies for detecting Hashimoto's thyroiditis. Arch Intern Med 1993; 153:862. Haubruck H, Mauch L, Cook NJ, et al. Expression of recombinant thyroid peroxidase by the bacillovirus system and its use in ELISA screening for diagnosis of autoimmune thyroid disease. Autoimmunity 1993; 15:275. Bagnasco M, Pesce G. Autoimmune thyroid disease: immunological model and clinical problem. Fundam Clin Immunol 1996; 4:7. Resetkova E, Nishikawa M, Mukuta T, et al. Homing of 51Cr-labelled human peripheral lymphocytes to Graves' disease tissue xenografted into SCID mice. Thyroid 1995; 5:293. Davies TF, Concepcion ES, Ben-Nun A, et al. T cell receptor V gene use in AITD: direct assessment by thyroid aspiration. J Clin Endocrinol Metab 1993; 76:660. Volpé R, Row VV. Role of antigen-specific suppressor T lymphocytes in the pathogenesis of autoimmune thyroid disease. In: Walfish PG, Wall JR, Volpé R, eds. Autoimmunity and the thyroid. Orlando, FL: Academic Press, 1985:79. Gerstein H, Rastogi B, Iwatani Y, et al. The decrease in nonspecific suppressor T lymphocytes in female hyperthyroid patients is secondary to the hyperthyroidism. Clin Invest Med 1987; 10:337. Grubeck-Loebenstein B, Derfler K, Kassal H, et al. Immunological features of non-immunogenic hyperthyroidism. J Clin Endocrinol Metab 1984; 60:150. Okita N, Row VV, Volpé R. Suppressor T lymphocyte deficiency in Graves' disease and Hashimoto's thyroiditis. J Clin Endocrinol Metab 1981; 52:528. Topliss DJ, Okita N, Lewis M, et al. Allosuppressor T lymphocytes abolish migration inhibition factor production in autoimmune thyroid disease: evidence from radiosensitivity experiments. Clin Endocrinol 1981; 15:335. Topliss DJ, How J, Lewis M, et al. Evidence for cell-mediated immunity and specific suppressor T lymphocyte dysfunction in Graves' disease and diabetes mellitus. J Clin Endocrinol Metab 1983; 57:700. Vento S, Hegarty JE, Bottazzo GF, et al. Antigen-specific suppressor cell function in autoimmune chronic active hepatitis. Lancet 1984; 1:1200. Vento S, O'Brien CJ, Cundy T, et al. Cellular immunity and specific defects of T-cell suppression in patients with autoimmune thyroid disorders. In: Pinchera A, Ingbar SH, McKenzie JM, Fenzi GF, eds. Thyroid autoimmunity. New York: Plenum Press, 1987:304. Noma T, Yata J, Shishiba Y, Inatsuki B. In vitro detection of antithyroglobulin antibody forming cells from the lymphocytes of chronic thyroiditis patients and analysis of their regulation. Clin Exp Immunol 1982; 49:465. Mori H, Hamada N, DeGroot LJ. Studies in thyroglobulin-specific suppressor T cell function in autoimmune thyroid disease. J Clin Endocrinol Metab 1985; 61:306. Tao TW, Gatenby PA, Leu SL, et al. Helper and suppressor activities of lymphocyte subsets on antithyroglobulin production in vitro. J Clin Endocrinol Metab 1985; 61:520. Benveniste P, Row VV, Volpé R. Studies of the immunoregulation of thyroid autoantibody production in man. Clin Exp Immunol 1985; 61:274. Iitaka M, Aquayo J, Iwatani Y, et al. Studies of the effect of suppressor T lymphocytes on the induction of antithyroid microsomal antibody-secreting cells in autoimmune thyroid disease. J Clin Endocrinol Metab 1988; 66:708. Iitaka M, Aquayo J, Iwatani Y, et al. In vitro induction of antithyroid microsomal antibody secreting cells in peripheral blood mononuclear cells from normal subjects. J Clin Endocrinol Metab 1988; 69:749. Yoshikawa N, Morita T, Resetkova E, et al. Reduced antigen-specific activation of suppressor T lymphocytes in autoimmune thyroid disease. J Endocrinol Invest 1993; 16:609. Mukuta T, Yoshikawa N, Resetkova E, et al. Activation of T cell subsets by synthetic TSH receptor peptides and recombinant glutamate decarboxylase in autoimmune thyroid disease and insulin-dependent diabetes mellitus. J Clin Endocrinol Metab 1995; 80:1264. Weetman AP, Volkman DJ, Burman KD, et al. The in vitro regulation of human thyrocyte HLA-DR antigen expression. J Clin Endocrinol Metab 1985; 61:817. Arreaza G, Yoshikawa N, Resetkova E, et al. Expression of intercellular adhesion molecule-1 on human thyroid cells before and after xenografting in nude and severe combined immunodeficient mice. J Clin Endocrinol Metab 1995; 80:3224. Yanagawa T, Mangklabruks A, Chang YB, et al. Human histocompatibility leukocyte antigen-DQA1*0501 allele associated with genetic susceptibility to Graves' disease in a Caucasian population. J Clin Endocrinol Metab 1993; 76:1569. McLachlan S. The genetic basis of autoimmune thyroid disease: time to focus on chromosomal loci other than the major histocompatibility complex (HLA in man). J Clin Endocrinol Metab 1993; 77:605A. How J, Topliss DJ, Strakosch C, et al. T lymphocyte sensitization and suppressor T lymphocyte defect in patients long after treatment for Graves' disease. Clin Endocrinol 1983; 18:61. Weetman AP, McGregor AM, Hall R. Evidence for an effect of antithyroid drugs on the natural history of Graves' disease. Clin Endocrinol 1984; 21:163. Ratanachaiyawong S, McGregor AM. Immunosuppressive effects of antithyroid drugs. Clin Endocrinol Metab 1985; 14:449. Volpé R. Evidence that the immunosuppressive effects of antithyroid drugs are mediated through actions on the thyroid cell: a review. Thyroid 1994; 4:217. Totterman TH, Karlsson FA, Bengtsson M, Mendel-Hartvig I. Induction of circulating activated suppressor-like T cells by methimazole therapy for Graves' disease. N Engl J Med 1987; 316:15. Volpé R. Immunoregulation in autoimmune thyroid disease. N Engl J Med 1987; 316:44. Eisenbarth GS. Autoimmune b cell insufficiency-diabetes mellitus type I. Triangle 1984; 23:111. Rossini AA, Mordes JP, Like AA. Immunology of insulin dependent diabetes mellitus. Annu Rev Immunol 1985; 3:289. Bach JF. Autoimmunity and type I diabetes. Trends Endocr Metab 1997; 8:71. Yoon JW, Austin M, Onodera T, Notkins AL. Virus induced diabetes mellitus: isolation of a virus from the pancreas of a child with diabetic ketoacidosis. N Engl J Med 1979; 300:1173. Gepts W. The pathology of the pancreas in human diabetes. In: Andreani A, DiMario U, Federlin KF, Heding LG, eds. Immunology and diabetes. London: Kimpton, 1984: 21.

69a. Varela-Calvino R, Sojarbi G, Arif S, Peakman M. T-cell reactivity to the P2C nonstructural protein of a diabetogenic strain of Coxsakie virus B4. Virology 2000; 274:56. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79.

Menser MA, Forrest JM, Bransby JM. Rubella infection and diabetes mellitus. Lancet 1987; 1:57. Karjalainen J, Martin JM, Knip M, et al. A bovine albumin peptide as a possible trigger of insulin dependent diabetes mellitus. N Engl J Med 1992; 327:302. Atkinson MA, Bowman MA, Kao KJ, et al. Lack of immune responsiveness to bovine serum albumin in insulin-dependent diabetes. N Engl J Med 1993; 329:l853. Todd JA, Bell JI, McDevitt HO. HLA-DQbeta gene contributes to susceptibility and resistance to insulin dependent diabetes mellitus. Nature 1987; 329:599. Nepom GT. Immunogenetics and IDDM. Diab Rev 1993; 1:93. Atkinson MA, MacLaren NK. Islet cell autoantigens in insulin dependent diabetes mellitus. J Clin Invest 1993; 92:1608. Harrison LC. Islet cell antigens in insulin-dependent diabetes: Pandora's box revisited. Immunol Today 1992; 13:348. Chase HP, Voss MA, Butler-Simon N, et al. Diagnosis of pre-type I diabetes. J Pediatr 1987; 111:807. Ziegler AG, Ziegler R, Vardi P, et al. Life table analysis of progression to diabetes of anti-insulin autoantibody-positive relatives of individuals with type l diabetes. Diabetes 1989; 38:1320. Dean BM, McNally JM, Bonifacio E, et al. Comparison of insulin autoantibodies in diabetes-related and healthy populations by precise displacement ELISA. Diabetes 1989; 38:12751.

80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113.

Sachs JA, Cudworth AG, Jaraquemade D, et al. Type I diabetes and the HLA-D locus. Diabetologia 1980; 18:41. Schatz DA, Winter WE, Maclaren NK. Immunology of diabetes mellitus. In: Volpé R, ed. Autoimmune diseases of the endocrine system. Boca Raton, FL: CRC Press, 1990:241. Tarn AC, Thomas JN, Dean BM, et al. Predicting insulin dependent diabetes. Lancet 1988; 1:845. McCulloch DK, Klaff LJ, Kahn SE, et al. Non-progression of subclinical beta cell dysfunction among first degree relatives of IDDM patients. Five year follow-up of the Seattle family study. Diabetes 1990; 39:549. Bonifacio E, Bingley P, Shattock M, et al. Quantification of islet cell antibodies and prediction of insulin dependent diabetes. Lancet 1990; 335:147. Riley WJ, Maclaren NK, Krischer J, et al. A prospective study of the development of diabetes in relatives of patients with insulin dependent diabetes. N Engl J Med 1990; 323:1167. Peig M, Gomis R, Ercilla G, et al. Correlation between residual b cell function and islet cell antibodies in newly diagnosed type I diabetes: follow-up study. Diabetes 1989; 38:1396. Spencer KM, Tarn A, Dean BM, et al. Fluctuating islet cell autoimmunity in unaffected relatives of patients with insulin dependent diabetes. Lancet 1984; 1:764. Lernmark A, Baekkescov S. Islet cell antibodies—theoretical and practical implications. Diabetologia 1981; 24:431. Brogren CH, Lernmark A. Islet cell antibodies in diabetes. J Clin Endocrinol Metab 1982; 11:409. Nerup J, Mandrup-Poulsen T, Molvig J. The HLA-IDDM association: implications for the etiology and pathogenesis of IDDM. Diabetes Metab Rev 1987; 3:779. Baekkeskov S, Aanstoot HJ, Fu Q, et al. The glutamate decarboxylase and 38 KD autoantigens in type I diabetes: aspects of structure and epitope recognition. Autoimmunity 1993; 15:24. Harrison LC, Honeyman MC, DeAizpurua HJ, et al. Inverse relation between humoral and cellular immunity to glutamic acid decarboxylase in subjects at risk of insulin dependent diabetes. Lancet 1993; 341:1365. Bottazzo GF, Dean BM, McNally JM, et al. In situ characterization of autoimmune phenomena and expression of HLA molecules in the pancreas in diabetic insulitis. N Engl J Med 1985; 313:353. Nerup J, Andersen OO, Bendixen G, et al. Anti-pancreatic cellular hypersensitivity in diabetes mellitus. Diabetes 1971; 20:424. Boitard C, Debray-Sachs M, Pouplard A, et al. Lymphocytes from diabetics suppress insulin release in vitro. Diabetologia 1981; 21:41. Boitard C, Yasunami R, Dardenne M, Bach JF. T cell mediated inhibition of the transfer of autoimmune diabetes in NOD mice. J Exp Med 1989; 169:1669. Fairchild RS, Kyner JL, Abdou NI. Suppressor cell dysfunction in insulin dependent diabetes. Diabetes 1980; 29:52A. Lohmann D, Krug J, Lampeter EF, et al. Defect of suppressor cell activity and cell mediated anti-beta cell cytotoxicity in type I diabetes mellitus. Diabetologia 1986;29:421. Boitard C, Timsit J, Larger E, et al. Pathogenesis of IDDM: immune regulation and induction of immune tolerance in the NOD mouse. Autoimmunity 1993; 15(Suppl):12. Sutherland DER, Sibley R, Xu XZ, et al. Twin to twin pancreas transplantation reversal and re-enactment of the pathogenesis of type I diabetes. Trans Assoc Am Physicians 1984; 97:80. Stiller CR, Dupre J. Immune interventional studies in type I diabetes mellitus: summary of the London (Canada) and Canadian-European experience. In: Eisenbarth GS, ed. Immunotherapy of diabetes and selected autoimmune diseases. Boca Raton, FL: CRC Press, 1989:73. Harrison LC, Colman PG, Dean B, et al. Increase in remission rate in newly diagnosed type I diabetic subjects treated with azathioprine. Diabetes 1985; 34:1306. Silverstein J, Maclaren N, Riley W, et al. Immunosuppression with azathioprine and prednisone in recent onset insulin dependent diabetes mellitus. N Engl J Med 1988; 319:599. Muir A, Luchetta R, Song HY, et al. Insulin immunization protects NOD mice from diabetes. Autoimmunity 1993; 15:58A. Boitard C. The differentiation of the immune system towards anti-islet autoimmunity: clinical prospects. Diabetologia 1992; 35:1101. Formby B, Shao T. T cell vaccination against autoimmune diabetes in non-obese diabetic mice. Ann Clin Lab Sci 1993; 23:137. Weiner HL. Treatment of autoimmune disease by oral tolerance to autoantigens. Autoimmunity 1993; 15:6. Kaufman DL, Clare-Saizier M, Tian J, et al. Spontaneous loss of T-cell tolerance to glutamic acid decarboxylase in murine insulin-dependent diabetes. Nature 1993; 366:69. Tisch R, Yang XD, Singer S, et al. Immune response to glutamic acid decarboxylase correlates with insulitis in non-obese diabetic mice. Nature 1993; 366:72. Bloise W, Wajchenberg BL, Moncada VY, et al. Atypical antiinsulin receptor antibodies in a patient with type B insulin resistance and scleroderma. J Clin Endocrinol Metab 1989; 68:227. Maron R, Elias D, Dejong BM, et al. Autoantibodies to the insulin receptor in juvenile onset insulin dependent diabetes. Nature 1983; 303:817. Addison T. On the constitutional and local effects of disease of the suprarenal glands. In: Wilks S, Daldy T, eds. A collection of the published writings of the late Thomas Addison, M.D., physician to Guy's Hospital. London: New Sydenham Society, 1868:209. Muir A, Schatz DH, MacLaren NK. Autoimmune Addison's disease. Springer Semin Immunopathol 1993; 14:275.

113a. de Carmo Silva R, Kater SA, Laureti S, et al. Autoantibodies against recombinant human steroidogenic enzymes 21-hydroxylase, side-chain cleavage and 17a-hydroxylase in Addison's disease and autoimmune polyendocrine syndrome Type III. Europ J Endocrinol 2000; 142:187. 114. Josse R. Autoimmune hypophysitis. In: Volpé R, ed. Autoimmune diseases of the endocrine system. Boca Raton, FL: CRC Press, 1990:331. 115. Bottazzo GF, Doniach D. Polyendocrine autoimmunity: an extended concept. In: Volpé R, ed. Autoimmunity and endocrine disease. New York: Marcel Dekker Inc, 1985:375. 115a. Ekwall O, Hedstrand H, Haavik J, et al. Pteridin-dependent hydroxylases as autoantigens in autoimmune polyendocrine syndrome type 1. J Clin Endocrinol Metab 2000; 85:2944. 116. Green ST, Ng JP, Chan-Lam D. Insulin dependent diabetes mellitus, myasthenia gravis, pernicious anemia, autoimmune thyroiditis and autoimmune adrenalitis in a single patient. Scott Med J 1988; 33:213. 117. Neufeld M, Blizzard RM. Polyglandular autoimmune disease. In: Pinchero A, ed. Autoimmune aspects of endocrine disorders. New York: Academic Press, 1980:357. 118. Irvine WJ. Autoimmunity in endocrine disease. In: Besser GM, ed. Advanced medicine, vol 13. Tunbridge Wells, England: Pittman, 1977;115. 119. Eisenbarth G, Maes M. Autoimmune polyendocrine syndromes. In: Volpé R, ed. The autoimmune endocrinopathies. Contemporary Endocrinology Series. Totowa, NJ: Humana Press, 1999:349. 120. Aaltonen J, Bjorses P, Sandkuijl L, et al. An autosomal locus causing autoimmune disease: autosomal polyglandular disease type I assigned to chromosome 21. Nat Genet 1994; 8:83.

CHAPTER 198 SHORT STATURE AND SLOW GROWTH IN THE YOUNG Principles and Practice of Endocrinology and Metabolism

CHAPTER 198 SHORT STATURE AND SLOW GROWTH IN THE YOUNG THOMAS ACETO, JR., DAVID P. DEMPSHER, LUIGI GARIBALDI, SUSAN E. MYERS, NANCI BOBROW, AND COLLEEN WEBER Normal Linear Growth, Postnatal Intrauterine Growth Restriction/Prematurity Intrauterine Growth Restriction Prematurity Therapy Changing Growth Patterns in Recent History Postnatal Somatic Growth in Infants with Atypical Fetal Growth Patterns Normal Sexual Maturation Precocious Puberty Idiopathic Precocious Puberty Precocious Puberty in Girls: Organic Versus Idiopathic Interpretation of Clinical Information History Useful Clinical Information Physical Examination Definitions Classification Short Stature with or without Slowing of Growth Familial Delayed Puberty Congenital Growth Hormone Deficiency Acquired Growth Hormone Deficiency Diagnostic Imaging in Growth Hormone Deficiencies Therapy and Prognosis of Growth Hormone Deficiencies Treatment of Glucocorticoid-Induced Growth Suppression Indications for Growth Hormone Treatment in Patients without Growth Hormone Deficiency Turner Syndrome Prader-Willi Syndrome Renal Dwarfism Complications of Growth Hormone Therapy Growth Hormone Insensitivity (Laron Syndrome) Retesting Young Adults with Childhood-Onset Growth Hormone Deficiency Growth Hormone Deficiency in Adults Young Adults with Growth Hormone Deficiency Treatment of Growth Hormone–Deficient Adults Growth Hormone for the Aged Growth Hormones and Cognitive Functioning Reversal of Early Atherosclerotic Changes Congenital Hypothyroidism Acquired Primary Hypothyroidism Maternal Hypothyroidism and Child Development Skeletal Dysplasias and other Syndromes Achondroplasia Hypochondroplasia Multiple Epiphyseal Dysplasia Syndrome Other Syndromes Normal or Abnormal Genetic Short Stature Slow Growth with or without Short Stature Atypical Crohn Disease Cushing Disease or Syndrome Iatrogenic Effects on Growth Psychosocial Management of Short Stature Chapter References

NORMAL LINEAR GROWTH, POSTNATAL After birth, many infants shift from a growth rate that is determined by maternal factors to one that is increasingly related to the infant's own genetic background, as reflected by midparental size. Thus, for approximately two-thirds of normal infants, the linear growth rate shifts percentiles during the first 18 months of life. The number of infants shifting upward in growth rate and the number shifting downward are approximately equal. Normal full-term infants who are relatively small at birth but whose genetic background dictates a larger size accelerate toward the new growth rate soon after birth and achieve a new channel of growth by an average age of 13 months (Fig. 198-1). Those infants who decelerate in their linear growth do so during the second to sixth months and achieve a new, lower growth channel by an average age of 13 months (Fig. 198-2). By the age of 18 months to 2 years, the shifts in growth rate usually cease, and growth proceeds along the same percentile.

FIGURE 198-1. The average linear growth of 18 middle-class normal infants who had been below the 10th percentile for length at full-term birth but who achieved the 50th percentile or better by the age of 2 years, at which time their stature correlated well with that of their parents. The catch-up in linear growth began soon after birth, and a new channel of growth had been achieved by 4 to 18 months. (From Smith DW, Truog W, Rogers JE, et al. Shifting linear growth during infancy: illustration of genetic factors in growth from fetal life through infancy. J Pediatr 1976; 89:225.)

FIGURE 198-2. The average linear growth of 23 normal middle-class infants who had been at or above the 90th percentile for linear growth at full-term birth but who fell to the 50th percentile or less by 2 years of age, at which time their stature correlated well with that of their parents. The deceleration of linear growth did not begin

until 2 to 7 months of age, and a new channel was achieved by 8 to 19 months. (From Smith DW, Truog W, Rogers JE, et al. Shifting linear growth during infancy: illustration of genetic factors in growth from fetal life through infancy. J Pediatr 1976; 89:225.)

INTRAUTERINE GROWTH RESTRICTION/PREMATURITY INTRAUTERINE GROWTH RESTRICTION Intrauterine growth restriction occurs in association with environmental, maternal, placental, and fetal factors.1,1a Whether different types of intrauterine growth restriction exist or whether the condition should be diagnosed solely on the basis of birth weight that is low for gestational age is unclear.2,3 Neonates with body weights of 2 SD below the mean, both may have abnormal short stature and should be studied for a genetically controlled disease such as congenital GHD or skeletal dysplasia (e.g., hypochondroplasia, multiple epiphyseal dysplasia syndrome). THERAPY AND PROGNOSIS No therapy is available for normal genetic short stature. Whether drug therapy for these normal children will ever be indicated is uncertain, given the difficulty of evaluating the long-term effects of drugs such as GH in healthy youngsters with normal short stature. Most parents are reassured by the knowledge that their normally short child is healthy and will grow to adulthood like themselves as a normally short and functional individual. Affected teenage boys should be encouraged to engage in sports that they enjoy and that are safe (wrestling, baseball, swimming, tennis, ice skating, sailing) and to avoid sports that pose a potential risk (tackle football). They should be reassured that normally short girls are delighted to date normally short young men. Those parents, usually fathers, who find the diagnosis difficult to accept should be reassured repeatedly. Therapy for abnormal genetic short stature depends on the underlying disease. SLOW GROWTH WITH OR WITHOUT SHORT STATURE DEPRIVATION SYNDROME Deprivation, caloric or emotional, slows weight gain and, eventually, linear growth. The type I deprivation syndrome has more of a nutritional component and the type II syndrome has more of a psychosocial component.98,99 Type I deprivation syndrome is seen in infants and young children. For various reasons, these patients have not received enough food or, in some cases, enough attention. The parent or caregiver may be disorganized, inadequately trained, misguided, overwhelmed, or disturbed. This condition also has been described in the second decade of life.100 Type II deprivation syndrome, the childhood variety, affects children older than 3 years and, occasionally, teenagers. Parents or caregivers, who frequently are alcoholics, abuse these children emotionally. Although the disorder occurs more often in the lower socioeconomic classes, the authors have documented the deprivation syndrome in the upper classes. Boys are affected most commonly.101 At the initial evaluation, hypopituitarism, including GHD, often is present. Without intervention, the prognosis for normal growth and development is guarded to dismal for patients with both types of deprivation syndrome. PRESENTING MANIFESTATIONS Infants with type I deprivation syndrome have slowing of growth, a scrawny appearance, and a relatively alert demeanor, although some look dejected. Kwashiorkor (see Chap. 127) rarely occurs in the United States. Eight patients, aged 14 to 27 months, were found to have consumed excessive amounts of fruit juice with resultant failure to thrive (abnormally low weight)102; however, these patients recovered after nutritional intervention. Sixteen extensively studied patients with nutritional dwarfing (aged 10 to 16 years) were found to have inappropriate eating habits and subnormal weight gain (accompanied by a proportionate decline in growth velocity), but no signs of emaciation. Parents reported that these children became satiated early during the course of a meal.100 Children with type II deprivation syndrome, those with “psychosocial dwarfism,” generally are withdrawn, grow extremely slowly, and have delayed sexual maturation. Most have an appropriate weight for their height, and some resemble patients with celiac dwarfism, with protuberant abdomens and wasted buttocks. Eventually, a history emerges of polydipsia, polyphagia, stealing of food, eating from garbage cans, and drinking from toilet bowls. Developmentally, patients in both groups perform suboptimally. ESTABLISHMENT OF THE DIAGNOSIS The gold standard for establishing the diagnosis of deprivation syndrome is the observation of accelerated weight gain in infants, and accelerated growth as well as weight gain in older children, when patients have new caretakers (e.g., hospitals or foster homes). Feeding should be increased gradually to the recommended number of kilocalories per kilogram of the ideal body weight for the patient's age. In infants, definitive weight gain occurs in ~2 weeks; in older children, acceleration of growth and weight gain can take several months. Laboratory studies are of little help in establishing the diagnosis. The bone age is retarded, particularly in patients with psychosocial dwarfism. Infants whose linear growth rates are slowing (e.g., dropping from the 50th to the 20th percentile from the ages of 6 to 12 months, respectively) present diagnostic difficulties. Such infants may be perfectly normal and may simply be experiencing a shift from an intrauterine growth rate influenced by maternal factors to an extrauterine growth rate dictated by their own genetic backgrounds (i.e., midparental height). These normal infants establish their permanent normal growth rate (e.g., along the 20th percentile) by the age of 18 to 24 months. THERAPY AND PROGNOSIS When the caretaker is disturbed, a new caretaker must be located. For some parents of malnourished infants, education in feeding is helpful. If a biologic parent is a psychologically disturbed caretaker, psychotherapy is essential. If the parent refuses psychotherapy or the infant does not improve rapidly, the physician must use every means necessary to place the child in a new permanent home.

With intervention, the long-term overall prognosis for Children with type I deprivation syndrome is generally good. For children with type II deprivation syndrome, the long-term prognosis for growth and sexual maturation is favorable, and intellectual ability improves to some extent (Fig. 198-23). However, both intellectual function and emotional development are likely to be permanently compromised.103

FIGURE 198-23. Growth curve of a girl with deprivation syndrome. Her growth as a young child was severely stunted, during which time she was being severely mistreated. She was removed from her home and responded remarkably to kindness and attention.

ATYPICAL CROHN DISEASE104 Crohn disease, a chronic inflammatory disease of the bowel of unknown etiology but with a strong genetic component, often interferes with growth and sexual maturation, probably as a result of chronic undernutrition, and secondarily, of low serum concentrations of IGF-I.105 Several factors contribute to the nutritional problems, including increased nutrient losses and malabsorption. PRESENTING MANIFESTATIONS Growth failure can herald Crohn disease. The weight often is more compromised than the height, and puberty is delayed but the patient looks well. Perianal fistulas are common. On questioning, patients may describe intermittent attacks of abdominal pain and diarrhea. CONFIRMATION OF THE DIAGNOSIS Barium contrast radiographs of the small and large bowels often are characterized by an irregular mucosa or a cobblestone-like pattern, a thickened bowel, and the presence of enteric fistulas. The segmental distribution of the lesions frequently is diagnostic. Biopsy samples of the rectal mucosa obtained by colonoscopy show typical granulomas. The erythrocyte sedimentation rate usually is elevated, the bone age is retarded, and the hemoglobin and serum albumin levels occasionally are depressed. THERAPY AND PROGNOSIS Control of the disease and provision of adequate nutrition are the prerequisites for growth, but an optimal method of accomplishing these goals has not been identified. Growth may accelerate with initial daily glucocorticoid therapy followed by alternate-day therapy in cases of stable disease. Calories have been administered by central or peripheral intravenous hyperalimentation, elemental diets, and specialized formulas.106 When the underlying disease is controlled, good nutrition alone, regardless of the method used to deliver it, stimulates growth. When disease activity cannot be stabilized and growth cannot be achieved with medical and nutritional support, surgical intervention should be considered. For growth to occur, resection must be performed before late puberty, all actively diseased bowel must be resected, and a prolonged disease-free postoperative period must be achieved. Ongoing nutritional therapy may augment the accelerated growth rate.107 The effects of various treatment plans on growth rate and final adult height require evaluation in large cooperative studies. CUSHING DISEASE OR SYNDROME Patients with Cushing syndrome secrete excess cortisol and other adrenocortical hormones, usually continuously but sometimes periodically.108 The underlying disease can be caused by an ACTH-secreting microadenoma of the pituitary (basophilic or mixed basophilic chromophobic, with a resultant bilateral adrenal hyperplasia) or by an adrenal tumor. Rarely, it is caused by a tumor secreting corticotropin-releasing hormone (CRH) or by an ectopic ACTH-producing tumor.109 The natural history of Cushing disease (adrenal hyperplasia) is unknown, but rare cases of spontaneous remission have been reported.110,111 Because many of the adrenal tumors are malignant, the mortality rate is high (see Chap. 75 and Chap. 83). After the age of 7 years, the most common underlying problem is an ACTH-producing pituitary adenoma.112,113 PRESENTING MANIFESTATIONS Although Cushing disease is rare in the young, infants as well as teenagers can be affected. Usually, help is sought for the increasingly abnormal appearance—moon facies, obesity (especially of the trunk and face), purplish striae, hypertension, acne, emotional lability, and virilization (Fig. 198-24). A review of growth data invariably reveals a pathologically slow growth rate over months or years. The slowing of growth occasionally precedes the abnormal appearance by several months or years.

FIGURE 198-24. Seventeen-year-old boy with Cushing syndrome caused by bilateral adrenal hyperplasia. Note moon facies, buffalo hump, and obesity, especially of the trunk.

CONFIRMATION OF THE DIAGNOSIS The diagnosis of Cushing disease depends on the demonstration of pathologically elevated cortisol secretion—that is, elevated urinary free cortisol levels that cannot be suppressed with low doses of dexamethasone (see Chap. 77). The response to administration of larger doses of dexamethasone helps to localize the lesion. Suppression with high doses of dexamethasone suggests the presence of adrenocortical hyperplasia caused by an abnormality of the hypothalamic-pituitary area, whereas failure to suppress is strongly suggestive of an adrenocortical tumor. Hypochloremic hypokalemic metabolic alkalosis can be present, and—in those cases caused by pituitary adenomas— late evening serum ACTH levels may be elevated. In a study of a short boy with periodic Cushing syndrome, the CRH test was found

to be helpful. ACTH and cortisol concentrations were undetectable both in the basal state and after stimulation with CRH. As expected, the patient had bilateral micronodular adrenal hyperplasia at surgery. In searching for a pituitary lesion, radiographs of the sella turcica, computed tomography of the pituitary area, MRI, petrosal sinus sampling, and, eventually, direct transsphenoidal visualization of the pituitary are helpful. For demonstrating the presence of an adrenal tumor, a radiograph of the abdomen can be useful to look for calcification of certain areas indicative of such a tumor. To reveal the tumor itself, ultrasonography, computed tomography, MRI, and radioactive iodocholesterol uptake can be used (see Chap. 88). In almost 300 patients, when plasma was sampled from the inferior petrosal sinuses with the conjunctive use of CRH, patients with Cushing disease could be distinguished from those with ectopic ACTH secretion.114 THERAPY AND PROGNOSIS Transsphenoidal microsurgery is the treatment of choice for patients with adrenal hyperplasia from a demonstrated pituitary tumor115 (see Chap. 23). Results are excellent when the tumor is visualized and removed at surgery. Many of these patients become permanently glucocorticoid deficient. The former approach, bilateral adrenalectomy, rarely is indicated. Long-term remission has been reported with pituitary irradiation (see Chap. 22). Ketoconazole can facilitate regression of the stigmata of Cushing disease.116 Surgical excision is the treatment of choice for demonstrable adrenal tumors (see Chap. 89). Well-localized adenomas have a good prognosis. However, microscopic examination does not always distinguish benign from malignant lesions. The results of chemotherapy for malignant tumors are disappointing. After successful therapy, the signs and symptoms of Cushing syndrome disappear, and many children grow in an accelerated fashion. As a group, they achieve a reasonable adult height.112 IATROGENIC EFFECTS ON GROWTH STIMULANT MEDICATION Certain neurostimulant drugs, especially methylphenidate but also pemoline and methamphetamine, can inhibit weight gain and growth before puberty but probably not after puberty. Desipramine does not inhibit the linear growth of children.117,118,119 and 120 PRESENTING MANIFESTATIONS Patients with attention deficit disorders who have been treated with neurostimulant drugs can have moderate slowing of growth and weight gain or even weight loss. Generally, these patients have received “high-normal” dosages (e.g., >1 mg/kg per day) of methylphenidate hydrochloride for many months. Anorexia is common and dose dependent, but abates with continued therapy. The results of physical examination usually are normal. ESTABLISHMENT OF THE DIAGNOSIS The diagnosis is clinical and can be confirmed only when linear growth accelerates after cessation of the medication. Discontinuance of therapy during summer vacations seems to result in accelerated growth, but catch-up is incomplete. If linear growth does not increase during a drug “holiday,” patients should be evaluated for other causes of slow growth. THERAPY AND PROGNOSIS A controlled study involving 124 preadolescent boys with attention deficit/hyperactivity disorder (ADHD) has been reported.120 Small but significant differences in height were found between children with and without ADHD. The height deficits were evident in early, but not late, adolescence, however, and were not related to the use of psychotropic medications. These data suggest that ADHD may be associated with temporary deficits in height gain through midadolescence that frequently normalize by late adolescence. This effect appears to be mediated by ADHD and not by its treatment. Thus, treatment with hGH is not indicated for this disorder.120 No definitive therapy is available, except for discontinuing drug therapy, decreasing a large dosage, or substituting desipramine. Thus, judicious prescribing of these drugs is important. During withdrawal, some children become temporarily hyperactive or depressed.

PSYCHOSOCIAL MANAGEMENT OF SHORT STATURE Current research builds on two decades of interest in psychosocial adjustment, personality and behavior factors, cognitive development, and school achievement, plus success in adult-hood of children with diagnoses of GHD, constitutional delay of growth,121,122,123,124,125 and 126 and Turner syndrome.127,128,129 and 130 Interest is increasing in the effect of GH therapy on the cognitive and behavioral functioning of short children, regardless of whether they demonstrate GHD.121,131,132,133 and 134 Societal bias toward tall stature from childhood through adulthood, even into the workplace, presents a challenge to the short individual in aspects of self-esteem, achievement, and acceptance. Therefore, their medical therapy is only one dimension of the care necessary to meet the multifaceted and variable needs of this population. The overall goal for short children is to enable them to become financially self-supporting adults who hold jobs commensurate with their intellectual and educational ability, who function independently in their social environment, and who are satisfied with their lives. To maximize the possibility of success, physicians must interact with their patients in age-appropriate ways, being especially sensitive to the physical environment, to the use of language, and to the power and influence of the therapeutic relationship. The multidisciplinary team approach, including psychological intervention, skilled and compassionate nursing, and educational and vocational counseling, is most efficacious. Achieving realistic treatment expectations is essential. In their interactions with short children, adults tend to infantilize, to overprotect, and to lower behavioral expectations. Some children respond by acting in a manner appropriate for their “height age” rather than their chronologic age. Short Children may show a tendency to withdraw from their peer group and to continue this social isolation through adolescence and into adulthood. Boys, especially those with delayed onset or absence of puberty, may avoid interaction with male peers and form nonromantic liaisons with girls or younger children. Further social isolation, including rejection of dating and heterosexual interaction, and poor school achievement are additional risks. Accompanying these withdrawal syndromes are lack of assertiveness and ambition, anxiety, low self-esteem, and dependency. GH therapy may add to the sense of well-being and overall health, thus leading to better psychological adjustment.131,134 Children of all ages must be allowed assertively and politely to correct an adult who mistakes their age or remarks about their height in a derogatory manner. The support of the entire family, including siblings and parents, plays a crucial role in helping them to be emotionally expressive, competent, self-reliant, and independent. Listening to parents, assessing their emotional capacity, and empowering and teaching them to meet the stresses of raising a child with a diagnosis of short stature are essential. Special issues relate to the language specific to the diagnosis of short stature. Sensitize parents to avoid the many “short”-related pejoratives in our language, such as calling someone “short-sighted” or referring to being “short-changed.” Gender inequality creates a greater burden for males, because short females may be viewed as “cute” or “petite.” Teaching parents to help their children to respond to teasing by bullies can be invaluable to the child who is not spontaneous in rebuffing a put-down with snappy repartee. Advise parents to encourage the development of hobbies, skills, and special talents that can help foster the high self-esteem that will cause peers to look up to their child. Specific suggestions include playing a musical instrument, singing, developing computer proficiency, debating, and engaging in arts and crafts such as painting, drawing, handicrafts, and sewing. Children must wear age-appropriate clothing, and teens should be allowed to feel that peer group fashion trends are permitted. Creating a physical environment at home that fosters self-sufficiency and inclusion in family routines, including chores, should be stressed. Specific suggestions include making stepstools available, relaxing rules about climbing on counters, rearranging usual kitchen configurations, providing light switch pull cords, and lowering closet rods. Whether it be blocks added to bike pedals or devices to make driving accessible to the teen, these aids are necessary to foster a sense of self-worth and peer group inclusion. Celebrations of the ritual rites of passage, developmental milestones, and birthdays, particularly entrance into the teenage years, are pivotal times to enhance self-esteem. Because short children tend to be at a physical disadvantage in body contact sports, encourage more personally self-competitive sports such as bettering one's own track record, rock climbing, tennis, golf, fishing, swimming, diving, skiing, martial arts, gymnastics, and wrestling in appropriate weight classes. Encourage individual problem solving; however, if, for example, the teen is self-conscious about changing clothes or showering in front of peers, altering of rigid school rules that may be changed only by the intervention of a parent or physician can save pain and humiliation. Allowing the individual to make his or her own choices and decisions fosters a sense of self-reliance.

Parents must be encouraged to be advocates for their short children at school, working closely with school personnel to preempt problems. Transition years when children move from elementary to middle and to high school may be times of increased stress when they have to redefine themselves in the hierarchy of their peer group. Special attention may be needed to insure that school achievement remains at the level of intellectual capacity. Relocating to a new community may increase stress and challenge academic and social adjustment. As a group, short children have normal intellectual function. A few patients with GHD caused by tumors or by cranial irradiation may experience varying degrees of intellectual impairment. Some girls with Turner syndrome are at greater risk for visuospatial perception deficits that cause difficulties with directional sense and map and graph reading, as well as some inability to use a computer mouse. Learning environments should maximize the child's skills while removing distractions, anxiety, or overload. Remediation education should be instituted if specific math deficits are diagnosed. Vocational planning should be consistent with each individual's pattern of strengths and weaknesses. The best means for part-time employment for teens and permanent employment for young adults may be through a friend or a nonjudgmental person who has had experience with people of short stature. Networks of such employers should be cultivated as a resource for this patient population. Role models in the community make ideal individuals to open doors and smooth the way for the short individual. The formal support groups that have been established—such as Human Growth Foundation (www.hgfound.org), Magic Foundation for Children's Growth (www.magicfoundation.org) and Little People of America (www.lpaonline.org)—are important conduits for peer group interaction as well as information and family networking. The groups are an effective advocate-lobbying group in the health care and political arenas. CHAPTER REFERENCES 1. Cowett RM, Stern L. The intrauterine growth retarded infant: etiology, prenatal diagnosis, neonatal management, and long-term follow-up. In: Lifshitz F, ed. Pediatric endocrinology. New York: Marcel Dekker, 1985:109. 1a.

Kramer MS, Seguin L, Lydon J, Galet L. Socioeconomic disparities in pregnancy outcome: why do the poor fare so poorly? Paediatr Perinat Epidemiol 2000; 14:194.

2. 3. 4. 5. 6. 7. 8.

Miller HC. Intrauterine growth retardation: past, present and future. Growth Genet Horm 1992; 8(3):5. Warshaw JB. Intrauterine growth restriction revisited. Growth Genet Horm 1992; 8(1):5. Cruise MO. A longitudinal study of the growth of low birth weight infants. I. Velocity and distance growth, birth to 3 years. Pediatrics 1973; 51:620. Kitchen WH, Doyle LW, Ford GW. Very low birthweight and growth in infants with atypical fetal growth patterns. Am J Dis Child 1977; 131:1078. Ross G, Lipper EG, Auld PAM. Growth achievement of very low birth weight premature children at school age. J Pediatr 1990; 117:307. Holmes GE, Miller HC, Khatab H, et al. Postnatal somatic growth in infants with atypical fetal growth patterns. Am J Dis Child 1977; 131:1078. Herman-Giddens ME, Slora EJ, Wasserman RC, et al. Secondary sexual characteristics and menses in young girls seen in office practice: a study from the pediatric research in office settings network. Pediatrics 1997; 99:505. 9. Palmert MR, Malvin HV, Boepple PA. Unsustained or slowly progressive puberty in young girls: initial presentation and long-term follow-up of 20 untreated patients. J Clin Endocrinol Metab 1999; 84:415. 10. Klein KO. Precocious puberty: who has it? Who should be treated? J Clin Endocrinol Metab 1999; 84:411. 10a. Lebrethon MC, Bourguignon JP. Management of central isosexual precocity: diagnosis, treatment, outcome. Curr Opin Pediatr 2000; 12:394. 11. Feuillam PP, Jones JV, Barnes K, et al. Reproductive axis after discontinuation of gonadotropin-releasing hormone analog. Treatment of girls with precocious puberty. Long-term follow-up comparing girls with hypothalamic hamartoma to those with idiopathic precocious puberty. J Clin Endocrinol Metab 1999; 84:44. 12. Butler GE, Sellar RE, Walker RF, et al. Oral testosterone undecaenoate in the management of delayed puberty in boys: pharmacokinetics and effects on sexual maturation and growth. J Clin Endocrinol Metab 1992; 75:37. 13. Wing-yee T, Boyukgelwiz A, Hindmarsh PC, et al. Long-term outcome of oxandrolone treatment in boys with constitutional delay of growth and puberty. J Pediatr 1990; 117:588. 14. LaFranchi S, Hanna CE, Mandel SH. Constitutional delay of growth: expected versus final adult height. Pediatrics 1991; 87:82. 15. Finkelstein JS, Neer RM, Biller BMK, et al. Osteopenia in men with a history of delayed puberty. N Engl J Med 1992; 326:600. 16. Wallis M. The molecular bases of growth hormone deficiency. Mol Aspects Med 1988; 10:429. 17. De Boer H, Blok G-J, Van der Veen E. Clinical aspects of growth hormone deficiency in adults. Endocr Rev 1995; 16:63. 18. Blum WF, Albertsson-Wikland K, Rosberg S, Rauke MB. Serum levels of IGF-I and IGFBP-3 reflect spontaneous growth hormone secretion. J Clin Endocrinol Metab 1993; 76:1612. 19. Frasier SD. A review of growth hormone stimulation tests in children. Pediatrics 1974; 53:929. 20. Lawson W. The diagnosis and treatment of endocrine disorders in childhood and adolescence. Springfield, IL: Charles C Thomas, 1950. 21. American Academy of Pediatrics Committee on Drugs and Committee on Bioethics. Considerations related to the use of recombinant human growth hormone in children. Pediatrics 1997; 99:122. 22. Spillotis BE, August GP, Hung W, et al. Growth hormone neurosecretory dysfunction: a treatable cause of short stature. JAMA 1984; 251:2223. 23. Daughaday WH. Recognition of growth hormone secretory disorders. (Editorial). JAMA 1984; 251:2251. 24. Moore WV, Donaldson DL, Hollowell JG, et al. Growth hormone secretory profiles: variation on consecutive nights. J Pediatr 1989; 115:51. 25. Frasier SD. A review of growth hormone stimulation tests in children. Pediatrics 1974; 53:929. 26. Rose SR, Ross JL, Uriarte M, et al. The advantage of measuring stimulated as compared with spontaneous growth hormone levels in the diagnosis of growth hormone deficiency. N Engl J Med 1988; 319:201. 27. Donaldson DL, Pan F, Hollowell JG, et al. Reliability of stimulated and spontaneous growth hormone (GH) levels for identifying the child with low GH secretion. J Clin Endocrinol Metab 1991; 72:647. 28. Loche S, Cambiaso P, Setzu S, et al. Final height after growth hormone therapy in non-growth-hormone-deficient children with short stature. J Pediatr 1994; 125:196. 29. Zadik Z, Mira U, Landau H. Final height after growth hormone therapy in peripubertal boys with a subnormal integrated concentration of growth hormone. Horm Res 1992; 37:150. 30. Frohman LA, Aceto T Jr, MacGillivray MH. Studies of growth hormone secretion in children: normal hypopituitary and constitutionally delayed. J Clin Endocrinol Metab 1967; 27:1409. 31. Cornblath M, Parker ML, Reisner SH, et al. Secretion and metabolism of growth hormone in premature and full term infants. J Clin Endocrinol Metab 1965; 25:209. 32. Martha PM Jr, Gorman KM, Blizzard RM, et al. Endogenous growth hormone secretion and clearance rates in normal boys, as determined by deconvolution analysis: relationship to age, pubertal status, and body mass. J Clin Endocrinol Metab 1992; 74:336. 33. Celniker AC, Chen AB, Wert RM Jr, Sherman BM. Variability in the quantity of circulating growth hormone using commercial immunoassays. J Clin Endocrinol Metab 1989; 68:469. 34. Granada ML, Sanmarti A, Lucas A, et al. Assay-dependent results of immunoassayable spontaneous 25-h growth hormone secretion in short children. Acta Paediatr Scand Suppl 1990; 370:63. 35. Baumann G. Growth hormone heterogeneity: genes, isohormones, variants, and binding proteins. Endocr Rev 1991; 12:424. 36. Chatelain P, Bouillat B, Cohen R, et al. Assay of growth hormone levels in human plasma using commercial kits: analysis of some factors influencing the results. Acta Paediatr Scand Suppl 1990; 370:56. 37. Rosenfeld RG, Albertsson-Wikland K, Cassorla F, et al. Diagnostic controversy: the diagnosis of childhood growth hormone deficiency revisited. J Clin Endocrinol Metab 1995; 80:1532. 38. Rosenfeld RG, Gargosky SE. Assays for insulin-like growth factors and their binding proteins: practicalities and pitfalls. J Pediatr 1996; 128:S52. 39. Werther GA. Growth hormone measurements versus auxologic treatment decisions: the Australian experience. J Pediatr 1996; 128:S47. 40. Strasburger CJ, Wu Z, Pflaum CD, Dressendorfer RA. Immunofunctional assay of human growth hormone in serum: a possible consensus for hGH measurement. J Clin Endocrinol Metab 1996; 81:2613. 41. L'Hermite-Balerjaux M, Copinschi G, Cauter EV. Growth hormone assays early to latest test generations compared. Clin Chem 1996; 42:1978. 42. Kaplan SA. Growth and growth hormone: disorders of the anterior pituitary. In: Kaplan SA, ed. Clinical pediatric and adolescent endocrinology, Philadelphia: WB Saunders, 1982:20. 43. Burstein S. Growth disorders after cranial irradiation in childhood. J Clin Endocrinol Metab 1994; 78:1280. 44. Ogilvy-Stuart AL, Clayton PE, Shalet SM. Cranial irradiation and early puberty. J Clin Endocrinol Metab 1994; 78:1282. 45. Blatt J, Bercu BB, Gillin JC, et al. Reduced pulsatile growth hormone secretion in children after therapy for acute lymphoblastic leukemia. J Pediatr 1984; 104:182. 46. Aceto T Jr, Frasier SD, Hayles AB, et al. Collaborative study of the effects of human growth hormone deficiency. I. First year of therapy. J Clin Endocrinol Metab 1972; 35:483. 47. Arslanian SA, Becker DJ, Lee PA, et al. Growth hormone therapy and tumor recurrence findings in children with brain neoplasms and hypopituitarism. Am J Dis Child 1985; 139:347. 48. Truilizi F, Scotti G, Di Natale B, et al. Evidence of a congenital midline brain anomaly in pituitary dwarfs: a magnetic resonance imaging study in 101 patients. Pediatrics 1994; 93:409. 49. Cacciari E, Zucchini S, Ambrosetto P, et al. Empty sella in children and adolescents with possible hypothalamic-pituitary disorders. J Clin Endocrinol Metab 1994; 78:767. 50. Baroncelli GI, Bertelloni S, Ceccarelli C, Saggese G. Measurement of volumetric bone mineral density accurately determines degree of lumbar undermineralization in children with growth hormone deficiency. J Clin Endocrinol Metab 1998; 83:3150. 51. Frasier SD, Foley TP Jr. Creutzfeldt-Jakob disease in recipients of pituitary hormones. J Clin Endocrinol Metab 1994; 78:1277. 52. Aceto T Jr, Sotos J, Garibaldi L, et al. Response to increasing doses of growth hormone (GH) in classic GH deficiency. Pediatr Res 1992; 31:71A. 53. Blethen SL, Compton P, Lippe B, et al. Factors predicting the response to growth hormone (GH) therapy in prepubertal children with GH deficiency. J Clin Endocrinol Metab 1993; 76:574. 54. Hedin L, Olsson B, Diczfalusy M, et al. Intranasal administration of human growth hormone (hGH) in combination with membrane permeation enhancer in patients with GH deficiency: a pharmacokinetic study. J Clin Endocrinol Metab 1993; 76:962. 55. Dhawan J, Pan LC, Pavlath GK, et al. Systemic delivery of human growth hormone by injection of genetically engineered myoblasts. Science 1991; 254:1509. 56. Barr E, Leiden JM. Systemic delivery of recombinant proteins by genetically modified myoblasts. Science 1991; 254:1507. 57. Allen DB, Julius R, Breen TJ, Attie KM. Treatment of glucocorticoid-induced growth suppression with growth hormone. J Clin Endocrinol Metab 1998; 83:2824. 58. Hopwood NJ, Hintz RL, Gertner JM, et al. Growth response of children with non-growth-hormone deficiency and marked short stature during three years of growth hormone therapy. J Pediatr 1993; 123:215. 59. Allen DB, Frasier SD, Foley TP Jr, Pescowitz OH. Growth hormone for Children with Down syndrome. (Editorial). J Pediatr 1993; 123:742. 60. Wilson DM, Lee PD, Morris AH, et al. Growth hormone therapy in hypophosphatemic rickets. Am J Dis Child 1991; 145:1165. 61. Rotenstein D, Reigel DH, Flom LL. Growth hormone accelerates growth of short children with neural tube defects. J Pediatr 1989; 115:417. 62. Angulo M, Castro-Magana C, Uy J. Pituitary evaluation and growth hormone treatment in Prader-Willi syndrome. J Pediatr Endocrinol 1991; 4:167. 63. Finkelstein BS, Silvers JB, Marrero U, et al. Insurance coverage, physician recommendations, and access to emerging treatments. Growth hormone therapy for childhood short stature. JAMA 1998; 279(9):663. 64. Saenger PH, Attie KM, DiMartino-Nardi J, et al. Metabolic consequences of 5-year growth hormone (GH) therapy in children treated with GH for idiopathic short stature. Genentech Collaborative Study Group. J Clin Endocrinol Metab 1998; 83:3115. 65. Rosenfeld RG, Frane J, Attie KM. Six-year results of a randomized prospective trial of human growth hormone and oxandrolone in Turner syndrome. J Pediatr 1992; 121:49. 65a. Chernausek SD, Attie KM, Cara JF, et al. Growth hormone therapy of Turner syndrome: the impact of age of estrogen replacement on final height. J Clin Endocrinol Metab 2000; 85:2439.

66. Holm VA, Cassidy SB, Butler MG, et al. Prader-Willi syndrome: consensus diagnostic criteria. Pediatrics 1993; 91:398. 67. Angulo M, Castro-Magana M, Mazur B, et al. Growth hormone secretion and effects of growth hormone therapy on growth velocity and weight gain in children with Prader-Willi syndrome. J Pediatr Endocrinol Metab 1996; 3:393. 68. Ritzen ME, Lindgren AC, Hagenäs L, Blichfeldt S, et al. Growth hormone treatment of children with Prader-Willi syndrome affects linear growth and body composition favourably. Acta Paediatr 1998; 87:28. 68a. Myers SE, Carrel AL, Whitman BY, Allen DB. Sustained benefit after 2 years of growth hormone upon body composition, fat utilization, physical agility, and growth in Prader-Willi syndrome. J Pediatr 2000; 137:42. 69. 70. 71. 72.

Haffner D, Schaefer F, Nissel R, et al. Effect of growth hormone treatment on the adult height of children with chronic renal failure. N Engl J Med 2000; 343:923. Hintz RL. Untoward events in patients treated with growth hormone in the USA. Horm Res 1992; 38(Suppl 1):44. Blethen SL. Pseudotumor cerebri: the national cooperative growth study. Contribution proceedings. 1993:14. Kohn B, Julius JR, Blethen SL. Combined use of growth hormone and gonadotropin-releasing hormone analogues: the national cooperative growth hormone study experience. German Study Group for Growth Hormone Treatment in Chronic Renal Failure. Pediatrics 1999; 104:1014. 73. Laron Z, Pertzeland A, Mannheimer S. Genetic pituitary dwarfism with high serum concentrations of growth hormone: a new inborn error of metabolism? J Med Sci 1966; 2:152. 74. Rosenfeld RG, Rosenbloom AI, Guevara-Aguirre J. Growth hormone (GH) insensitivity due to primary GH receptor deficiency. Endocr Rev 1994; 15:369. 75. Clemmons DR, Underwood LE. Uses of human insulin-like growth factor-I in clinical conditions. J Clin Endocrinol Metab 1994; 79:4. 75a. Laron Z. The essential role of IGF-I: Lessons from the long-term study and treatment of children and adults with Laron syndrome. J Clin Endocrinol Metab 1999; 84:4397. 76. Juul A, Kastrup KW, Pedersen SA, et al. Growth hormone (GH) provocative retesting of 108 young adults with childhood-onset GH deficiency and the diagnostic value of insulin-like growth factor I (IGF-I) and IGF-binding protein-3. J Clin Endocrinol Metab 1997; 82:1195. 77. Rosen T, Bengtsson BA. Premature mortality due to cardiovascular disease in hypopituitarism. Lancet 1990; 336:285. 78. Sönksen PH, Weissberger AJ. Growth hormone deficiency in adults. Growth Genet Horm 1998; 14:41. 79. Baum HB, Katznelson L, Sherman JC, et al. Effects of physiological growth hormone (GH) therapy on cognition and quality of life in patients with adult-onset GH deficiency. J Clin Endocrinol Metab 1998; 83:3184. 80. Zimmerman D, Saenger PH, Gharib H. AACE clinical practice for growth hormone use in adults. Endocr Pract 1998; 4:165. 81. Aleman A, Verharr HJJ, DeHaan EHF, et al. Insulin-like growth factor-I and cognitive function in healthy older men. J Clin Endocrinol Metab 1999; 84:471. 82. Pfeiffer M, Verhorner R, Zizek B, et al. Growth hormone treatment reverses early atherosclerotic changes in GH-deficient adults. J Clin Endocrinol Metab 1999; 84:453. 83. Klein RZ, Arnold MB, Bigos ST, et al. Correlation of cognitive test scores and adequacy of treatment in adolescents with congenital hypothyroidism. J Pediatr 1994; 124:383. 84. Sekadde CB, Slaunwhite WR, Aceto T Jr. Rapid radioimmunoassay of triiodothyronine in clinical radioassay procedures: a compendium. In: Besch PK, ed. American Association of Clinical Chemists, 1975:292. 85. Jones KL. Achondroplasia syndrome. In: Jones KL, ed. Smith's recognizable patterns of human malformation, 5th ed. Philadelphia: WB Saunders, 1997:298. 85a. Hunter I, Greene SA, MacDonald TM, Morris AD. Prevalence and aetiology of hypothyroidism in the young. Arch Dis Child 2000; 83:207. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134.

Sekadde CB, Slaunwhite WR Jr, Aceto T Jr, Murray K. Administration of thyroxine once a week. J Clin Endocrinol Metab 1975; 39:759. Pantsiotou S, Stanhope R, Uruena M, et al. Growth prognosis and growth after menarche in primary hypothyroidism. Arch Dis Child 1991; 66:838. Man EB, Jones WS. Thyroid function in human pregnancy, V. Am J Obstet Gynecol 1969; 104:898. Haddow JE, Klein RZ, Mitchell ML, et al. Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. N Engl J Med 1999; 341:549. Beighton P, Giedion ZA, Gorlin R, et al. International classification of osteochondrodysplasias. Am J Med Genet 1992; 44:223. Hall BD. Approach to skeletal dysplasia. Pediatr Clin North Am 1992; 39:279. Shapiro F. Epiphyseal disorders. N Engl J Med 1987; 317:1702. Jones KL. Osteochondrodysplasias. In: Jones KL, ed. Smith's recognizable patterns of human malformation, 5th ed. Philadelphia: WB Saunders, 1997. Taybi H, Lachman RS. Radiology of syndromes, metabolic disorders, and skeletal dysplasias, 3rd ed. Chicago: Year Book, 1990:671. Saleh M, Burton M. Leg lengthening: patient selection and management in achondroplasia. Orthop Clin North Am 1991; 22:589. Lavini F, Renzi-Brivio L, de Bastianai G. Psychologic, vascular, and physiologic aspects of lower limb lengthening in achondroplastics. Clin Orthop 1990; 250:138. Jones KL. Hypochondroplasia syndrome. In: Jones KL, ed. Smith's recognizable patterns of human malformation, 5th ed. Philadelphia: WB Saunders, 1997:304. Powell GF, Brasel JA, Blizzard RM. Emotional deprivation and growth retardation simulating idiopathic hypopituitarism. I. Clinical evaluation of the syndrome. N Engl J Med 1967; 276:1271. Powell GF, Brasel JA, Raiti S, et al. Emotional deprivation and growth retardation simulating idiopathic hypopituitarism. II. Endocrine evaluation of the syndrome. N Engl J Med 1967; 276:1279. Sandberg DE, Smith MM, Fornari V, et al. Nutritional dwarfing: is it a consequence of disturbed psychosocial functioning? Pediatrics 1991; 88:926. Rudolf MCJ, Hochberg Z. Annotation. Are boys more vulnerable to psychosocial growth retardation? Dev Med Child Neurol 1990; 32:1022. Smith MM, Lifshitz F. Excess fruit juice consumption as a contributing factor in nonorganic failure to thrive. Pediatrics 1994; 93:438. Money J, Annecillo C, Kelly JF. Growth of intelligence: failure and catch-up associated respectively with abuse and rescue in the syndrome of abuse dwarfism. Psychoneuroendocrinology 1983; 8:309. Rosenthal SR, Snyder JD, Hendricks KM, Walker WA. Growth failure and inflammatory bowel disease. Approach to treatment of a complicated adolescent problem. Pediatrics 1983; 72:481. Kirschner BS. Growth and development in chronic inflammatory bowel disease. Acta Paediatr Scand Suppl 1990; 366:98. Polk DB, Hattner JAT, Kerner JA Jr. Improved growth and disease activity after intermittent administration of a defined formula diet in children with Crohn's disease. J Parenter Enteral Nutr 1992; 16:499. Lipson AB, Savage MO, Davies PSW, et al. Acceleration of linear growth following intestinal resection for Crohn's disease. Eur J Pediatr 1990; 149:687. Muguruza MTG, Chrousos GP. Periodic Cushing syndrome in a short boy: usefulness of the ovine corticotropin releasing hormone test. J Pediatr 1989; 115:270. Preeyasombat C, Sikikulchayanonta V, Mahaclok Elert-Wattana P, et al. Cushing's syndrome caused by Ewing's sarcoma secreting corticotropin releasing factor-like peptide. Am J Dis Child 1992; 146:1103. Putnam TI, Aceto T Jr, Abbassi V, Kenny FM. Cushing's disease with a spontaneous remission. Pediatrics 1972; 50:477. Kammer H, Barton M. Spontaneous remission of Cushing's disease: a case report and review of the literature. Am J Med 1979; 67:519. McArthur RG, Cloutier MD, Hayles AB, Sprague RF. Cushing's disease in children: findings in 13 cases. Mayo Clin Proc 1972; 47:318. Thomas CG, Smith AT, Griffith JM, Askin FB. Hyperadrenalism in childhood and adolescence. Ann Surg 1984; 199:538. Oldfield EG, Doppman JL, Nieman LK, et al. Petrosal sinus sampling with and without corticotropin-releasing hormone for the differential diagnosis of Cushing's syndrome. N Engl J Med 1991; 325:897. Tyrell JB, Brooks RM, Fitzgerald PA, et al. Cushing's disease: selective transsphenoidal resection of pituitary microadenomas. N Engl J Med 1978; 298:753. Sonino N, Boscaro M, Merola G, Mantero F. Prolonged treatment of Cushing disease by ketoconazole. J Clin Endocrinol Metab 1985; 61:718. Roche AF, Lipman RS, Overall JE, Hung W. The effects of stimulant medication on the growth of hyperkinetic children. Pediatrics 1979; 63:847. Dickinson LC, Lee J, Ringdahl IC, et al. Impaired growth in hyperkinetic children receiving pemoline. J Pediatr 1979; 94:538. Vincent J, Varley CK, Leger P. Effects of methylphenidate on early adolescent growth. Am J Psychiatry 1990; 147:501. Spencer T, Biederman J, Wilens T. Growth deficits in children with attention deficit hyperactivity disorder. Pediatrics 1998; 102:501. Sandberg D. Short stature in middle childhood: a survey of psychosocial functioning in a clinic-referred sample. In: Stabler B, Underwood L, eds. Growth, stature, and adaptation. Chapel Hill, NC: University of North Carolina Press, 1994:19. Stabler B, Underwood L, eds. Growth, stature, and adaptation. Chapel Hill, NC: University of North Carolina Press, 1994. Zimet G. Psychosocial functioning of adults who were short as children. In: Stabler B, Underwood L, eds. Growth, stature, and adaptation. Chapel Hill, NC: University of North Carolina Press, 1994:73. Sartorio A, Conti A, Molinari E, et al. Growth, growth hormone and cognitive functions. Horm Res 1996; 45:23. Stabler B, Clopper RR, Siegel PT, et al. Links between growth hormone deficiency, adaptation and social phobia. Horm Res 1996; 45:30. Burman P, Deijen JB. Quality of life and cognitive function in patients with pituitary insufficiency. Psychother Psychosom 1998; 67:154. Lagrou K, Xhouret-Heinrichs D, Heinrichs C, et al. Age-related perception of stature, acceptance of therapy, and psychosocial functioning in human growth hormone-treated girls with Turner's syndrome. J Clin Endocrinol Metab 1998; 83:1494. McCauley E. Self-concept and behavioral profiles in Turner Syndrome. In: Stabler B, Underwood L, eds. Growth, stature, and adaptation. Chapel Hill, NC: University of North Carolina Press, 1994:181. Skuse D. Psychosocial functioning in the Turner Syndrome: a national survey. In: Stabler B, Underwood L, eds. Growth, stature, and adaptation. Chapel Hill, NC: University of North Carolina Press, 1994:151. Rovet J. School outcome in Turner Syndrome. In: Stabler B, Underwood L, eds. Growth, stature, and adaptation. Chapel Hill, NC: University of North Carolina Press, 1994:165. Cuttler L, Silvers JB, Singh J, et al. Short stature and growth hormone therapy. A national study of physician recommendation patterns. JAMA 1996; 276:531. Hintz RL. Growth hormone treatment of idiopathic short stature. Horm Res 1996; 46:208. Rao JK, Julius JR, Breen TJ, Blethen SL. Response to growth hormone in attention deficit hyperactivity disorder: effects of methylphenidate and pemoline therapy. Pediatrics 1998; 102(Suppl):497. Stabler B, Siegel PT, Clopper RR, et al. Behavior change after growth hormone treatment of children with short stature. J Pediatr 1998; 133:366.

CHAPTER 199 ENDOCRINOLOGY AND AGING Principles and Practice of Endocrinology and Metabolism

CHAPTER 199 ENDOCRINOLOGY AND AGING DAVID A. GRUENEWALD AND ALVIN M. MATSUMOTO Principles of Geriatric Endocrinology Impaired Homeostasis Nonspecific and Atypical Presentation Difficulties in Laboratory Evaluation Geriatric Assessment and Treatment Hypothalamus and Pituitary Gland Hypothalamus Posterior Pituitary: Antidiuretic Hormone Anterior Pituitary Pituitary Adenomas and the Empty Sella Syndrome Pineal Gland and Melatonin Thyroid Glands Physiology Laboratory Diagnosis of Thyroid Disease in Older Patients Diagnosis of Hyperthyroidism Diagnosis of Hypothyroidism Hyperthyroidism Hypothyroidism Nodular Thyroid Disease and Thyroid Cancer Disorders of Parathyroid Glands and Calcium Metabolism Aging and Regulation of Serum Calcium Vitamin D Deficiency Paget Disease of Bone Osteoporosis Management of Osteoporosis Hypercalcemia Primary Hyperparathyroidism Hypercalcemia of Malignancy Adrenal Cortex Physiology Laboratory Diagnosis of Adrenocortical Disease in Older Patients Hyperadrenocorticism Hypoadrenocorticism Aldosterone and Renin Adrenal Androgens Catecholamines Female Reproductive and Endocrine Function: Menopause Male Reproductive and Endocrine Function Hormonal Changes Sexual Activity and Erectile Dysfunction Obesity Carbohydrate Metabolism and Diabetes Mellitus Epidemiology Pathogenesis Clinical Presentation and Complications of Diabetes in Older Patients Diagnostic Criteria Management Lipid Levels Chapter References

The number of elderly people is growing faster than the population at large. The number of Americans older than age 65 years is expected to increase from 35 million in 2000 to 78 million in 2050. Furthermore, the number of the “oldest old,” those older than age 85, is expected to increase from 4 million in 2000 to almost 18 million in 2050.1 Thus, the latter, most frail group of elderly people with the greatest burden of age-associated diseases is the group that is growing the most rapidly.1a Because endocrine diseases such as osteoporosis, type 2 diabetes mellitus, and hypothyroidism are extremely common in older people, adult endocrinologists will see an increasing proportion of elderly patients in their practices in the future. With aging, changes occur in many parameters of endocrine function, such as decreases in growth hormone (GH) and gonadal steroid levels, and increases in cholesterol levels and adiposity. Many of these changes predispose to morbidity and mortality in later life; for example, ovarian failure affects bone mass and fracture risk. The effects of other age-related alterations (e.g., declining GH and testosterone levels) are of uncertain significance. Furthermore, the clinical presentation, diagnosis, treatment, and prognosis of certain endocrine diseases are altered with aging, greatly increasing the clinical challenges of evaluation and management in elderly patients. In turn, endocrine diseases in geriatric patients often have profound effects on functional status and quality of life, and these issues are often much more important to patients than the underlying diseases per se. Ultimately, the primary goal of medical management in elderly people is not to eliminate disease, but rather to help the older person achieve the highest possible level of functioning and quality of life.

PRINCIPLES OF GERIATRIC ENDOCRINOLOGY IMPAIRED HOMEOSTASIS Aging is characterized by a decline in functional reserve of major body organs, leading to impaired ability to restore equilibrium after environmental stresses. This age-related impairment of homeostatic regulation is evident in many endocrine functions but may become clinically evident only during acute or significant long-term stress. For example, fasting blood glucose levels exhibit very little change with normal aging, but after challenge with a glucose load, glucose levels increase more in healthy elderly people than in young adults. The function of endocrine systems may be maintained through homeostatic mechanisms and/or changes in hormone metabolism that offset the loss of function. For example, pituitary luteinizing hormone (LH) secretion and serum LH levels are increased in many elderly men and testosterone metabolism is decreased, thus compensating for a reduction in testicular testosterone secretion. In some cases, however, these changes are insufficient to maintain normal function with aging even under basal conditions. One example is aldosterone production, which declines disproportionately to its clearance rate with aging, a situation that leads to age-related decreases in basal plasma aldosterone levels. Several principles of geriatric endocrinology illustrate the complexity and challenge of evaluating frail older patients with endocrine disease. These include the atypical presentations of illness; the presence of multiple coexisting medical problems; the large number of symptoms, signs, and abnormal laboratory findings often present in individual elderly patients; underreporting of symptoms; and problems in the cognitive, psychiatric, social, economic, and functional domains. Failure to appreciate these challenges and to appropriately assess older patients with these issues in mind may result in missed or incorrect diagnoses, inappropriate treatments, and poor functional outcomes. NONSPECIFIC AND ATYPICAL PRESENTATION Endocrinopathies commonly present in elderly people with nonspecific, muted, or atypical symptoms and signs. For example, hypothyroidism and hyperthyroidism may present similarly in older adults with nonspecific symptoms, including weight loss, fatigue, weakness, constipation, and depression. The presentation of endocrine disease in geriatric patients may also be atypical compared with that in younger patients (e.g., apathy, depression, and psychomotor retardation may be associated with hyperthyroidism, and marked hyperglycemia and hyperosmolarity without ketoacidosis may be present in elderly patients with type 2 diabetes). In some patients, regardless of the cause of the illness, its manifestations may occur in the most compromised body system. Thus, in an older patient with underlying gait and balance abnormalities, falling may be the primary symptom of diseases as diverse as pneumonia, myocardial infarction, uncontrolled diabetes mellitus, or hypothyroidism. Illnesses may present in the guise of other disabling geriatric syndromes such as delirium, urinary incontinence, and dementia. Endocrine disorders may produce or be

associated with any or all of these syndromes, so endocrinologists must have a basic understanding of these disorders. Several excellent reviews of these geriatric syndromes are available.2,3 and 4 DIFFICULTIES IN LABORATORY EVALUATION In addition to the atypical or nonspecific presentations of endocrine disease described earlier, with aging, it is increasingly common for illnesses such as hypothyroidism to present without any symptoms. The presence of disease may be appreciated only on routine laboratory screening, as in the case of asymptomatic hypercalcemia secondary to hyperparathyroidism. Furthermore, the presence of multiple medical problems and the use of multiple medications may confound the evaluation of older patients. For example, decreased serum thyroxine (T4) and triiodothyronine (T3) levels and alterations in the levels of serum thyroid-stimulating hormone may occur in elderly patients who are systemically ill or are taking certain medications (e.g., glucocorticoids, dopamine) but are euthyroid (euthyroid sick syndromes), giving a misleading impression of an endocrine abnormality. The evaluation of the older patient is further complicated by the fact that normal ranges for endocrine laboratory tests are usually established in healthy young subjects and may not reflect normal values in healthy elderly people. Moreover, normative data for older populations are often confounded by the inclusion of subjects with age-associated diseases. Finally, most studies of aging and endocrine function in humans are crosssectional rather than longitudinal and, therefore, may not accurately predict age-related changes within a given individual. Indeed, variability among individuals is a hallmark of aging. GERIATRIC ASSESSMENT AND TREATMENT The onset of functional decline may be an important, and sometimes the only, clue to the development of an acute illness or exacerbation of a chronic disease in geriatric patients. Accordingly, a structured geriatric assessment should be a part of the clinical evaluation, especially in frail elderly patients. Functional assessment can detect impairments in physical function, cognition, emotional status, sensory capabilities, and activities of daily living that are not detected by standard clinical examinations,5 and these impairments are often much more important to patients than the underlying diseases that give rise to them. Such an assessment can also help to determine the response to treatment and to predict the patient's ultimate degree of disability. A useful approach to general outpatient screening of older patients for functional disability has been suggested.6 Patients with evidence of functional impairment on screening examination may benefit from a comprehensive functional assessment by an interdisciplinary care team. However, comprehensive evaluation is time consuming and expensive; therefore, it should be targeted to the most appropriate patients: frail or ill elderly people with a real or anticipated functional decline (including patients on the verge of requiring institutionalization), those with inadequate primary medical care, and those with poor economic and social support systems.7 Treatment decisions involving geriatric patients with endocrine disease must consider age-associated factors such as alterations in clearance rate and target-organ effects, coexisting medical illnesses, and the medications taken by the patient. Older patients consume a disproportionate share of medications compared with the population at large. Moreover, drug toxicities are more frequent and severe in the elderly than in young patients receiving the same drug regimen.3 Dysfunction in multiple organ systems together with cognitive and visual impairment further predispose older patients to adverse drug effects. As a result, older people are at high risk for the development of medication side effects and drug interactions secondary to polypharmacy. To minimize these risks, dosage levels for hormone replacement and medications must be adjusted for changes in clearance rate with aging, and patients should receive the lowest dosage of medication needed to achieve the therapeutic effect. New medications should be initiated using low doses and increased very gradually as needed. Finally, the medication regimen should be reviewed periodically, and medications no longer needed should be discontinued.

HYPOTHALAMUS AND PITUITARY GLAND HYPOTHALAMUS Studies directly assessing the effects of aging on parameters of hypothalamic neuroendocrine function in humans have not been performed. Some of these effects, however, can be inferred by assessing age-related alterations in circadian and ultradian rhythms (e.g., pulsatile release) of pituitary hormones and by determining pituitary hormonal responsiveness to administration of hypothalamic releasing hormones or to agents that either block end-organ feedback (e.g., clomiphene and metyrapone) or stimulate hypothalamic-pituitary hormonal secretion (e.g., stimulation of antidiuretic hormone [ADH] secretion by hypertonic saline administration or stimulation of GH secretion by insulin-induced hypoglycemia). Age-related blunting of the circadian rhythm of LH pulse frequency has been observed in healthy elderly men, suggesting altered regulation of the gonadotropin-releasing hormone (GnRH) pulse generator with aging.8 Furthermore, LH pulse frequency is relatively decreased despite reduced testosterone levels in some healthy elderly men as compared with young men, implying decreased GnRH pulse frequency in these older men.9 Interestingly, despite this impairment in baseline LH pulse frequency with aging, fasting decreases LH pulse frequency in healthy young men but not in older men, indicating altered reproductive axis regulation in response to a fasting stress.10 Finally, administration of naloxone, an opioid antagonist, does not increase LH pulse frequency in healthy older men as it does in young men, suggesting altered hypothalamic opioid regulation of LH secretion.11 In contrast to the findings for the reproductive axis, adrenocorticotropic hormone (ACTH) pulse frequency, cortisol levels, and ACTH response to corticotropin-releasing hormone (CRH) stimulation are unchanged in healthy elderly compared with young men, suggesting that hypothalamic regulation of pituitary/adrenocortical function may be relatively unimpaired by aging.12 Hypothalamic-pituitary feedback sensitivity to some end-organ hormones is altered with aging. For example, most studies have found increased feedback sensitivity to testosterone with aging,13,14 whereas glucocorticoid feedback sensitivity is decreased with aging.15 POSTERIOR PITUITARY: ANTIDIURETIC HORMONE The bulk of evidence suggests an increase in basal ADH levels with aging.16 Furthermore, aging is associated with an increased ADH responsiveness to osmotic stimuli such as hypertonic saline infusion, and pharmacologic inhibition of ADH secretion (e.g., with ethanol infusion) is impaired in elderly subjects compared with young adults.17 Taken together with the age-associated decline in glomerular filtration rate, the increased prevalence of conditions such as congestive heart failure and hypothyroidism, and the use of sulfonylurea or diuretic medication, these changes in ADH secretion predispose elderly people to the development of hyponatremia by impairing free water clearance. Elderly people are also at increased risk for dehydration and hypernatremia. Although ADH secretory capacity is unimpaired with aging, the renal response to ADH is blunted, possibly due to chronic exposure to elevated levels, resulting in decreased maximal urinary concentrating capacity.16 Furthermore, baroreceptor responsiveness to ADH declines with aging, such that release in response to hypotension or hypovolemia is decreased and the risk of volume depletion is higher. Other factors predisposing older adults to water depletion include the impairment in thirst responses to dehydration18 and the common occurrence of states that limit access to free water (e.g., altered mental status, immobility, and surgery). ANTERIOR PITUITARY GROWTH HORMONE The GH axis undergoes significant alterations in many healthy elderly people (see Chap. 12). In adult men, GH secretion declines progressively after age 40, and by age 70 to 80 approximately half of all men have no significant GH secretion over a 24-hour period. Levels of plasma somatomedin C (insulin-like growth factor-I [IGF-I]) show a corresponding decline, such that by age 70 to 80, ~40% of subjects exhibit plasma IGF-I levels similar to those found in GH-deficient children (see Chap. 12 and Chap. 173). These low IGF-I levels in octogenarians correlate with an absence of significant nocturnal GH pulses.19 Circulating levels of IGF-binding protein-3 (IGFBP-3) also decrease with aging, but whether this affects the bioactivity of IGF-I in older adults is unclear.20 Based on animal studies, this dramatic decrease in GH secretion with aging is thought to be due primarily to decreased hypothalamic secretion of GH-releasing hormone (GHRH) and increased hypothalamic somatostatin production, rather than to an age-related decrease in pituitary GHRH responsiveness.19,21 Furthermore, the ability of exogenous IGF-I to suppress serum GH levels decreases with advancing age in humans, suggesting that declining GH secretion with aging is not due to increased sensitivity to IGF-I negative feedback.22 In line with these observations, normal GH and IGF-I levels can be achieved in GH-deficient elderly subjects with GHRH administration.19 Of note, the magnitude of the increase in pulsatile GH secretion induced by fasting was found to be similar in elderly and young adult human subjects, although the absolute levels of GH secretion of older subjects were ~50% lower in both fed and fasted conditions.23 These findings suggest that age-related hyposomatotropism may be partially reversible by lifestyle modifications such as changes in diet or exercise. The GH response to various secretagogues (e.g., insulin, arginine, and L -dopa) and to GHRH is normal or reduced with aging. As with the thyrotropin-releasing hormone (TRH) stimulation test, a normal GH response may be useful in verifying intact pituitary (somatotrope) function. The insulin-tolerance test is considered the

definitive test to diagnose GH deficiency in adults,24 but in elderly people this test is associated with increased risk due to the high prevalence of ischemic heart disease in this population. The arginine-stimulation test25 and the combined administration of arginine and GHRH24 have been proposed as alternative methods for the assessment of GH secretion in older patients. Many age-associated changes in body composition, such as increased adiposity and decreased muscle and bone mass, are similar to those associated with GH deficiency in younger patients.21,26 This observation has led to the hypothesis that decreased GH secretion with aging contributes to alterations in body composition (including diminished muscle and bone mass) and increased frailty in older adults, and to the suggestion that GH supplementation might be clinically useful in preventing or reversing these age-related changes. As noted earlier, compared with young adults, many healthy older adults are deficient in GH and IGF-I, and frail elderly people living in nursing homes have even lower IGF-I levels than healthy elderly subjects.27 Whether young adult or age-adjusted reference ranges are the most appropriate standards to use for older adults is unclear, however, and definitive criteria for clinically significant GH deficiency in elderly people have not been established. Short-term GH replacement in older men with low plasma IGF-I levels was found to increase lean body mass by 9% and reduce fat mass by 14%.28 These beneficial effects were sustained in a similar study over 1 year of follow-up and regressed partially after cessation of treatment, results which suggest that hyposomatotropism contributes to age-related alterations in body composition.29 “Physiologic” replacement in elderly people, however, may require a lower dosage than in young adults.30 Side effects such as carpal tunnel syndrome and gynecomastia were common in older subjects with plasma IGF-I levels exceeding 1.0 U/mL during treatment,19 even though normal IGF-I levels in young adults may be as high as 1.5 U/mL. Moreover, whether GH supplementation can achieve meaningful improvements in functional status or quality of life in elderly people is unclear. In healthy older men with low IGF-I levels but intact functional capacity, GH replacement increased lean body mass and reduced adiposity but did not yield any discernible improvement in functional capacity.31 GH supplementation, however, possibly could improve functioning in frail GH- and IGF-I–deficient elderly people with preexisting functional deficits. Other abnormalities of GH secretion, including GH deficiency in adults with hypothalamic-pituitary disease, are covered in Chapter 12. PROLACTIN No clinically significant changes in basal prolactin levels occur with aging. The amplitude of nocturnal pulsatile prolactin secretion, however, is lower in elderly than in young men; this may be due to age-related alterations in dopaminergic regulation of prolactin secretion.32 Furthermore, several medications commonly used by elderly patients, including phenothiazines, metoclopramide, and cimetidine, inhibit dopamine secretion and sometimes cause elevated prolactin levels. Hypothyroidism increases hypothalamic TRH release, which in turn stimulates prolactin secretion. When hyperprolactinemia does occur, its clinical manifestations are usually subtle and often unrecognized. Prolactin excess has antigonadotropic effects; therefore, hyperprolactinemia causes secondary hypogonadism and may contribute to sexual dysfunction and bone loss. Less common manifestations of hyperprolactinemia in older people include gynecomastia and, rarely, galactorrhea. ADRENOCORTICOTROPIC HORMONE No significant age-related changes occur in basal ACTH and cortisol levels, or in cortisol responses to exogenous ACTH stimulation. However, evidence obtained using stimuli such as metyrapone, insulin-induced hypoglycemia, or ovine CRH, with and without vasopressin, indicates that cortisol and ACTH responses to stimuli at or above the level of the anterior pituitary are increased or prolonged with aging.33,34 Furthermore, the sensitivity of the hypothalamic–pituitary–adrenal (HPA) axis to glucocorticoid negative feedback is decreased with aging (see discussion of adrenal cortex physiology later). THYROID-STIMULATING HORMONE Conflicting data have been reported on the effect of aging on thyroid-stimulating hormone (TSH) levels. Some studies have found unchanged or slightly increased TSH levels in normal elderly people, with elevated TSH occurring more commonly in females. However, TSH was found to decrease with aging in healthy subjects who were carefully selected to exclude subclinical primary hypothyroidism.35,36 In fact, primary hypothyroidism is very common in older adults, with 3% of older men and 7% of elderly women having TSH levels >10 µU/mL.37 In other cases, TSH levels were suppressed by concurrent glucocorticoid use, and by fasting and stress, which were associated with severe nonthyroidal illnesses. Thyroid responsiveness to TSH administration is preserved with normal aging, but the TSH response to TRH is diminished or even absent in healthy elderly people, particularly in men. Therefore, an abnormal TRH-stimulation test cannot be used to support a diagnosis of hyperthyroidism in elderly people. A normal TSH response to TRH may be useful in ruling out hyperthyroidism, but this test is rarely needed with the widespread availability of highly sensitive TSH assays. Finally, in young adults, TSH secretion exhibits a circadian variation, with the highest levels of TSH released during the night. This nocturnal TSH peak is blunted with aging, however, suggesting hypothalamic dysfunction.38 GONADOTROPINS In the early perimenopausal phase of the menopausal transition, the number of ovarian follicles gradually declines, leading to a reduction in inhibin-B production.39 The decrease in negative feedback at the pituitary due to reduced inhibin-B levels is thought to be the primary stimulus leading to an increase of the serum follicle-stimulating hormone (FSH) levels that sustain inhibin-A and estradiol production until late in the menopausal transition.39 Ultimately, inhibin A and estradiol levels decline and FSH levels increase substantially, marking the progression to late perimenopausal status—although LH levels do not increase during this period. After menopause, FSH levels are increased to a greater extent than are LH levels, although by 15 years postmenopause, LH levels fall to below premenopausal levels. Postmenopausal women exhibit an exaggerated gonadotropin response to GnRH, due to loss of negative feedback from ovarian hormones. Furthermore, older postmenopausal women exhibit lower basal 24-hour mean LH levels and greater suppression of LH and FSH secretion by estradiol than younger women with premature ovarian failure, suggesting hypothalamic-pituitary alterations in aging women.40 Aging men exhibit higher basal LH and FSH levels than younger men, but gonadotropin levels often remain within the normal range. Testosterone levels are decreased in many healthy older men with elevated gonadotropin levels, implying primary testicular failure. Furthermore, decreased LH pulse frequency (an indicator of hypothalamic GnRH pulse generator activity) is evident in some healthy elderly men despite reduced testosterone levels.10,41 Evidence of altered pituitary function is also seen with aging, with slightly impaired LH responses to GnRH administration in elderly men compared with young men.42 Decreased testosterone levels together with inappropriately normal (i.e., not elevated) gonadotropin levels is a common finding in both healthy and systemically ill elderly men, suggesting secondary hypogonadism. Additional clinical and hormonal evidence of pituitary dysfunction is needed in these cases to justify imaging studies to rule out pituitary tumors. PITUITARY ADENOMAS AND THE EMPTY SELLA SYNDROME The incidence of pituitary tumors is not markedly altered with aging.43 Autopsy studies reveal that pituitary “incidentalomas” are common, occurring in 13% to 27% of subjects.44 Most functioning adenomas are microscopic prolactin-producing tumors (see Chap. 13). In contrast, nonfunctioning adenomas are the most common type of pituitary tumors diagnosed during life in older adults, comprising 61% to 73% of cases.44 Many of these apparently nonfunctioning adenomas, however, actually produce quantities of gonadotropins (especially FSH) or the a subunit of these glycoprotein hormones. Nonsecreting tumors and tumors that secrete LH, FSH, or a subunit are usually large at the time they are diagnosed, because few or no symptoms of hormonal hypersecretion occur. These tumors typically present with a mass effect, including visual field abnormalities and headaches, as incidental findings on imaging studies, or with manifestations of panhypopituitarism. Vision changes are the most common presentation of pituitary adenoma in elderly patients,45 and for pituitary tumors to present with symptoms of hormonal overproduction (e.g., acromegaly or Cushing disease) is uncommon. As in younger adults, management of large pituitary tumors usually involves transsphenoidal decompression and debulking, along with assessment of anterior pituitary hormone function and replacement of hormone deficiencies. Prolactinomas are managed medically with dopamine agonists, such as bromocriptine, pergolide, or cabergoline. Both transsphenoidal pituitary surgery and radiotherapy appear to be effective and relatively well tolerated by older patients who are appropriate candidates. The clinician may find it appropriate, however, to manage elderly patients with normal endocrine status and no visual field defects who are asymptomatic or at high surgical risk with only serial magnetic resonance imaging (MRI) and visual field assessment.44 Panhypopituitarism is difficult to diagnose in older patients because the symptoms are nonspecific and are difficult to distinguish from other common age-related symptoms. Among the presentations of hypopituitarism reported in case series of elderly patients are postural hypotension, recurrent falls, hyponatremia, weakness, weight loss, immobility, drowsiness and confusion, and urinary incontinence.44 Elderly patients with panhypopituitarism require replacement of thyroid hormone and cortisol; the indications for replacement of estrogen, testosterone, and GH in these patients have not been as clearly established.44 With the widespread use of brain-imaging techniques such as computerized tomography (CT) and MRI, both pituitary masses and the empty sella syndrome are being identified with greater frequency. An MRI study of healthy young and elderly subjects reported that pituitary height and volume tends to decrease with aging, and empty sella was observed in 19% of elderly subjects but in none of the young subjects. No relationship was seen between pituitary volume and anterior pituitary hormone

levels,46 however, confirming other reports that clinically apparent pituitary dysfunction is uncommon in empty sella syndrome.47 This study also found that the posterior pituitary bright signal on T1-weighted MRI, which is thought to reflect stored ADH-neurophysin complex, was not detected in 29% of healthy elderly subjects, whereas it was detected in all young adult subjects.46 None of the subjects had clinical manifestations of diabetes insipidus, but plasma osmolarity and ADH levels were higher in the elderly subjects. Thus, the absence of the posterior pituitary bright signal on T1-weighted MRI appears to reflect a physiologic depletion of ADH neurosecretory granules rather than a pathologic occurrence. Based on the foregoing, the functional significance of an incidental finding of empty sella or altered posterior pituitary bright signal on MRI in apparently healthy elderly patients is unclear. In such patients, conservative management is appropriate, although visual field testing, and measurement of serum anterior pituitary hormone levels in patients with empty sella, are indicated to rule out subclinical pituitary dysfunction and suprasellar involvement.

PINEAL GLAND AND MELATONIN Interest is increasing in age-related alterations in the secretion of melatonin, a hormone involved in the organization of circadian and seasonal biorhythms that is produced by the pineal gland (see Chap. 10). Melatonin production is inhibited by light exposure, resulting in a robust circadian variation (levels high at night and low during the day) that is controlled by pacemaker neurons in the suprachiasmatic nucleus of the hypothalamus.48 Melatonin has sedative effects, suggesting a role in sleep production, but its precise physiologic role has not been fully determined. Melatonin production is minimal in young infants but increases markedly after 3 months of age, reaching maximal nighttime levels by the age of 1 to 3 years. After early childhood, melatonin secretion gradually decreases, with a progressive decline continuing into old age.49 The significance of this age-related decline in melatonin secretion is unclear, but proposals have been made in the lay press that age-related melatonin “deficiency” is related to a wide variety of age-associated conditions, including immune system deficiencies, cancer, and the aging process itself.50 Based on such publicity, and the ready availability of melatonin supplements in the United States without a prescription, these supplements are enjoying wide use. The physiologic and pharmacologic effects of melatonin have not been thoroughly studied, however, and the long-term risks and benefits of melatonin-replacement therapy remain to be determined.

THYROID GLANDS PHYSIOLOGY The changes in thyroid physiology occurring with normal aging are summarized in Table 199-1. A reduction with aging in T4 secretion by the thyroid is balanced by a decrease in T4 clearance rate; therefore, serum T4 levels do not change significantly with normal aging. Levels of the binding proteins thyroid-binding globulin (TBG) and thyroid-binding prealbumin (TBPA) are not markedly affected by aging; hence, no significant change occurs in total T4 and T3 resin uptake (T3 RU). T3 levels are normal in healthy elderly people until extreme old age, when T3 declines slightly,35,51 but in the setting of nonthyroidal illness, extrathyroidal conversion of T4 to T3 by 5' deiodinase is often impaired, resulting in decreases in circulating T3 levels.52 Low serum total T3 (low T3 syndrome) is the most common thyroid function test abnormality in nonthyroidal illness and occurs in ~70% of hospitalized patients.36,53 Serum reverse T3 (rT3) levels are increased in some elderly people, but this is associated with decreased food intake52 or nonthyroidal illness53 rather than aging per se, and is due to a reduction in the metabolic clearance rate of rT3 due to impaired 5' deiodinase activity. The effects of aging on TSH levels was discussed previously.

TABLE 199-1. Alterations in Thyroid Physiology and Hormones with Aging

LABORATORY DIAGNOSIS OF THYROID DISEASE IN OLDER PATIENTS Muted, atypical, and often asymptomatic presentation of thyroid disease is the rule rather than the exception in elderly people, as discussed later. Accordingly, laboratory screening is the most reliable means to identify hypothyroidism and hyperthyroidism in the geriatric population. Office-based screening of women older than age 50 with a sensitive TSH test has been recommended to detect unsuspected hypothyroidism and hyperthyroidism, as the prevalence of overt thyroid dysfunction is 1.4% in this group.54 Furthermore, the yield of the TSH test for hypo- and hyperthyroidism is sufficiently high to warrant testing elderly people of either sex who present for medical care— especially those with a recent decline in ability to perform activities of daily living or cognitive deterioration—and at the time of admission to the hospital, psychiatric unit, or nursing home. In addition, the possibility of altered thyroid status should be considered any time an elderly patient's clinical status deteriorates without a clear explanation. In the nursing home population, routine screening and/or yearly monitoring of TSH levels have been recommended by some clinicians because of the high frequency of abnormalities in this group55,56 and 57; other physicians advocate an individualized approach to laboratory testing based on current health and functional status, and the wishes of the patient or surrogate decision maker regarding testing and treatment.58 DIAGNOSIS OF HYPERTHYROIDISM Highly sensitive TSH assays are adequate to screen for hyperthyroidism in relatively healthy elderly outpatients. The diagnosis should be confirmed using the free T4 test, however, because systemic illnesses, malnutrition, and some medications commonly used by older patients (glucocorticoids, dopamine agonists, phenytoin) may suppress TSH levels.59 Furthermore, many euthyroid patients with multinodular goiters have low TSH and normal T4 levels, suggesting subclinical hyperthyroidism. Most asymptomatic elderly patients with low serum TSH levels are euthyroid; the majority of these patients have isolated suppression of TSH with normal T4 and T3 levels and normal TSH on repeat testing 4 to 6 weeks later.59,60 Additional testing is needed to confirm the diagnosis of hyperthyroidism in some older patients. “T3 toxicosis” is more common with aging, especially in patients with toxic thyroid nodules. Patients with this disorder may exhibit a normal free T4 or free T4 index. In such cases, determination of free or total T3 level is needed to diagnose hyperthyroidism. In contrast to thyrotoxicosis in young patients, an elevated T3 level is specific but not sensitive for hyperthyroidism in an elderly patient. T3 levels are elevated in only half of elderly hyperthyroid patients, compared with 87% of younger patients.60 This is probably due to decreased T4 to T3 conversion secondary to normal aging and nonthyroidal disease in older people. As in younger patients, the high T4 syndrome (euthyroid hyperthyroxinemia) may also cause diagnostic confusion. Common causes of the high T4 syndrome in elderly patients include acute psychiatric illness, use of some drugs, or other conditions associated with an acute reduction in T4 to T3 conversion (acute fasting, use of b-adrenergic blocking agents or glucocorticoids), and states associated with increased TBG (estrogen or opiate use, hepatitis). Patients with high T4 syndrome have normal TSH levels and elevated total T4 levels, often with elevated TBG concentrations (e.g., in liver diseases such as chronic hepatitis). Free T4 levels may be elevated as well (e.g., in the setting of acute psychiatric illness). Abnormalities in thyroid function testing often resolve with treatment of the underlying condition or discontinuation of the responsible medication. As noted earlier, the TRH-stimulation test is generally not helpful in diagnosing hyperthyroidism in older people, because the increase in TSH levels after TRH administration is often blunted or absent even in healthy elderly persons. A finding of an increase of >3 µU/mL in TSH over baseline levels after TRH administration, however, excludes hyperthyroidism at any age. Thyroid scanning and measurement of radioactive iodine uptake (RAIU) are sometimes helpful in confirming the

diagnosis of hyperthyroidism and in defining the type of hyperthyroidism. DIAGNOSIS OF HYPOTHYROIDISM In diagnosing hypothyroidism, the serum TSH level is the most sensitive indicator of primary hypothyroidism in elderly as in younger adults. Although an elevated TSH level alone usually indicates primary hypothyroidism, TSH levels may be increased transiently during the recovery phase of an acute illness. Primary hypothyroidism should, therefore, be confirmed by the finding of a reduced free T4 index or level in association with an elevated TSH level, or a persistently elevated TSH level at a more distant time from the illness. Some elderly patients with serious nonthyroidal illnesses have a decreased serum free T4 index without elevation in serum TSH levels (low T4 syndrome). In this syndrome, levels of thyroid hormone–binding proteins are decreased, and with severe illness, T4 binding to TBG is decreased due to the presence of a T4-binding inhibitor. Serum free T4 levels are usually normal, and levels of rT3 are usually elevated. These patients do not appear to benefit from thyroid hormone replacement. However, two other clinical situations may present with decreased serum free T4 and free T4 index and relatively normal TSH levels. Elderly patients with primary hypothyroidism may have suppressed TSH levels from fasting, acute illnesses such as head trauma, and use of medications (dopamine, phenytoin, glucocorticoids). TSH levels are not usually suppressed into the normal range in these patients, however. The other clinical situation is secondary hypothyroidism, which is uncommon in the elderly as in younger adults. Unlike in the low T4 syndrome, in secondary hypothyroidism rT3 levels are typically decreased, and panhypopituitarism is usually present. If secondary hypothyroidism is suggested by an inappropriately normal to low TSH level together with a low free T4, measurement of serum testosterone and gonadotropin levels and/or an ACTH-stimulation test may be useful.

HYPERTHYROIDISM Hyperthyroidism (see Chap. 42) is common in elderly people. Fifteen percent of patients with thyrotoxicosis are older than age 60, and estimates of the prevalence of hyperthyroidism in this population range from 0.5% to 3%.36 Graves disease is the most common cause of hyperthyroidism in elderly people in the United States; however, toxic multinodular goiter and toxic adenomas are more common in older than in young adults. Many elderly patients with multinodular goiter have subclinical hyperthyroidism, with undetectable levels of TSH, free T4 and T3 levels within the normal range, and normal basal RAIU that cannot be fully suppressed with exogenous thyroid hormone. As noted earlier, atypical disease presentations are common in geriatric patients with hyperthyroidism (Table 199-2). As with hypothyroidism, the presentation of thyrotoxicosis in elderly patients is often vague, atypical, or nonspecific, with symptoms often occurring in the most impaired organ system. Classic findings such as goiter, nervousness, ophthalmopathy, hyperactive reflexes, increased sweating, tremor, and heat intolerance are much less common in elderly than in younger patients with hyperthyroidism, whereas tachycardia, muscle atrophy, anorexia, and atrial fibrillation (often without a rapid ventricular response) are more common manifestations.36,60,61 Hyperthyroidism is present in 13% to 30% of elderly patients with atrial fibrillation, and estimates are that a low TSH concentration confers a three-fold increase in the risk of developing atrial fibrillation within a decade after the TSH abnormality is detected.62 Furthermore, thyrotoxicosis in older patients may present with congestive heart failure. Unlike the situation with younger people, constipation is a more common presenting complaint than frequent bowel movements. Constitutional symptoms such as weight loss, weakness, nausea, and anorexia may be prominent; this may prompt clinicians to order unnecessary and extensive evaluations to rule out gastrointestinal malignancies. Other individuals present with apathetic hyperthyroidism, a common clinical manifestation in older patients but rarely if ever present in younger individuals. These patients demonstrate an absence of signs and symptoms of adrenergic hyperstimulation (e.g., tremor or hyperkinesis), a blunted affect or depression, and confusion or slowed mentation. Finally, hyperthyroidism increases the risk of developing osteoporosis and should be ruled out in those who present with decreased bone density. Not all studies, however, have found a relationship between low TSH levels and bone loss.63,64

TABLE 199-2. Clinical Manifestations of Hyperthyroidism in the Elderly

Although the issue is somewhat controversial,54,65 evidence exists that treatment of subclinical hyperthyroidism may be justified in elderly people, because of the increased risk of osteoporosis and cardiovascular complications.36,65 An accurate assessment of thyroid function is essential, however, before initiating such treatment. For now, data from randomized, controlled trials of treatment for subclinical hyperthyroidism are not available to clarify the risks and benefits of this practice. Administration of radioactive iodine (RAI) is the treatment of choice for thyrotoxicosis in most older patients. Patients with toxic multinodular goiter may require large or repeated doses of RAI. Treatment with antithyroid drugs such as propylthiouracil or methimazole is useful before RAI administration, both for symptom control and to reduce the likelihood of release of stored thyroid hormone and exacerbation of thyrotoxicosis from radiation thyroiditis. These drugs should be discontinued 1 week before RAI administration to optimize RAI uptake. In frail patients, an antithyroid drug can be resumed 3 to 5 days after RAI and continued for 1 to several months depending on the response to RAI. If not contraindicated, b-blockers can be given before and maintained after RAI for relief of symptoms such as tachycardia, tremor, and restlessness. After RAI treatment, T4 levels must be carefully followed for the emergence of hypothyroidism, or persistent or recurrent hyperthyroidism requiring retreatment. Moreover, the clearance rate of other medications may decrease after RAI treatment of the thyroid, such that drug levels may increase and dosage may need to be adjusted.

HYPOTHYROIDISM Hypothyroidism (see Chap. 45) is very common in the geriatric population. The reported prevalence of hypothyroidism in healthy elderly adults varies considerably depending on the population studied, with prevalence estimates of overt hypothyroidism ranging from 0.5% to 5% and of subclinical hypothyroidism, from 5% to 20% of adults older than age 60 years.36 Most hypothyroidism in older adults is due to chronic autoimmune thyroiditis, as in younger patients. The diagnosis of hypothyroidism is often overlooked in elderly people for several reasons. First, hypothyroidism usually has an insidious onset and a slow rate of progression. Second, physicians often fail to recognize typical clinical features of hypothyroidism that in older people are similar to “normal” age-related changes (e.g., dry skin, poor skin turgor, weakness, slowed mentation, constipation) or to manifestations of coexisting disease (e.g., anemia, congestive heart failure, diabetes). Finally, as in hyperthyroidism, atypical or uncommon clinical findings may be the dominant or sole presenting manifestations of hypothyroidism in elderly patients (Table 199-3). In large studies of older patients with overt hypothyroidism, the diagnosis of hypothyroidism was recognized on clinical examination in 50 years of age, although significant geographic variation is found.81 Commonly affected sites include the pelvis, spine, femur, and skull. Most people with Paget disease are asymptomatic, and the disorder is usually identified incidentally when radiographs are obtained for an unrelated indication, or an otherwise unexplained elevation in serum alkaline phosphatase is noted. In symptomatic patients, pain is the most common presenting symptom, usually localized to the affected bones. In one-third of cases, however, the pain is due to secondary osteoarthritic changes, often in the hips, knees, and vertebrae. Bony deformities occur in ~15% of patients at the time of diagnosis, usually involving the long bones of the lower extremities and often presenting as a bowing of the affected extremity. With involvement of the skull, compression of the eighth cranial nerve may result in sensorineural hearing loss. In patients with Paget disease of the hip joint, the results of total hip arthroplasty are comparable to those of hip arthroplasties performed in patients unaffected by Paget disease.82 The bisphosphonates are the current treatment of choice and are effective in suppressing the accelerated bone turnover and bone remodeling that is characteristic of this disease.

OSTEOPOROSIS Osteoporosis is a major cause of morbidity and mortality in older people. Estimates are that 1.5 million fractures annually are a direct result of osteoporosis. By the age of 90 years, ~32% of women and 17% of men have fractured a hip.83 Patients who fracture a hip are at a 10% to 20% increased risk of mortality over the following year84 and have increased morbidity, including institutionalization and impairment in mobility and functional status. Two clinical syndromes of osteoporosis have been proposed on epidemiologic and biochemical grounds.85 Type I (“postmenopausal”) osteoporosis is thought to occur in women typically between 51 and 75 years old and is characterized by an accelerated rate of bone loss, mainly in trabecular bone, with distal radius and vertebral fractures. The loss of the direct restraining effects of estrogen on bone-cell function is thought to be the most important mediator of this accelerated bone loss, leading to increased sensitivity of bone to PTH and increased calcium release from bone.86 PTH levels are slightly decreased at steady state in these patients. Type II (“senile”) osteoporosis is characterized by a late, slow phase of bone loss occurring in both men and women older than 70 years. This condition is associated with progressive secondary hyperparathyroidism and loss of both trabecular and cortical bone with vertebral and hip fractures. The proposal was originally made that type II osteoporosis is due primarily to decreased serum 1,25(OH)2D levels and calcium malabsorption, resulting in a secondary increase in PTH levels and bone resorption. Subsequently, however, the proposal has been put forward that both the secondary hyperparathyroidism and the decreased bone formation characterizing this slow phase of bone loss are manifestations of underlying estrogen deficiency in both elderly men and elderly women, with loss of estrogen action resulting in net calcium wasting and in an increase in the level of dietary calcium intake required to maintain bone homeostasis.86 Aging men have decreased circulating levels of both bioavailable estrogen and testosterone. Bioavailable estrogen levels appear to be a better correlate of bone mass than testosterone levels, supporting the hypothesis that estrogen deficiency is an important cause of bone loss in elderly men.86 As in younger patients, secondary causes of osteoporosis, osteomalacia, and primary hyperparathyroidism must be considered in the assessment of elderly osteopenic patients (Table 199-4). Osteomalacia due to vitamin D deficiency is relatively common in elderly people, particularly in those with a history of gastrectomy, chronic renal failure, malabsorption, or anticonvulsant use. Screening for osteomalacia is warranted in frail elderly people, including measurement of serum calcium and phosphate levels, which are decreased in vitamin D deficiency, and alkaline phosphatase level, which is increased. Findings of decreased 24-hour urinary calcium and serum 25-hydroxyvitamin D levels and increased PTH levels help to confirm the diagnosis.

TABLE 199-4. Causes of Osteopenia

MANAGEMENT OF OSTEOPOROSIS The management of osteoporosis and osteomalacia is discussed in detail in Chapter 64 and Chapter 63, respectively. The following points should be emphasized in the care of frail older patients with osteoporosis. First, older people at risk for falling should be identified. Important risk factors for falls in elderly adults include cognitive impairment, abnormalities of gait and balance, use of multiple medications, use of psychoactive medications, nocturia, disabilities of the lower extremities, and environmental factors such as household clutter. Second, a performance-oriented assessment of gait and balance should be undertaken. Direct observation of gait and balance is probably more useful than the standard neuromuscular examination in identifying patients with increased fall risk and treatable mobility problems.87,88 and 89 Assessment should include gait parameters (such as initiation of gait, step height, step length, step symmetry, path deviation, and trunk stability) and balance parameters (including ability to rise from a chair, immediate and sustained standing balance [with eyes open and closed], stability after sternal nudge, turning balance, and stability while looking upward or bending down). Management of patients at risk for falling may include treatment of underlying conditions contributing to falls (e.g., Parkinson disease, postural hypotension, foot disorders); minimization of medications increasing the risk of falling; gait retraining and strengthening; provision of assistive devices (e.g., a cane or walker); and environmental alterations (e.g., elimination of obstacles and clutter, provision of proper lighting and handrails).88 Secondary causes of osteoporosis, such as alcoholism, glucocorticoid excess, hypogonadism in men, hyperthyroidism, and multiple myeloma, should be sought and treated. For prophylaxis and for treatment of established osteoporosis, most women (and men) should ingest at least 1 to 1.5 g of elemental calcium per day beginning in the perimenopausal period, unless a contraindication such as a history of nephrolithiasis or hypercalciuria is present. Calcium carbonate is an inexpensive formulation that is acceptable for most patients, although calcium citrate is better absorbed by patients with achlorhydria. This should be accompanied by a daily multivitamin containing 400 to 800 IU of vitamin D. Intake of >1000 IU per day may cause increased bone resorption, although if biochemical evidence of osteomalacia is found, higher doses of vitamin D may be needed. Also, exercise is important.89a Aside from calcium and vitamin D supplementation, estrogen-replacement therapy has been considered by some to be the treatment of first choice for established postmenopausal osteoporosis, based on long-term experience, well-documented effectiveness in preventing bone loss, and the potential for other benefits aside from its effects on bone. However, relatively few data are available on the effects of estrogen use on rates of fracture (especially hip fracture) in postmenopausal women.90 Estrogens act primarily to prevent bone resorption, but the ability to restore lost bone mass may be minimal. Furthermore, evidence exists that the beneficial effects of estrogen replacement are more marked in women who begin treatment within five years after menopause.90 Therefore, maximum benefit from estrogen therapy is obtained by beginning use as soon as possible after menopause. Estrogens also play an important role in the treatment of type II osteoporosis,91 however, and other studies have reported similar benefits from estrogen use on bone density, both in women older than age 70 and in younger women.83 The optimal duration of estrogen use has not been established, but because cessation of estrogen treatment leads to resumption of bone loss, continuing treatment until age 65 to 70 or longer in those without contraindications to their use may be reasonable. Alternatively, the suggestion has been made that starting estrogen use at age 65 may provide almost as much protection against osteoporotic fractures as starting at menopause, and may reduce the risks of long-term estrogen therapy (see Chap. 100 and Chap. 223 for

discussions of the risks).92,93 The selection of estrogen replacement regimens for elderly women is discussed later in the Menopause section. The bisphosphonates alendronate and etidronate have been demonstrated to decrease the fracture rate in women with postmenopausal osteoporosis and are effective alternatives to estrogen therapy, especially for women who are concerned about the potential adverse effects of estrogens. Continuous high-dose etidronate may lead to impaired bone mineralization; therefore, etidronate must be given intermittently at a lower dosage (e.g., 400 mg per day for 2 weeks every 3 months).94 An additive benefit of combined intermittent etidronate and cyclical estrogen and progesterone replacement on hip and spine bone mineral density was demonstrated over a 4-year period in postmenopausal women with established osteoporosis.95 Low-dose alendronate (5 mg per day) prevents bone loss in postmenopausal women without established osteoporosis.96,97 This approach is appropriate for those women who are at high risk for future fractures and who are unable or unwilling to take estrogen.98 The efficacy and safety of other bisphosphonates (e.g., risedronate) in the prevention and treatment of osteoporosis are currently under investigation. Calcitonin may be less effective in the prevention of bone loss in osteoporotic patients than estrogens or bisphosphonates, and its long-term efficacy in fracture prevention has not been as well documented. Raloxifene, a selective estrogen-receptor modulator, has been shown to increase bone density in postmenopausal women,99 although to a smaller extent than estrogen replacement, and data regarding reduction in fractures are limited.100 Although raloxifene does not appear to stimulate the endometrium in these women, it does decrease total and low-density lipoprotein (LDL) cholesterol levels. Unlike estrogens, however, raloxifene does not increase high-density lipoprotein (HDL) levels, and raloxifene is associated with an increased incidence of hot flushes, leg cramps, and thromboembolic events. Furthermore, no long-term data are available to assess the effects of raloxifene on the incidence of coronary heart disease (CHD) events, cognitive function, or the incidence of breast, ovarian, and uterine cancers.100 At present, recommendation of other potential osteoporosis treatments such as fluoride, androgenic steroids, and parathyroid hormone injections is premature until studies are available that document improvements in osteoporotic fracture rates without significant adverse effects.

HYPERCALCEMIA PRIMARY HYPERPARATHYROIDISM Primary hyperparathyroidism occurs most commonly in adults between 45 and 60 years of age, although it may develop at any age. The condition occurs more often in women than in men by a ratio of nearly 3:1. The annual incidence of the disease is ~1 per 1000.101 The diagnosis is often suspected based on a finding of elevated serum calcium levels on routine laboratory testing; indeed, asymptomatic hyperparathyroidism is by far the most common presentation.102 Compared with younger patients, however, elderly people with primary hyperparathyroidism more often present with neuropsychiatric and neuromuscular symptoms, and with osteoporosis associated with fractures.103 Other common complaints in older patients include altered mental status, fatigue, depression, weakness, personality change, memory loss, anorexia, and constipation. The approach to the diagnosis and treatment of elderly patients with primary hyperparathyroidism is similar to that of younger patients (see Chap. 58).102 Serum calcium level is high and phosphate level is often low to low normal, whereas alkaline phosphatase level is often high normal or mildly increased. Assays for intact PTH are the diagnostic tests of choice, and the presence of primary hyperparathyroidism is established by a high-normal or elevated PTH level in the setting of hypercalcemia. (Nonparathyroid causes of hypercalcemia are associated with undetectable or clearly decreased PTH levels. The most common cause of hypercalcemia in hospitalized patients is a malignancy producing a PTH-related protein [PTHrP] that can be measured in reference laboratories.) The skeletal effects of primary hyperparathyroidism are selective, with a reduction in cortical bone but relative protection against cancellous bone loss. Accordingly, bone densitometry is an important part of the evaluation of these patients, and measurements at the forearm and hip are the best indicators of cortical bone density (the latter is the most important site because of its significance as a risk factor for hip fracture). Surgery is the only definitive treatment and is indicated in surgical candidates with markedly elevated serum calcium levels (>12 mg/dL), overt manifestations of primary hyperparathyroidism (e.g., nephrolithiasis), marked hypercalciuria, and markedly reduced cortical bone density. For the patients who are managed conservatively, checking serum calcium levels every 6 months and monitoring 24-hour urine calcium excretion, creatinine clearance, and bone densitometry annually is appropriate. These patients should be instructed to avoid thiazide diuretics and to avoid dehydration, but avoidance of dairy products is unnecessary. Most patients followed expectantly remain stable over time. Medical therapy for hyperparathyroidism may include administration of estrogens in women, oral phosphate in patients with low serum phosphate levels, and bisphosphonates. HYPERCALCEMIA OF MALIGNANCY In most patients with malignancy-related hypercalcemia, an obvious neoplasm is evident on examination and routine diagnostic evaluation. Diagnostic possibilities include humoral hypercalcemia of malignancy, in which a PTHrP is produced by a cancer, usually of squamous cell type (e.g., lung, head, and neck), that occurs commonly in the elderly. In addition, multiple myeloma and some lymphatic tumors secrete osteoclast-activating factors, many of which are cytokines (e.g., lymphotoxin, interleukin-1, tumor necrosis factor). Treatment of hypercalcemia of malignancy includes volume repletion followed by immediate forced diuresis with saline infusion and furosemide. Parenteral bisphosphonates (e.g., pamidronate), calcitonin, or mithramycin should also be given, and the underlying malignancy should be treated if possible. Glucocorticoid therapy is reserved primarily for myeloma and lymphatic tumors. In elderly patients with advanced malignancies, short life expectancy, and poor functional status, not treating the hypercalcemia may be appropriate and may provide a more comfortable mode of exit for these terminally ill patients.

ADRENAL CORTEX PHYSIOLOGY Decreased cortisol production is offset by decreased cortisol clearance, resulting in unchanged basal serum cortisol levels with aging. Urinary free cortisol levels are the same in elderly persons as in young adult subjects.104 Stimulation of cortisol secretion by exogenous ACTH is unaltered with aging.105 Furthermore, cortisol and ACTH responses to metyrapone, insulin-induced hypoglycemia, ovine CRH, and perioperative stress are normal or slightly prolonged in elderly subjects,106,107 indicating an intact HPA axis responsiveness to stimulation with aging. In addition, ACTH pulse frequency is similar in healthy young and healthy elderly men, suggesting that baseline hypothalamic regulation of glucocorticoid function is intact with aging.12 Clear evidence now exists, however, that feedback sensitivity to glucocorticoids decreases with aging.15,108 Although the clinical implications of this decreased responsiveness to glucocorticoid feedback inhibition are uncertain, some have hypothesized that decreased negative feedback results in prolonged glucocorticoid exposure; this, in turn, damages hippocampal neurons regulating glucocorticoid secretion, leading to additional glucocorticoid hypersecretion and further damage to mechanisms regulating glucocorticoid feedback inhibition.34,109 This process may be involved in mediating a “glucocorticoid cascade” of neurodegeneration in Alzheimer disease and, to a lesser extent, in the normal aging brain.110 Thus, although the issue is still controversial, age-related glucocorticoid dysregulation may be a potentially modifiable risk factor for Alzheimer disease. LABORATORY DIAGNOSIS OF ADRENOCORTICAL DISEASE IN OLDER PATIENTS Adrenal hyperfunction and hypofunction are less common in elderly than in middle-aged adults. However, manifestations that are associated with either adrenal hyperfunction (e.g., hypertension, obesity, and diabetes mellitus) or adrenal insufficiency (e.g., orthostatic hypotension and weight loss) occur more commonly in older than in young adults. Therefore, adrenal disease must be considered in the evaluation of elderly patients with these manifestations, and patients with suggestive findings on physical examination or laboratory screening should undergo further assessment. In addition, benign adrenal masses are common incidental findings observed during imaging procedures of the abdomen. Benign adenomas usually range in size from 1 to 6 cm and weigh 10 to 20 g, whereas malignant tumors generally weigh >100 g.106,111 Incidentally detected adrenal masses of 6 cm should probably be removed in patients who are appropriate candidates for surgery, although the natural history of incidentally discovered adrenal lesions in the elderly is unknown. Several factors may interfere with testing of HPA axis function in the elderly. First, the excretion of steroids commonly measured in urine is decreased with renal impairment, and measurements are unreliable if the creatinine clearance is 5% of bone mass compared to baseline. Calcium supplementation with 1500 mg per day of elemental calcium, together with vitamin D, 800 IU per day, should be ensured to minimize bone loss caused by a corticosteroid-induced decrease in intestinal calcium absorption and an increase in urinary calcium losses. The exogenous administration of corticosteroids induces suppression of gonadotropins and sex steroids; therefore, sex hormone–replacement therapy should be considered unless contraindicated. For postmenopausal women, appropriate hormone-replacement regimens are the same as for women not receiving corticosteroids. Men with low testosterone levels should also be given hormone replacement, either with testosterone enanthate or testosterone cypionate, 100 to 200 mg intramuscularly every 2 weeks, or with daily transdermal application of a testosterone patch. The bisphosphonate alendronate (5 or 10 mg per day) or etidronate (administered cyclically, 400 mg per day for 14 days every 3 months) are effective in the primary and secondary prevention of bone loss in patients receiving glucocorticoid therapy.114,115 and 116 Whether bisphosphonate treatment reduces fracture risk in patients with established corticosteroid-related osteoporosis is not yet clear, however. Calcitonin is also effective in the prevention and treatment of glucocorticoid-induced bone loss.113,117 The restriction of dietary sodium and the administration of thiazide diuretics are useful to decrease corticosteroid-induced hypercalciuria, although the effects of these interventions on bone density have not been fully investigated. HYPOADRENOCORTICISM Hypoadrenocorticism is covered in detail in Chapter 76. As in younger adults, iatrogenic adrenal failure secondary to long-term glucocorticoid administration is the most common cause of hypoadrenocorticism in elderly people. Only very uncommonly does autoimmune adrenocortical insufficiency present initially in an elderly patient. Some nonautoimmune causes of adrenal insufficiency occur more commonly in older adults, however, including tuberculosis, adrenal hemorrhage in patients taking anticoagulants, and metastatic involvement of the adrenals.106 Some elderly patients with chronic adrenal insufficiency (e.g., secondary to hypopituitarism) present with nonspecific symptoms of “failure to thrive”—such as weight loss, anorexia, weakness, and decreased functional status. Moreover, one-third of older patients with adrenal insufficiency do not have hyperkalemia at initial presentation. Of note, compared with adrenal insufficiency in younger adults, adrenal insufficiency in the elderly has historically been more often fatal and more commonly diagnosed only at autopsy.118 Therefore, a high index of suspicion is required to detect this treatable, life-threatening problem in many older patients. Recovery of HPA axis responsiveness after cessation of glucocorticoid therapy is variable and in some individuals may not be complete even after several months. A number of factors may put older adults at higher risk to develop iatrogenic adrenal insufficiency. Elderly people who are on complicated medication regimens or who are cognitively impaired may become confused about their medications or forget to take them. High medication costs or medication-related side effects may cause older patients to discontinue their medicines abruptly without consulting their physicians. In addition, as noted earlier, the clinical manifestations of adrenocortical insufficiency are often nonspecific. When adrenocortical insufficiency due to cessation of long-term glucocorticoid therapy is suspected in an older person, an appropriate course is to perform the ACTH-stimulation test and institute therapy. As with younger adults, older people with persistent adrenocortical insufficiency should be given glucocorticoid replacement and coverage for major surgery and other stressful events until HPA axis function has recovered. ALDOSTERONE AND RENIN Aldosterone secretion and clearance rates decrease with aging. In contrast to cortisol levels, however, basal plasma aldosterone levels are not maintained at normal levels in elderly subjects but decline by ~30% in healthy octogenarians compared with younger adults.119 In response to dietary sodium restriction, aldosterone secretion increases three-fold in the young but only two-fold in older adults. These changes in aldosterone secretion with aging are thought to be due to corresponding reductions in plasma renin activity, which is decreased in the basal state and in response to sodium restriction or upright posture.120 Furthermore, conversion of inactive to active renin is thought to be impaired with aging.121 Age-related increases in atrial natriuretic hormone (ANH) secretion also contribute to the age-related decrease in aldosterone secretion by directly inhibiting aldosterone release and by inhibiting renal renin secretion, plasma renin activity, and angiotensin II levels. In addition, renal responsiveness to ANH is increased with aging, suggesting that ANH is an important contributor to age-related renal sodium losses.16 Declining aldosterone levels with aging predispose older people to renal salt wasting. Several other factors place elderly people at higher risk for volume depletion and dehydration, including decreased thirst sensation, increased ANH levels and renal ANH sensitivity, decreased renal ADH responsiveness, and possibly decreased renal-tubular responsiveness to aldosterone.16 In addition, age-related hypoaldosteronism is characterized by a decreased aldosterone response to hyperkalemia, which may contribute to an increased susceptibility to hyperkalemia if other potassium regulatory systems fail.122 Accordingly, elderly patients with hyporeninemic hypoaldosteronism are at higher risk of becoming hyperkalemic during treatment with potassium-sparing diuretics, b-blocking agents, or nonsteroidal antiinflammatory drugs. Older diabetic patients with renal insufficiency are particularly susceptible to this complication. ADRENAL ANDROGENS Adrenal androgen secretion declines progressively beginning in the third decade, with plasma levels of the principal adrenal androgen, dehydroepiandrosterone (DHEA), declining to just 10% to 20% of young adult levels by the eighth and ninth decades.123,124 This age-related decrease in DHEA levels is due to a reduction in adrenal DHEA secretion rather than to an increase in DHEA metabolism. Furthermore, the DHEA response to adrenal stimulation by ACTH is markedly decreased in older people. However, levels of DHEA and its sulfate, DHEAS, exhibit marked interindividual variability at all ages. Epidemiologic data suggest that DHEA levels may play an important role in the deterioration of a variety of physiologic functions with aging. For example, low DHEA levels have been reported in states of poor health (e.g., after surgery or accidents) or in the setting of immunologic dysregulation (e.g., active rheumatoid arthritis or acquired immunodeficiency syndrome).125 Other studies have reported positive correlations between plasma DHEA levels and vigor and longevity, and inverse correlations with cancers and cardiovascular disease.126,127 Furthermore, high-functioning community-dwelling elderly subjects have higher levels of DHEAS than low-functioning subjects,128 such that DHEAS levels are linked to functional status. As a result of these associations, considerable interest has been generated in the potential therapeutic effects of DHEA administration in older adults. Data from randomized controlled trials in which DHEA was administered in dosages of 50 to 100 mg per day over a 6- to 12-month period have shown subjective improvements in physical and psychological well-being, increased serum IGF-I levels and, at higher dosages, increased lean body mass and muscle strength at the knee.129 Levels of circulating lipids, glucose, and insulin as well as bone density were comparable in treated and placebo groups. Concerns remain, however, regarding the potential for androgenization in women, gynecomastia in men, possible adverse effects on lipoprotein metabolism at supraphysiologic doses, and potential hepatotoxicity.130,131 Furthermore, DHEA is metabolizable to estrogens and to androgens, including testosterone and dihydrotestosterone,125,129 and its effects on the risk of breast cancer in women and of prostate cancer in men have not been determined. Thus, although these data are potentially promising, the long-term safety and efficacy of DHEA treatment have not been established. CATECHOLAMINES Norepinephrine (NE) is the principal neurotransmitter released by sympathetic postganglionic neurons. After release, most NE is taken up again into the axon

terminals, whereas only a small fraction is released into the circulation. Substantial evidence indicates that sympathetic nervous system (SNS) activity is increased with aging in humans. Basal plasma NE levels, and the NE secretory response to various stimuli such as upright posture and exercise, are increased with aging.132,133 In contrast, circulating epinephrine levels and epinephrine responses to various stimuli exhibit little change with aging. Although sympathetic tone is increased with aging, physiologic responsiveness to both a- and b-adrenergic receptor–mediated stimulation appears to decrease with aging.134 Decreased catecholamine responsiveness is thought to be due to changes at both the receptor and the postreceptor level. The clinical effects of this age-related increase in sympathetic tone may include the development of hypertension. A number of studies have reported a correlation between hypertension and increased plasma NE levels in elderly people.135 Body fat content is independently associated with both aging and plasma NE levels, however, suggesting that obesity may contribute to hypertension by increasing sympathetic tone, independent of the effects of aging per se. Certain diseases commonly associated with aging may give rise to autonomic insufficiency with orthostatic hypotension, including Parkinson disease, multiple system atrophy, and diabetes mellitus. Moreover, certain drugs that interfere with SNS function may also cause orthostatic hypotension; these include antihypertensive agents such as clonidine and a-methyldopa and psychoactive drugs such as phenothiazines and tricyclic antidepressants. Other factors such as volume depletion, prolonged bed rest, and venous insufficiency may also cause or exacerbate postural hypotension. The management of orthostatic hypotension includes the treatment of hypovolemia, discontinuation of medications that may exacerbate postural hypotension, and instructing the patient to sit with legs dangling for several minutes before getting out of bed, to elevate the head of the bed at night, and to use elastic support stockings and an abdominal binder to promote venous return. Useful medications include caffeine, fludrocortisone to expand plasma volume, and midodrine, a sympathomimetic amine. Postprandial hypotension is a common disorder involving SNS dysfunction in elderly people. It possibly results from inadequate SNS compensation for pooling of blood in the splanchnic vessels after a meal, impaired baroreceptor reflex function, impaired peripheral vasocontriction, release of vasoactive gastrointestinal peptides, and/or inadequate postprandial increases in cardiac output.136 This condition is especially common in elderly hypertensive patients. Postprandial hypotension may be an important cause of syncope in elderly patients and should be considered in the evaluation of older people with unexplained syncope.137 The management of this condition should include the avoidance of dehydration, discontinuation of unnecessary drugs that could exacerbate postprandial hypotension, consumption of frequent small meals, avoidance of alcohol, and avoidance of strenuous exercise within 2 hours after meals. Caffeine use has not been shown to be helpful in this condition.

FEMALE REPRODUCTIVE AND ENDOCRINE FUNCTION: MENOPAUSE The mean age for menopause in women, 51 years, has not changed significantly over the last century. Because life expectancy for women has increased markedly over the same period, however, most women can expect to spend more than one-third of their lives in the postmenopausal state. Accordingly, clinicians caring for perimenopausal and postmenopausal women must work closely with them to consider the potential impact of menopausal changes on future health and functional status. Failure of ovarian end-organ function occurs by the fifth or sixth decade, with cessation of ovarian follicular development, estradiol secretion, and menstruation, and unresponsiveness to gonadotropin stimulation. Premenopausally, estradiol (E2) and estrone (E1) are secreted by maturing ovarian follicles and are produced by aromatization of ovarian and adrenal androgenic precursors in peripheral tissues. An elevation in serum FSH levels heralds the onset of menopause and is the most sensitive clinically available indicator of cessation of ovarian follicle development.138 As noted in the discussion of gonadotropins earlier, levels of inhibin B appear to decline slightly before FSH levels increase,39 but an inhibin-B assay is not yet clinically available. After menopause, circulating E2 and E1 are derived almost entirely from aromatization of adrenal androstenedione, and E1 levels are higher than those of E2. FSH levels increase to a greater extent than LH levels. Obese postmenopausal women exhibit increased production of adrenal androgenic precursors and increased peripheral aromatization of androgens to estrogens, especially to E1, which is associated with a decreased risk of osteoporosis. The manifestations of estrogen deficiency experienced by menopausal women are listed in Table 199-5. Vasomotor symptoms occur in three out of four women during the perimenopausal period139 and may be associated with other symptoms such as palpitations, faintness, fatigue, and vertigo. These symptoms, along with atrophy of the sexual tissues, are relieved by estrogen replacement. Clonidine, methyldopa, or medroxy-progesterone may occasionally be effective for treating hot flushes in women who are unable to take estrogens.

TABLE 199-5. Manifestations of Estrogen Deficiency in Menopausal Women

As discussed earlier, long-term treatment with estrogens in postmenopausal women may prevent or delay the occurrence of osteoporosis and fractures. Furthermore, the incidence of coronary artery disease rises sharply in postmenopausal women, and some evidence from observational studies has suggested that estrogen replacement in postmenopausal women decreases the risk of cardiovascular disease by up to 50% compared with those not receiving estrogens.83,140 This apparent cardioprotective effect of estrogens has been attributed to favorable effects on lipid metabolism, including decreased serum LDL and increased HDL cholesterol levels, and to other effects of estrogens on clotting mechanisms, glucose regulation, blood vessels, and myocardial tissue.141 Cardiac catheterization studies have reported that women receiving estrogen replacement develop less severe coronary atherosclerosis, and that the greatest benefit of estrogens on coronary disease is realized by women with more severe atherosclerosis at the time of the initial catheterization.142,143 The cardiovascular benefits of estrogen-replacement therapy have been called into question, however, by the results of the Heart and Estrogen/Progestin Replacement Study (HERS)—a randomized, placebo-controlled trial of hormone replacement for secondary prevention of CHD in postmenopausal women—which found no reduction in the rate of coronary events over a period of >4 years, with a tendency toward a higher coronary event rate in the treatment group in the first year and fewer events in years 4 and 5.144 Furthermore, thromboembolic events were increased within the first year of therapy. Based on these data, insufficient evidence exists at this time to support the initiation of hormone-replacement therapy specifically to prevent coronary events in women with established CHD. Estrogen replacement possibly may be beneficial in the primary prevention of coronary events; however, data from randomized, controlled trials addressing this issue will not be available until the results of studies such as the Women's Health Initiative and the Women's International Study of Long Duration Oestrogen after Menopause (WISDOM) are completed.145 Furthermore, given the trend toward a cardiovascular benefit after several years of treatment in the HERS study, continuing therapy for this indication in women who are already receiving replacement may be appropriate. Nevertheless, even long-term therapy may be associated with some risks. One large prospective observational study reported that overall mortality was reduced by 37% in women currently taking hormone replacement, but the magnitude of this apparent benefit decreased to 20% in those receiving therapy for >10 years.146 Some, but not all, evidence suggests that estrogen use may improve cognitive function in postmenopausal women with Alzheimer disease (AD). Retrospective studies have found an inverse correlation between the dosage and duration of estrogen-replacement therapy and the incidence of AD,147,148 and estrogen administration in small open trials appeared to improve cognition and mood in women with AD.149,150 Moreover, some aspects of cognitive function were found to improve with transdermal estrogen replacement in a small double-blind, placebo-controlled trial of postmenopausal women with AD.151 Larger prospective studies are needed, however, to document any alleged therapeutic role for estrogens in the treatment of AD. Established risks associated with estrogen replacement include a nearly six-fold increase in the incidence of endometrial cancer in those receiving unopposed estrogen therapy. Coad-ministration of a progestin protects against this complication, however. The lifetime risk of developing breast cancer appears to be 10% to 30% higher in postmenopausal women receiving estrogen-replacement therapy for >10 years, and this risk must be weighed against the potential benefits of hormone

replacement.141 Hypercoagulability is a potential side effect of estrogen therapy. Estrogen use may also predispose to cholelithiasis and hypertriglyceridemia. Table 199-6 summarizes the contraindications to estrogen-replacement therapy.

TABLE 199-6. Contraindications to Postmenopausal Hormone Replacement

In summary, further clarification of the risks and benefits of postmenopausal hormone-replacement therapy must await the completion of additional prospective, randomized, controlled trials such as the Women's Health Initiative. For the time being, the beneficial effects on bone density and urogenital atrophy, and—in the immediate postmenopausal period—the amelioration of vasomotor symptoms such as hot flushes, are the primary indications for hormone-replacement therapy in some postmenopausal women. An estrogen-progestin combination is appropriate for women with an intact uterus to decrease the risk of endometrial cancer, although inclusion of a progestin may partially attenuate some of the favorable effects of estrogens on lipids, and the effects of progestins on cardiovascular risk are uncertain. Estrogens should be used alone in women who have had a hysterectomy. Many older women may prefer continuous rather than cyclic hormone administration, because objectionable resumption of menses often occurs with cyclic therapy. An appropriate continuous combined hormonal regimen for many of these women is 0.625 mg daily of an oral conjugated estrogen, together with 2.5 mg daily of medroxyprogesterone acetate. Potential disadvantages to this approach include the possibility of irregular bleeding and the relative lack of information on the effects of continuous progesterone on reduction in endometrial cancer risk and serum lipid levels. Furthermore, some women may experience weight gain and depression with daily progesterone administration. Patients should be informed that spotting often occurs during the first few months on this regimen, but that most women become amenorrheic after a year of treatment. For women who find a cyclic regimen suitable, 0.625 mg of conjugated estrogen may be used daily, with addition of a 10- to 14-day course of medroxy-progesterone acetate, 5 to 10 mg daily, every 3 or 4 months to induce withdrawal bleeding. This approach has the disadvantage of a greater likelihood of inducing menses, but menses will occur at a predictable time, and more is known about the effects of cyclic hormone regimens on uterine cancer risk and lipids. With either approach, very gradual introduction of conjugated estrogen may reduce unpleasant symptoms such as breast swelling, beginning with 0.3 mg every other day, advancing to 0.3 mg per day after 1 month and 0.625 mg per day the following month. Alternatives to conjugated estrogens include oral, transdermal, and vaginal estradiol preparations. Alternatives to estrogen replacement are available for women who are unable or unwilling to take estrogens—specifically, the selective estrogen-receptor modulators such as raloxifene. As noted earlier, raloxifene has beneficial effects on bone mass and serum lipid levels. The increase in bone mass, however, may be less with raloxifene than with estrogen therapy, and although the reduction in total and LDL cholesterol levels is similar with estrogens and raloxifene, the latter does not increase HDL cholesterol levels.100 The effects of selective estrogen-receptor modulators on cardiovascular mortality and mortality due to all causes are unknown. In contrast to estrogens, raloxifene commonly produces hot flushes and is, therefore, not indicated for the treatment of vasomotor symptoms of menopause. Raloxifene has estrogen-antagonist effects on breast and uterine tissues, and does not cause endometrial hyperplasia. Tamoxifen, another estrogen antagonist, reduces the risk of invasive breast cancer in women who are at increased risk for breast cancer or who have a history of breast carcinoma in situ.152 If raloxifene is shown to have similar effects, it may be particularly useful in women with risk factors for breast cancer or a history of the disease.

MALE REPRODUCTIVE AND ENDOCRINE FUNCTION HORMONAL CHANGES The age-related changes in reproductive function in men are less dramatic than those that occur in aging women. In aging men, reproductive changes occur gradually, exhibit considerable variation among individuals, and usually do not result in severe hypogonadism. A modest degree of primary testicular failure is evident in many healthy elderly men, as evidenced by decreases in daily sperm production, diminished total and free testosterone levels, and reduced testosterone responses to exogenous gonadotropin administration, together with increased serum gonadotropin levels.42,153 Some medical illnesses and malnutrition may further impair testicular function, as can some medications (Table 199-7).153a In more frail older men, testicular failure is extremely common. For example, among male nursing home residents, 45% of individuals have been reported to exhibit testosterone levels within the hypogonadal range.154 Some men, however, maintain serum testosterone levels within the normal range even after age 80.42,153 In addition to primary testicular failure, subtle age-related changes occur in hypothalamic-pituitary control of testicular function, including decreased gonadotropin responsiveness to exogenous GnRH administration42 and decreased LH pulse frequency in some healthy aging men.41 Furthermore, many healthy older men have inappropriately normal gonadotropin levels (i.e., not elevated above the normal range) in the presence of low testosterone levels, suggesting secondary testicular failure.

TABLE 199-7. Medications Associated with Decreased Serum Testosterone Levels and Hypogonadism

A small number of aging men exhibit more obvious testicular failure, with total testosterone levels clearly below the lower limits of the normal range and clear manifestations of androgen deficiency (e.g., decreased libido and potency, osteoporosis, gynecomastia, and hot flushes). In the absence of contraindications, androgen replacement is indicated for these older men, as for young men. More often, however, the clinician is faced with an older patient with slightly decreased serum testosterone levels (e.g., 2.5 to 3.0 ng/mL) and nonspecific symptoms that may include impotence, loss of libido, muscle weakness, or osteopenia, and whether these men should be treated is unclear. In young hypogonadal men, androgens are important for the maintenance of normal bone and muscle mass, sexual drive, and erectile function.42 The hypothesis has been raised that age-related androgen deficiency may contribute to declining muscle and bone mass and other concomitants of aging, but few randomized, controlled studies have been performed to determine whether androgen supplementation in older men is beneficial and whether the benefits outweigh the risks. Two reports involving a small number of older hypogonadal men found that testosterone administration for 3 months increased lean body mass, reduced biochemical indices of bone turnover, and increased libido, without adverse effects on lipid metabolism or symptoms of prostatism.155,156 Another study of testosterone replacement over 12 months in older men with low testosterone levels found increases in grip strength but failed to demonstrate changes in body composition.157 Several of the subjects were withdrawn from treatment, however, due to elevations in hematocrit. Finally, visceral fat mass and fasting blood glucose levels declined in middle-aged men receiving testosterone.158 Although some placebo-controlled studies have shown improvements in some measures of strength with testosterone therapy, no data are yet available to show whether testosterone improves functional performance or quality of life in elderly men.159 At the moment, larger and longer term studies are needed to determine the risks and benefits of androgen-replacement therapy before this approach can be recommended for aging men

with slightly reduced testosterone levels. Androgen-replacement therapy is discussed in detail in Chap. 119. SEXUAL ACTIVITY AND ERECTILE DYSFUNCTION In general, sexual activity and libido decline with aging, although some healthy elderly men exhibit stable or increased sexual desire with aging.160,161 Important determinants of sexual behavior in elderly men include perceived health status and the level of sexual activity during the younger years. Kinsey estimated the prevalence of impotence to be 55% by the age of 75 years.162 In contrast to the earlier view that most impotence is psychogenic, erectile dysfunction is now thought to have an organic basis in the large majority of cases. Furthermore, the prevalence of organic causes increases with advancing age.163 Important contributors to impotence in older men include arterial and venous abnormalities, neuropathies, use of medications, and coexisting medical illnesses. The most common cause of erectile dysfunction in older men is vascular disease, with half of all men older than age 50 exhibiting evidence of impaired penile blood flow. Venous insufficiency (failure to occlude venous outflow) may occur due to leakage, arteriovenous malformations, or increased shunting between the corpora cavernosa and the glans. Although overt hypogonadism is present in 50% of patients by causing a reduction in bone matrix deposition. The osteoporosis is thought to be multifactorial, with hypogonadism, elevated cortisol levels, poor nutrition with inadequate intake of calcium and vitamin D, and excess physical activity contributing to the clinical picture. Hypoestrogenemia may be a major contributing factor in the development of bone loss. The degree of spinal osteopenia seen in patients with anorexia nervosa, however, is greater than that seen in other states of comparable hypoestrogenemia (i.e., hyperprolactinemia and hypothalamic amenorrhea). Studies have failed to show a significant reversal or prevention of bone loss by estrogen/progesterone supplementation combined with calcium supplementation71; this contrasts with their marked efficacy in the treatment of postmenopausal women. Although an inadequate intake of calcium and vitamin D can lead to osteopenia, calcium intake and vitamin D levels do not correlate with bone density in anorexia nervosa,72 and calcium supplementation has failed to lead to increased bone density in this patient group.71 Prolonged undernutrition is likely a major factor in the development of bone loss. Bone density correlates with body composition indices such as body mass index (BMI) and fat mass in women with anorexia nervosa. IGF-I is a nutritionally regulated hormone with potent effects on bone formation. The IGF-I levels, which are reduced and correlate with bone loss in anorexia, increase with weight gain.49,73 IGF-I may be important for bone formation 49 and may be useful in the treatment of osteoporosis in anorexic patients. Plasma levels of the adrenal androgen dehydroepiandrosterone (DHEA), which can also be converted to estrogen, decrease in patients with anorexia nervosa.74 These patients may have decreased adrenal 17–20 lyase activity, contributing to increased cortisol and decreased DHEA production.75,76 As estrogens tend to inhibit bone resorption whereas androgens promote bone formation,77 this would indicate that low DHEA levels may play a role in the development of osteoporosis in this patient group. Indeed, one study74 indicates that DHEA treatment decreases N-terminal cross-linked telopeptide of type I collagen (a marker of bone resorption), while increasing osteocalcin (a marker of bone formation). LEPTIN Leptin is a protein product of the ob gene in mice (see Chap. 186). It is produced by adipose tissue and appears to convey information regarding the bodily adipose tissue mass to neural systems in the hypothalamus that are involved in the regulation of food intake (see Chap. 125, Chap. 126 and Chap. 186). The leptin level is positively correlated with the percentage of body fat in both obese individuals and those of normal body weight. Serum leptin levels are low in patients with anorexia nervosa48 (in apparent relationship to BMI). This correlation becomes uncoupled at extremely low BMIs, however, indicating a possible threshold effect. Although one study78 showed that short-term refeeding (3 days) of anorexic patients who had very low BMIs did not increase leptin levels, more long-term nutritional rehabilitation and weight gain leads to increased serum leptin. Higher leptin levels have been found in weight-recovering patients than in comparable controls.48 These findings suggest that factors in addition to body fat mass have a role in regulating the serum leptin level. Indeed, insulin, cytokines, and possibly GH and corticosteroids may stimulate the release of leptin independently of body fat.79,80, 81,82,83 and 84 The relatively higher level of leptin seen in patients with anorexia nervosa has been hypothesized to contribute to the difficulties these patients have with weight restoration and maintenance. Leptin appears to affect several neuroendocrine mechanisms, and animal studies have shown that, during starvation, leptin plays a role in activating the HPG and thyroidal axes (while depressing the HPA axis).85(See also ref. 84a.) Some have proposed that the decreased leptin levels observed in anorexic patients may be responsible for the amenorrhea observed in these patients. Indeed, all weight-recovered anorexic patients with normal menstrual cycles have leptin levels above a certain threshold.86 However, leptin levels are no different in weight-recovered amenorrheic and eumenorrheic patients. This indicates that leptin is a necessary, but not a sufficient, factor for the resumption of menses in patients with anorexia nervosa.87 CENTRAL NEUROPEPTIDE SYSTEMS Appreciation of the richness and complexity of the CNS regulation of food intake has increased profoundly. Much of this regulation takes place within the hypothalamus; the arcuate nucleus, ventro-medial nucleus, lateral hypothalamic area, and paraventricular nucleus play major roles. Multiple neuropeptide systems appear to have a part in food intake regulation; these include proopiomelanocortin (POMC), neuropeptide Y (NPY), galanin, CRH, cholecystokinin, and more newly described peptides (e.g., orexin/hypocretin). Many of these neuropeptides have been found both in the brain and in the gut. Alterations in some of these neuropeptide systems have been described in patients with anorexia nervosa.88 Altered endogenous opioid activity may play a role in the disturbed feeding behavior seen in anorexia nervosa. Animal studies indicate that opioid agonists increase and opioid antagonists decrease food intake.89 Underweight anorectics are reported to have lower CSF b-endorphin levels than healthy volunteers, whereas long-term weight-restored patients have normal CSF b-endorphin levels. Therefore, reduced b-endorphin activity may play a role in the refusal of food observed in anorectics. NPY is one of the most potent stimulants of feeding behavior seen in the CNS. NPY, which is found in the hypothalamus, stimulates feeding behavior when injected intracerebroventricularly in animals. Studies of persons with anorexia nervosa indicate that underweight anorexic patients have significantly higher levels of CSF NPY than do controls and that this level normalizes after long-term weight restoration.88 Possibly, this elevation represents a homeostatic mechanism to stimulate feeding, which, however, is ineffective within this patient group because of down-regulation of NPY receptors. NPY may play a role in the amenorrhea seen in anorexic patients by reducing LH, as NPY is involved in regulating the release of LHRH into the hypophysial circulation90 (see Chap. 125). Although these substances are technically difficult to study directly in human subjects, further research on the roles of these and other substances in food-intake-regulation is likely to have important implications for the pathophysiology of eating disorders and may identify new targets for pharmacologic treatment of these disorders. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th ed. Washington: American Psychiatric Press, 1994. Gershon ES, Badner JA, Goldin LR, et al. Closing in on genes for manic depressive illness and schizophrenia. Neuropsychopharmacology 1998; 18(4):233. DePaulo JJ, McMahon FJ. Recent developments in the genetics of bipolar disorder. Cold Spring Harb Symp Quant Biol 1996; 61(783):783. Berrettini W. Progress and pitfalls: bipolar molecular linkage studies. J Affect Disord 1998; 50(2–3):287. Baron M. Genetic linkage and bipolar affective disorder: progress and pitfalls. Mol Psychiatry 1997; 2(3):200. Stokes PE, Sikes CR. The hypothalamic-pituitary-adrenocortical axis in major depression. Neurol Clin 1988; 6(1):1. Joffe RT, Levitt AJ. The thyroid and depression. In: Joffe RT, Levitt AJ, eds. The thyroid axis and psychiatric illness. Washington: American Psychiatric Press, 1993:195. Gold PW, Loriaux DL, Roy A, et al. Responses to corticotropin-releasing hormone in the hypercortisolism of depression and Cushing's disease: pathophysiologic and diagnostic implications. N Engl J Med 1986; 314:1329. Amsterdam JD, Winokur A, Abelman E, et al. Cosyntropin (ACTH 1–24) stimulation test in depressed patients and healthy subjects. Am J Psychiatry 1983; 140(907):907. Orth DN. Corticotropin-releasing hormone in humans. Endocr Rev 1992; 13:164. Mitchell AJ. The role of corticotropin releasing factor in depressive illness: a critical review. Neurosci Biobehav Rev 1998; 22(5):635. Nemeroff CB, Bissette G, Akil H, Fink M. Neuropeptide concentrations in the cerebrospinal fluid of depressed patients treated with electroconvulsive therapy. Corticotrophin-releasing factor, beta-endorphin and somatostatin. Br J Psychiatry 1991; 158(59):59.

13. De Bellis MD, Gold PW, Geracioti TD Jr, et al. Association of fluoxetine treatment with reductions in CSF concentrations of corticotropin-releasing hormone and arginine vasopressin in patients with major depression. Am J Psychiatry 1993; 150(4):656. 14. Heinrichs SC, Menzaghi F, Pich EM, et al. The role of CRF in behavioral aspects of stress. Ann N Y Acad Sci 1995; 771:92. 15. Deak T, Nguyen KT, Ehrlich AL, et al. The impact of the nonpeptide corticotropin-releasing hormone antagonist antalarmin on behavioral and endocrine responses to stress. Endocrinology 1999; 140(1):79. 16. Geracioti TD, Orth DN, Ekhator NN, et al. Serial cerebrospinal fluid corticotropin-releasing hormone concentrations in healthy and depressed humans. J Clin Endocrinol Metab 1992; 74:1325. 17. Gold PW. The endocrinology of melancholic and atypical depression: relation to neurocircuitry and somatic consequences. Proc Assoc Am Physicians 1999; 111(1):22. 18. Kling MA, Roy A, Doran AR, et al. Cerebrospinal fluid immunoreactive corticotropin-releasing hormone and adrenocorticotropin secretion in Cushing's disease and major depression: potential clinical implications. J Clin Endocrinol Metab 1991; 72(2):260. 19. Dorn LD, Burgess ES, Dubbert B, et al. Psychopathology in patients with endogenous Cushing's syndrome: “atypical” or melancholic features. Clin Endocrinol (Oxf) 1995; 43(4):433. 20. Weissman MM, Klerman GL. Sex differences and the epidemiology of depression. Arch Gen Psychiatry 1977; 3:98. 21. Altman N, Sachar EJ, Gruen PH, et al. Reduced plasma LH concentration in postmenopausal depressed women. Psychosom Med 1975; 37:274. 22. Rivest S, Rivier C. The role of corticotropin-releasing factor and interleukin-1 in the regulation of neurons controlling reproductive function. Endocr Rev 1995; 16:177. 23. Sirinathsinghji DJS, Rees LH, Rivier J, Vale W. Corticotropin-releasing factor is a potent inhibitor of sexual receptivity in the female rat. Nature 1983; 305:232. 24. Chrousos GP, Torpy DJ, Gold PW. Interactions between the hypothalamic-pituitary-adrenal axis and the female reproductive system: clinical implications (NIH Conference). Ann Intern Med 1998; 129:229. 25. Bebbington PE, Dunn G, Jenkins R, et al. The influence of age and sex on the prevalence of depressive conditions: report from the National Survey of Psychiatric Morbidity. Psychol Med 1998; 28:9. 26. McEwen BS. Clinical Review 108: the molecular and neuroanatomical basis for estrogen effects on the central nervous system. J Clin Endocrinol Metab 1999; 84(6):1790. 27. Arpels JC. The female brain hypoestrogenic continuum from the premenstrual syndrome to menopause. J Reprod Med 1996; 41:633. 28. Fink G, Sumner BE, Rosie R, et al. Estrogen control of central neurotransmission: effect on mood, mental state, and memory. Cell Mol Neurobiol 1996; 16:325. 29. Seeman MV. Psychopathology in women and men: focus on female hormones. Am J Psychiatry 1997; 154:1641. 30. Archer JS. Relationship between estrogen, serotonin, and depression. Menopause 1999; 6:71. 31. Gregoire AJP, Kumar R, Everitt B, et al. Transdermal oestrogen for treatment of severe post-natal depression. Lancet 1996; 347:930. 32. Rubinow DR, Schmidt PJ, Roca CA. Estrogen-serotonin interactions: implications for affective regulation. Biol Psychiatry 1998; 44:839. 33. Joffe II, Cohen LS. Estrogen, serotonin, and mood disturbance: where is the therapeutic bridge? Biol Psychiatry 1998; 44:798. 34. Esposito S, Prange AJ, Golden RN. The thyroid axis and mood disorders: overview and future prospects. Psychopharmacol Bull 1997; 33(2):205. 35. Joffe RT. The use of thyroid supplements to augment antidepressant medication. J Clin Psychiatry 1998; 5(26):26. 36. Whybrow PC, Prange AJ. A hypothesis of thyroid-catecholamine receptor interaction: its relevance to affective illness. Arch Gen Psychiatry 1981; 38:106. 37. Haggerty JJ, Silva SG, Marquardt M, et al. Prevalence of antithyroid antibodies in mood disorders. Depress Anxiety 1997; 5(2):91. 38. Woeber KA. Subclinical thyroid dysfunction. Arch Intern Med 1997; 157(10):1065. 39. Mendlewicz J, Linkowski P, Kerkhofs M, et al. Diurnal hypersecretion of growth hormone in depression. J Clin Endocrinol Metab 1985; 60:505. 40. Michelson D, Stratakis C, Hill L, et al. Bone mineral density in women with depression. N Engl J Med 1996; 335(16):1176. 40a. Reginster JY, Deroisy R, Paul I, et al. Depressive vulnerability is not an independent risk factor for osteoporosis in postmenopausal women. Maturitas 1999; 33:133. 41. Kling MA, Rubinow DR, Doran AR, et al. Cerebrospinal fluid immunoreactive somatostatin concentrations in patients with Cushing's disease and major depression: relationship to indices of corticotropin-releasing hormone and cortisol secretion. Neuroendocrinology 1993; 57(1):79. 42. Kramer MS, Cutler N, Feighner J, et al. Distinct mechanism for antidepressant activity by blockade of central substance P receptors. Science 1998; 281(5383):1640. 42a. Cotrufo P, Monteleone P, d'Istria M, et al. Aggressive behavioral characteristics and endogenous hormones in women with bulimia nervosa. Neuropsychobiology 2000; 42:58. 43. 44. 45. 46. 47. 48. 49.

Yanovski SZ. Biological correlates of binge eating. Addict Behav 1995; 20(6):705. Birketvedt GS, Florholmen J, Sundsfjord J, et al. Behavioral and neuroendocrine characteristics of the night-eating syndrome. JAMA 1999; 282:657. Neumarker KJ. Mortality and sudden death in anorexia nervosa. Int J Eat Disord 1997; 21(3):205. Licinio J, Wong ML, Gold PW. The hypothalamic-pituitary-adrenal axis in anorexia nervosa. Psychiatry Res 1996; 62(1):75. Gold PW, Kaye W, Robertson GL, Ebert M. Abnormalities in plasma and cerebrospinal fluid arginine vasopressin in patients with anorexia nervosa. N Engl J Med 1983; 308:1117. Eckert ED, Pomeroy C, Raymond N, et al. Leptin in anorexia nervosa. J Clin Endocrinol Metab 1998; 83(3):791. Grinspoon S, Baum H, Lee K, et al. Effects of short term rhIGF-I administration on bone turnover in osteopenic women with anorexia nervosa. J Clin Endocrinol Metab 1996; 81:3864.

49a. Seed JA, Dixon RA, McCluskey SE, Yang AH. Basal activity of the hypothalamic–pituitary–adrenal axis and cognitive function in anorexia nervosa. Eur Arch Psychiatry Clin Neurosci 2000; 250:11. 50. Doerr P, Fichter M, Pirke KM, Lund R. Relationship between weight gain and hypothalamic pituitary adrenal function in patients with anorexia nervosa. J Steroid Biochem 1980; 13:529. 51. Gold PW, Gwirtsman H, Avgerinos PC, et al. Abnormal hypothalamic-pituitary-adrenal function in anorexia nervosa. Pathophysiologic mechanisms in underweight and weight-corrected patients. N Engl J Med 1986; 314(21):1335. 52. Kling MA, Demitrack MA, Whitfield HJJ, et al. Effects of the glucocorticoid antagonist RU 486 on pituitary-adrenal function in patients with anorexia nervosa and healthy volunteers: enhancement of plasma ACTH and cortisol secretion in underweight patients. Neuroendocrinology 1993; 57(6):1082. 53. Kaye WH, Gwirtsman H, George DT, et al. Elevated cerebrospinal fluid levels of immunoreactive corticotropin releasing hormone in anorexia nervosa: relationship to state of nutrition, adrenal function, and intensity of depression. J Clin Endocrinol Metab 1987; 64:203. 54. Mortola JF, Rasmussen DD, Yen SS. Alterations of the adrenocorticotropin-cortisol axis in normal weight bulimic women: evidence for a central mechanism. J Clin Endocrinol Metab 1989; 68(3):517. 55. Demitrack MA, Kalogeras KT, Altemus M, et al. Plasma and cerebrospinal fluid measures of arginine vasopressin secretion in patients with bulimia nervosa and in healthy subjects. J Clin Endocrinol Metab 1992; 74:1277. 56. Nishita JK, Ellinwood EHJ, Rockwell WJK, et al. Abnormalities in the response of plasma arginine vasopressin during hypertonic saline infusion in patients with eating disorders. Biol Psychiatry 1989; 26:73. 57. Chiodera P, Volpi R, Marchesi C, et al. Reduction in the arginine vasopressin responses to metoclopramide and insulin-induced hypoglycemia in normal weight bulimic women. Neuroendocrinology 1993; 57:907. 58. Morimoto Y, Oishi T, Hanasaki N, et al. Interactions among amenorrhea, serum gonadotropins and body weight in anorexia nervosa. Endocrinol Jpn 1980; 27:191. 59. Warren MP, Vande Wiele RL. Clinical and metabolic features of anorexia nervosa. Am J Obstet Gynecol 1973; 117:435. 60. Garfinkel RE, Brown GM, Stancer HC, Moldofsky H. Hypothalamic-pituitary function in anorexia nervosa. Arch Gen Psychiatry 1975; 32:739. 61. Vigersky RA, Loriaux DL, Anderson AE, Lipsett MB. Anorexia nervosa: behavioral and hypothalamic aspects. Clin Endocrinol Metab 1976; 5:517. 62. Luger A, Deuster PA, Kyle SB, et al. Acute hypothalamic-pituitary-adrenal responses to the stress of treadmill exercise: physiologic adaptations to physical training. N Engl J Med 1987; 316(21):1309. 63. Kaye WH, Jimerson DC, Lake CR, Ebert MH. Altered norepinephrine metabolism following long-term weight recovery in patients with anorexia nervosa. Psychiatry Res 1985; 14:333. 64. Landsberg L, Young JB. Endocrine changes in anorexia nervosa; an interpretation based on the metabolic adaptation to caloric restriction. In: Brown GM, Relchlin KS, eds. Neuroendocrinology and psychiatric disorder. New York: Raven Press, 1984:349. 65. Obarzanek E, Lesem M, Jimerson D. Resting metabolic rate of anorexia nervosa patients during weight gain. Am J Clin Nutr 1994; 60:666. 66. Altemus M, Hetherington MM, Flood M, et al. Decrease in resting metabolic rate during abstinence from bulimic behavior. Am J Psychiatry 1991; 148:1071. 67. Obarzanek E, Lesem MD, Goldstein DS, Jimerson DC. Reduced resting metabolic rate in patients with bulimia nervosa. Arch Gen Psychiatry 1991; 48:456. 68. Gianotti L, Broglio F, Aimaretti G, et al. Low IGF-I levels are often uncoupled with elevated GH levels in catabolic conditions. J Endocrinol 1998; 21:115. 69. Gianotti L, Rolla M, Arvat E, et al. Effect of somatostatin infusion on the somatotrope responsiveness to growth hormone-releasing hormone in patients with anorexia nervosa. Biol Psychiatry 1999; 45:334. 70. Thissen J, Keteslegers JM, Underwood LE. Nutritional regulation of the insulin like growth factors. Endocr Rev 1994; 15:80. 71. Klibanski A, Biller BM, Schoenfeld DA, et al. The effects of estrogen administration on trabecular bone loss in young women with anorexia nervosa. J Clin Endocrinol Metab 1995; 80(3):898. 72. Carmichael KA, Carmichael DH. Bone metabolism and osteopenia in eating disorders. Medicine 1995; 74:254. 73. Counts DR, Gwirtsman H, Carlsson LMS, et al. The effects of anorexia nervosa and refeeding on growth-hormone-binding protein, the IGFs and the IGF binding proteins. J Clin Endocrinol Metab 1992; 75:762. 74. Gordon CM, Grace E, Jean ES, et al. Changes in bone turnover markers and menstrual function after short-term oral DHEA in young women with anorexia nervosa. J Bone Miner Res 1999; 14(1):136. 75. Devesa J, Perez-Fernandez R, Bokser L, et al. Adrenal androgen secretion and dopaminergic activity in anorexia nervosa. Horm Metab Res 1987; 20:57. 76. Zumoff B, Walsh BT, Katz JL, et al. Subnormal plasma DHEA to cortisol ratio in anorexia nervosa: a second hormonal parameter of ontogenic regression. J Clin Endocrinol Metab 1983; 56:668. 77. Raisz LG, Wiita B, Artis A, et al. Comparison of the effects of estrogen alone and estrogen plus androgen on biochemical markers of bone formation and resorption in postmenopausal women. J Clin Endocrinol Metab 1996; 81:37. 78. Balligand JL, Brichard SM, Brichard V, et al. Hypoleptinemia in patients with anorexia nervosa: loss of circadian rhythm and unresponsiveness to short-term refeeding. Eur J Endocrinol 1998; 138(4):415. 79. Carro E, Senaris R, Considine RV, et al. Regulation of in vivo growth hormone secretion by leptin. Endocrinology 1997; 138:2203. 80. Bereis K, Vosmeer SUK. Effects of glucocorticoids and of growth hormone on serum leptin. Eur J Endocrinol 1996; 135:663. 81. Sarraf P, Frederich RC, Turner EM, et al. Multiple cytokines and acute inflammation raise mouse leptin levels: potential role in inflammatory anorexia. J Exp Med 1997; 185:171. 82. Grunfeld C, Zhao C, Fuller J, et al. Endotoxin and cytokines induce expression of leptin, the ob gene product, in hamsters. A role for leptin in the anorexia of infection. J Clin Invest 1996; 97:2152. 83. Malmstrom R, Taskinen MR, Daronen SLY-JH. Insulin increases plasma leptin concentrations in normal subjects and patients with NIDDM. Diabetologia 1996; 39:993. 84. Boden G, Chen X, Mazzoli MIR. The effect of fasting on serum leptin in normal human subjects. J Clin Endocrinol Metab 1996; 81:3419. 84a. Monteleone PD; Lieto A, Tortorella A, et al. Circulating Peptin in patients with anorexia nervosa, bulimia nervosa or binge-eating disorder: relationship to body weight, eating patterns, psychopathology and endocrine changes. Psychiatry Res 2000; 94:121. 85. 86. 87. 88. 89. 90.

Ahima R, Pkabakaran D, Mantzoros C, et al. Role of leptin in the neuroendocrine response to fasting. Nature 1996; 382:250. Kopp W, Blum WF, Von PS, et al. Low leptin levels predict amenorrhea in underweight and eating disordered females. Mol Psychiatry 1997; 2(4):335. Audi L, Mantzoros CS, Vidal PA, et al. Leptin in relation to resumption of menses in women with anorexia nervosa. Mol Psychiatry 1998; 3(6):544. Kaye WH. Neuropeptide abnormalities in anorexia nervosa. Psychiatry Res 1996; 62(1):65. Morley JE, Levine AS, Yim GK, Lowry MT. Opioid modulation of appetite. Neurosci Biobehav Rev 1983; 7:281. Kalra SP, Allen LG, Clark JT, et al. Neuropeptide Y—an integrator of reproductive and appetitive functions. In: Moody TW, ed. Neural and endocrine peptides and behavior. New York: Plenum Press, 1986:353.

CHAPTER 202 RESPIRATION AND ENDOCRINOLOGY Principles and Practice of Endocrinology and Metabolism

CHAPTER 202 RESPIRATION AND ENDOCRINOLOGY PRASHANT K. ROHATGI AND KENNETH L. BECKER Influence of Hormones on Fetal Lung Maturation and the Surfactant System Glucocorticoids Thyroid Hormones Estrogen Prolactin Insulin Other Hormones Influence of Disorders of the Endocrine System on the Respiratory System Hypothalamus Pituitary Gland Thyroid Gland Disorders of Calcium and Bone Metabolism Adrenal Cortex Female Endocrine System Male Endocrine System Diabetes Mellitus Chapter References

Hormones play a vital role in normal fetal lung maturation and also influence postnatal pulmonary structure and function. Therefore, it is not surprising that endocrine disease can exert profound effects on the respiratory system. INFLUENCE OF HORMONES ON FETAL LUNG MATURATION AND THE SURFACTANT SYSTEM Pulmonary development is a continuous process that begins in the early embryo and continues for several years during postnatal life. A critical stage in prenatal lung development is the alveolar stage, beginning at 26 to 28 weeks, which is associated with the formation of rudimentary alveoli, the maturation of the surfactant system (type II pneumocytes and their product surfactant), and the preparation of the liquid-filled lungs for their eventual extrauterine function of gas exchange.1 A close relationship exists between fetal viability and the maturation of the surfactant system.2 Pulmonary surfactant (Fig. 202-1 and Fig. 202-2), which lines the alveolar spaces and small airways, is a complex mixture of lung proteins (10%) and lipids (90%), with dipalmitoylphosphatidylcholine (lecithin) being the major surface-active phospholipid component.

FIGURE 202-1. Diagram of the formation and secretion of pulmonary surfactant by the type II epithelial cells (granular pneumocytes). After the migration of the lamellar bodies to the apex of the cell, they are extruded by exocytosis into the alveolar subphase and expand into tubular myelin (TM) figures. At the air-liquid interface, the figures spread into a monolayer (M). (ER, endoplasmic reticulum; G, Golgi apparatus; LB, lamellar bodies in which surfactant is stored.) (From Goerke J. Lung surfactant. Biochim Biophys Acta 1974; 344:241.)

FIGURE 202-2. The concept of interfacial tension and the physiologic role of surfactant: A rubber balloon filled with water or air (left) tends to empty because of the retractile pressure that is generated by the elasticity of the rubber. Similarly, in the lungs (right), the alveoli also generate retractile pressure because of tissue elasticity; this pressure is roughly proportional to the lung volume and is maximal at total lung capacity (i.e., the lung volume at maximal inspiration). Another dominant component of the total retractile pressure in the alveolus is that generated by surface tension due to (or because of) the air-liquid interface. Surface tension is a physical phenomenon inherent at interfaces between two dissimilar phases (in alveoli, it is the air-liquid interface), which generates a surface force that reduces the surface area (and consequently the volume) of the air-liquid interface. The retractile pressure caused by surface tension can be estimated from the Laplace relation P = 2T/r (P is pressure, T is surface tension, and r is alveolar radius). It is obvious from the relationship of T and r that the retractile pressure from surface tension increases with the decrease in the size of the alveoli and would be maximal toward the end of expiration. Thus, the transpulmonary pressure required to counterbalance the retractile force of tissue elasticity would be the least toward the end of expiration; however, the transpulmonary pressure required to counteract the surface tension and prevent the alveoli from total collapse during expiration would be highest during end-expiration. The pulmonary surfactant, which lines the alveoli and becomes more concentrated in the air-liquid interface during expiration, progressively reduces this interfacial tension during expiration and, thereby, decreases the levels of transpulmonary pressures required to prevent alveolar collapse. Furthermore, surfactant reduces the work of breathing and also prevents alveolar edema.

Apoproteins consist of four distinct lung-specific surfactant-associated proteins, SP-A, SP-B, SP-C, and SP-D. SP-A and SP-D are hydrophilic glycoproteins that contain a collagenous aminoterminal domain and a noncollagenous lectin-like carboxyterminal domain. SP-B and SP-C are hydrophobic proteins. The genes for these proteins are expressed only in type II alveolar epithelial cells of the lungs and perhaps in certain bronchiolar epithelial cells that share the same embryonic lineage. Not only are these genes located on different chromosomes, but they respond differently to hormonal influences, suggesting that they are independently regulated. These proteins mediate the transformation of lamellar bodies into tubular myelin, accelerate the rate of adsorption and spreading of the phospholipids into a monolayer, help in phospholipid recycling and metabolism by facilitating the uptake of phospholipids by type II pneumocytes, and enhance pulmonary defense by modulating phagocytosis, bacterial killing, and chemotaxis of alveolar macrophages.3,4 The major function of surfactant is to reduce the surface tension that otherwise would be generated by the presence of the fluid-gas interface within the alveoli and small airways, thereby protecting alveoli from collapsing, reducing the work of breathing, and preventing alveolar edema. Neonates who lack a mature, “complete” surfactant system, or who have a deficiency of a specific component, develop neonatal respiratory distress syndrome (NRDS). The normal maturation of type II

pneumocytes, and the synthesis and secretion of surfactant, is influenced by several hormones5 (Table 202-1).

TABLE 202-1. The Influence of Hormones on Fetal Lung Maturation and Surfactant Synthesis

GLUCOCORTICOIDS The important physiologic role that glucocorticoids play in the regulation of the fetal surfactant system is suggested by the following observations in humans: (a) a surge in cortisol levels in amniotic fluid before the maturation of the surfactant system as evidenced by a rise in the lecithin/sphingomyelin (L/S) ratio in the amniotic fluid, (b) an inverse relationship between cortisol levels in cord blood and NRDS, and (c) the demonstration that the antepartum administration of glucocorticoids accelerates maturation of the surfactant system with a decrease in the incidence of NRDS in high-risk pregnancies.5a In animal studies, the antepartum administration of exogenous glucocorticoids is associated with binding of the hormone to its receptors in the cytosol and nucleus of type II pneumocytes, slowing of cytodifferentiation of type II pneumocytes into type I pneumocytes, ultrastructural evidence of maturation of type II pneumocytes with abundant lamellar inclusion bodies, and enhanced choline phosphate cytidyltransferase (CPCT) activity in the lung with an increase in choline incorporation into phosphatidylcholine. Moreover, there is a several-fold increase in surfactant phospholipid in bronchoalveolar lavage fluid and in the lungs, a significant increase in the L/S ratio in amniotic fluid, physiologic evidence of greater pulmonary distensibility and deflation stability, and an enhanced survival of the prematurely delivered fetus. In lung explant studies, glucocorticoids induce increases in SP-B and SP-C proteins and their messenger RNAs (mRNAs), whereas a variable response is observed in SP-A, depending on the dose and duration of glucocorticoid administration. In vivo studies have demonstrated an increase in SP-A protein and mRNA, an increase in SP-B mRNA, and a decrease in SP-C mRNA. THYROID HORMONES Thyroid hormones influence the maturation of the fetal lung and regulate surfactant synthesis. Low levels of thyroid hormones in cord blood have been found in neonates who develop NRDS. Interestingly, the intraamniotic injection of thyroxine to high-risk pregnant women whose infants were to be delivered prematurely induced changes in amniotic fluid typical for the mature fetal lung; perhaps more significantly, none of the premature babies developed NRDS. Type II pneumocytes have nuclear receptors for triiodothyronine. Hypothyroidism is associated with smaller type II pneumocytes, smaller and fewer lamellar bodies, and decreased synthesis of surfactant. After the administration of thyroid hormones, these morphologic changes reverse: There is increased choline incorporation into phosphatidylcholine subsequent to increased CPCT activity, increased fatty acid synthesis because of increased activity of fatty acid synthetase and acetyl coenzyme A carboxylase, augmented protein synthesis, improved lung mechanics, and increased recovery of surfactant from lung. Thyroid hormones do not enhance any of the surfactant proteins. In fact, they decrease SP-A and SP-B mRNA. ESTROGEN In humans, a surge in blood estrogen levels precedes the increase in the L/S ratio in the amniotic fluid, and plasma estrogen levels are decreased in infants with NRDS. In addition, the administration of estrogen to pregnant women increases the L/S ratio in amniotic fluid. In animal studies, the administration of estrogen to mothers produces the following effects in the fetus: accelerated morphologic maturation of the lung, increased choline incorporation into phosphatidylcholine by stimulating the activity of CPCT, and an increased amount of surface-active phospholipids recovered by lung lavage. These effects are mediated by estrogen-binding macromolecules in the cytosol of lung cells and not by classic estrogen receptors, which are lacking in the lung. Estrogen increases SP-A and SP-B mRNA, and decreases SP-C mRNA. PROLACTIN The circumstantial evidence for a role of prolactin in the maturation of the surfactant system of humans includes the presence of receptors for prolactin in fetal lungs, the surge in prolactin levels in fetal blood that precedes and parallels the increase in the L/S ratio in amniotic fluid, and the presence of low cord blood prolactin levels in infants with NRDS. The role of this hormone in surfactant production also is supported by animal studies that have shown an increase in total phospholipids, phosphatidylcholine, and lecithin in lung tissue extracts of animals treated with prolactin, and a reduction in lung lavage phospholipids when animals were treated with bromocriptine, an inhibitor of prolactin secretion. INSULIN The incidence of NRDS is increased six-fold in infants of mothers with diabetes because of a delay in the maturation of the lung and its surfactant system. Insulin receptors are present in fetal lung. Glucose freely crosses the placenta from mother to fetus and induces the fetal pancreas to produce more insulin. There is conflicting evidence concerning whether hyperinsulinemia or hyperglycemia is responsible for these adverse effects on the lung. In animals, fetal lung maturation has been studied by inducing diabetes in pregnant mothers with alloxan or streptozocin. These studies revealed morphologic evidence of delayed maturation, such as poorly developed alveoli, less well-differentiated epithelium, and intracellular accumulation of lamellar bodies in type II pneumocytes. Moreover, there was less retention of air on deflation, suggesting a propensity for alveolar collapse; reduced surface activity in lung lavage in association with diminished recovery of lecithin; decreased activity of CPCT; and diminished synthesis of lecithin, phosphatidylcholine, and phosphatidyl glycerol. Some of these findings may be due to an antagonism by insulin of the normal effects of cortisol on the enzymes of surfactant synthesis. Insulin decreases SP-A and SP-B proteins and their mRNAs, and has no effect on SP-C protein or its mRNA. Adequate glucose control of maternal diabetes appears to reduce the risk of fetal pulmonary immaturity.6 OTHER HORMONES Thyrotropin and corticotropin accelerate maturation of the surfactant system, presumably by increasing fetal thyroid hormones and glucocorticoids.7 Interestingly, corticotropin also may act independently of its corticotropic effect. The administration of phospholipids extracted from the hypothalamus accelerates the maturation of the surfactant system, and this effect is not the result of increased availability of substrate for surfactant synthesis. Epidermal growth factor also may play a role in lung maturation.8 In addition, decreased concentrations of mammalian bombesin have been found in the lungs of infants with NRDS.

INFLUENCE OF DISORDERS OF THE ENDOCRINE SYSTEM ON THE RESPIRATORY SYSTEM HYPOTHALAMUS Hypothalamic disorders can influence the respiratory tract by influencing the autonomic nervous system, by affecting the respiratory drive, or by producing hypopituitarism. The hypothalamus influences the cyclic changes in nasal resistance to airflow by affecting the sympathetic tone that controls blood flow, thereby modifying the turgidity of the mucosa and the humidification of inspired air. Patients with Kallmann syndrome, who have hypothalamic atrophy, lack cyclic changes in nasal resistance.9 The hypothalamus also regulates the diurnal variation of bronchomotor tone by influencing the vagal tone; however, abnormalities of bronchomotor tone have not been evaluated in hypothalamic disorders. Intense hypothalamic stimulation following an acute intracranial pathology associated with increased intracranial pressure has been implicated in the pathogenesis of life-threatening neurogenic pulmonary edema.10 In humans, hypothalamic dysfunction, initially suspected because of hyperphagia in patients with hypopituitarism, has been associated with central hypoventilation, depressed ventilatory response to inspired carbon dioxide, and hypoventilation during exercise.11 Abnormal breathing patterns, including sleep apnea, have been described in Prader-Willi and Kleine-Levin syndromes associated with hypothalamic dysfunction.12 Interestingly, corticotropin-releasing hormone is a respiratory stimulant in humans.13

PITUITARY GLAND Of the hormones produced by the pituitary gland, only growth hormone is known to have significant effects on the respiratory system. The effects of diminished as well as excessive secretion of growth hormone on the respiratory system depend on the age at which the disorder begins. This hormone influences growth and development maximally during childhood, to a limited extent in adulthood, and minimally, if at all, during fetal development. GROWTH HORMONE DEFICIENCY Newborns with hypopituitarism are of normal length and weight and have normally developed lungs. Retardation in linear growth is apparent within the first few years of life and is associated with a proportionate decrease in the size of the thorax and lungs, but there is no clinically recognizable respiratory impairment or changes in arterial blood gases. Although unproven, it is believed that these lungs would demonstrate fewer and smaller alveoli, as well as smaller bronchi and bronchioles. In acquired causes of hypopituitarism, the growth is normal until somatotropic deficiency develops. A child who acquires growth hormone deficiency during adolescence will have poorly developed paranasal sinuses and smaller lungs.14 The lungs would be expected to show small alveoli that are normal in number, because alveolar multiplication in humans is complete by preadolescence, and any further increase in lung volumes results primarily from increased size of the alveoli. Patients with adult-onset hypopituitarism, who are being treated with replacement of thyroid, adrenal, and sex hormones, but not with growth hormone, have no clinically recognizable respiratory symptoms. However, pulmonary function studies demonstrate reduced static lung volumes, with disproportionate reductions in functional residual capacity and residual volume as well as increased lung elastic recoil, but normal respiratory muscle strength, flow rates, and gas exchange. Recurrent nasal polyps have been noted in Sheehan syndrome.15 Although pituitary adenylate cyclase–activating peptide 38 is a potent endogenously produced dilator of human airways, its pathophysiologic effects are unknown.15a GROWTH HORMONE EXCESS Persons with acromegaly die prematurely, and deaths attributed to respiratory disease are three times more frequent. This probably is related to adverse functional consequences of structural changes in the upper and lower respiratory tract, as described in Table 202-2.16 On physical examination, the overgrowth of the ribs and the enlarged vertebral bodies may produce striking thoracic deformity (Fig. 202-3).

TABLE 202-2. Effect of Some Endocrine and Metabolic Disorders on the Respiratory System

FIGURE 202-3. Kyphotic, deformed chest in a 65-year-old man with acromegaly. Note the increased anteroposterior diameter of the barrel-shaped thorax related to the transverse position of the thickened and lengthened ribs.

Pulmonary Function Changes. Persons with acromegaly have large lungs.17 There is a significant correlation between the duration of acromegaly and lung size as measured by total lung capacity. The increase in total lung capacity with normal diffusing capacity seen in humans with acromegaly has been attributed to increase in size and perhaps to increase in number of alveoli.18 Upper Airways Obstruction. Persons with acromegaly have intrathoracic and extrathoracic airways obstruction. They are especially predisposed to upper airway obstruction because of (a) encroachment of the lumen of the upper airways by generalized thickening of the mucosa, (b) macroglossia (Fig. 202-4), and (c) narrowing of the opening between the vocal cords.19 The last problem may result from hypertrophy of the vocal cords or from fixation of the vocal cords by any of the following mechanisms: enlargement of arytenoid cartilages with impaired mobility of cricoarytenoid joints; paralysis of the recurrent laryngeal nerve because of demyelination or stretching by overgrown laryngeal cartilaginous structures, or by an enlarged thyroid gland, which can occur in acromegaly; and myopathic changes in laryngeal muscles. The spectrum of obstruction ranges from physiologic evidence of upper airway obstruction without clinical manifestations, to chronic upper airway obstruction with hoarseness and dyspnea on exertion, to the development of acute upper airway obstruction with stridor during the induction of anesthesia, on extubation, or after acute viral upper respiratory tract infections (this may require emergency tracheotomy).

FIGURE 202-4. Macroglossia in a 70-year-old man with acromegaly. Note also the cranial nerve palsy, involving the extraocular muscles, that was due to the enlarging pituitary tumor. Other causes of macroglossia include hypothyroidism, amyloidosis, the Beckwith-Wiedemann syndrome, Down syndrome, Hurler syndrome (mucopolysaccharidosis I), hemangioma, lymphangioma, and plexiform neurofibromatosis.

Sleep Apnea. Sleep apnea, primarily the obstructive form, has been recognized in patients with acromegaly and may be responsible for daytime somnolence, noisy

snoring, and life-threatening arrhythmias during sleep.20,21,21a During these obstructive apneic episodes, there is no airflow at the entrance to the upper airways, in spite of persistent thoracic and abdominal respiratory movement, suggesting occlusion of the upper airways. The fact that these apneic episodes occur only during sleep indicates that the occlusion is not related solely to anatomic narrowing of the airways, but also to superimposed abnormalities in upper airway function, which cause the oropharyngeal airway to collapse and occlude the airway. The collapse of the oropharynx is related to the interaction of at least three physiologic factors, all of which are operative in acromegaly: (a) Narrowing of the oropharyngeal lumen occurs caused by hyperplasia of soft tissues; (b) changes in the oronasopharynx take place, producing an increase in upstream airflow resistance, thereby leading to the development of greater subatmospheric pressure in the airways during inspiration; and (c) most importantly, during sleep, the inadequate increase in the tone of the upper airway muscles during inspiration results in their inability to prevent collapse of the upper airways during inspiration. Nasopharyngeal endoscopy performed during episodes of obstructive apnea has demonstrated collapse of the posterior and lateral hypopharyngeal walls during inspiration, before any posterior movement of the tongue.22 The sleep apnea episodes may be responsive to treatment with bromocriptine and octreotide.23,24 Polysomnographic studies in patients with acromegaly have found central sleep apnea to occur in ~50% of cases. The mechanism for altered respiratory control during sleep is not clear, but may be related to interaction of depressed levels of somatostatin, which is known to be involved in the control of breathing, and elevated levels of insulin-like growth factor-I.25,26 In this regard, octreotide therapy has been reported to be useful.27 THYROID GLAND (SEE TABLE 202-2) HYPERTHYROIDISM Pulmonary Function Changes. Alterations in pulmonary function in persons with hyperthyroidism stem from the development of myopathy affecting respiratory muscles, including the diaphragm; increased central respiratory drive; and increased metabolism.28 Total lung capacity and functional residual capacity usually are normal. However, vital capacity and maximum breathing capacity are reduced, primarily because of diminished muscle strength. Lung compliance is slightly decreased. Airway resistance is unaffected. In persons with hyperthyroidism, basal oxygen consumption and the absolute oxygen consumption at each level of work are increased, although the increment in oxygen consumption per unit increase in workload is comparable to that in normal persons. This requires an augmentation of ventilation. Because of muscle weakness and decreased lung compliance, this augmentation is achieved by increasing the frequency of breathing, rather than the tidal volume, and leads to increased dead-space ventilation, a disproportionate increase in minute ventilation, and dyspnea. Central respiratory drive in response to hypercapnia and hypoxia is increased, and, again, this disproportionate increase in minute ventilation is achieved predominantly by augmenting the frequency of breathing.29,30 Bronchial Asthma. The increase in bronchial reactivity and the worsening of coexistent asthma with the onset of hyperthyroidism are well established, although the mechanism remains unclear.31 Hypotheses explaining the biochemical link between asthma and hyperthyroidism have centered around altered metabolism of catecholamines, corticosteroids, or arachidonic acid derivatives.32 Although plasma catecholamine levels are low in hyperthyroidism, this is an unlikely explanation because the intracellular levels of cyclic adenosine monophosphate are higher than normal, and the tissue responsiveness to catecholamines is increased; both of these should cause improvement of bronchial asthma. In asthmatic patients with hyperthyroidism, there is an increased requirement for corticosteroids. However, serum cortisol is normal in hyperthyroidism despite increased endogenous catabolism, because of increased production. A finding, of uncertain significance, is that there is a shift in the metabolism of hydrocortisone toward its inactive 11-ketone metabolites. Hyperthyroidism is associated with an increase in cellular membrane phospholipids, the major source for arachidonic acid, and with an increase in the phospholipase activity that triggers the release of arachidonic acid from membrane phospholipids. Once freed, the arachidonic acid is metabolized by one of the two membrane-bound enzymatic pathways (see Chap. 172). The lipoxygenase pathway leads to production of leukotrienes, the mediators of immunologically mediated asthma. The cyclooxygenase pathway leads to production of endoperoxides and prostaglandins that can produce bronchoconstriction. Furthermore, in hyperthyroidism, the activity of prostaglandin 15-hydroxydehydrogenase, which in activates prostaglandins, is reduced. It is postulated that with the onset of hyperthyroidism, the accumulation of leukotrienes and prostaglandins in the lung, along with enhanced sensitivity of persons with asthma to these products, leads to a worsening of the asthma. Pulmonary Hypertension. Primary pulmonary hypertension has been described with hyperthyroidism associated with Graves disease in neonates and adults, and multinodular toxic goiter.33,34 Reversal of pulmonary hypertension with establishment of the euthyroid state suggests that this is due to an increase in pulmonary vascular reactivity; however, an autoimmune pathogenetic mechanism also has been suggested. HYPOTHYROIDISM Pulmonary Function Changes. The primary effects of hypothyroidism on the respiratory system should be differentiated from secondary effects of obesity and cardiac dysfunction, which frequently accompany hypothyroidism. Hypothyroidism is associated with dysfunction of the diaphragm and other muscles of respiration.35 Hypothyroid patients without obesity or cardiac dysfunction have normal lung volumes, airway conductance, and arterial blood gas values. However, maximum breathing capacity often is reduced as a consequence of muscle weakness, and pulmonary diffusing capacity can be reduced as a result of anemia. In hypothyroid, obese patients, the total lung capacity, vital capacity, maximum breathing capacity, and peak expiratory flow rate are reduced because of the combination of the obesity and muscle weakness. In severe hypothyroidism, the reduction of diffusing capacity is more than just a result of anemia, because it is not corrected by blood transfusion, suggesting the presence of structural changes within the alveolar membrane. All the abnormalities of pulmonary function correct with thyroxine administration. Bronchial reactivity increases in persons with hypothyroidism who do not have asthma.36 Alveolar Hypoventilation. Hypoxic and hypercapnic ventilatory drives are depressed in hypothyroidism. In severe hypothyroidism, a combination of depressed ventilatory drive and respiratory muscle weakness may produce alveolar hypoventilation and CO2 narcosis, which may be lethal.37,38 It responds to thyroid-replacement therapy. Sleep Apnea Syndrome. The recognition that symptoms of excessive daytime somnolence, apathy, and lethargy are shared by patients with hypothyroidism and those with sleep apnea syndrome led to polysomnographic studies in patients with hypothyroidism.39,40 Disturbances of sleep architecture and the presence of sleep apnea, predominantly obstructive sleep apnea, were found to be frequent, even when the patients were not obese. The frequency of apnea decreases with thyroxine therapy, even if there is no change in body weight.41 Multiple factors contribute to the pathogenesis of obstructive sleep apnea. They include narrowing of the upper airway orifice caused by myxomatous tissue infiltration of the tongue and pharyngeal structures; hypertrophy and hypotonia of the genioglossus caused by abnormalities of intrinsic contractile properties of the muscle; a preferential decrease in phasic inspiratory neural output to the upper airway musculature, including the genioglossus, as compared with the intercostal and diaphragmatic muscles; and a decrease in inspiratory effort. These findings suggest that the genioglossus and other pharyngeal muscles are unable to oppose the tendency for pharyngeal closure during inspiration when the pressure in the oropharynx is subatmospheric. Pleural Effusion. Most pleural effusions result from heart disease, pericardial effusion, or ascites related to hypothyroidism. The mechanism responsible for these effusions is not well understood, but probably is related to alterations in pleural capillary permeability or extravasation of hygroscopic mucoproteins into the cavity. These effusions may be small or massive, unilateral or bilateral, and borderline between exudates and transudates, and they often exhibit little evidence of inflammation. Initially, neutrophils and, subsequently, lymphocytes are predominant in these effusions.42 Pulmonary Edema. The development of pulmonary edema secondary to upper airways obstruction related to deposition of myxomatous tissue in the larynx has been described in a patient with hypothyroidism.43 In this respect, it is known that thyroid hormone upregulates alveolar epithelial fluid clearance.43a DISORDERS OF CALCIUM AND BONE METABOLISM (SEE TABLE 202-2) HYPERPARATHYROIDISM AND OTHER HYPERCALCEMIC DISORDERS Hypercalcemia of diverse causes, when associated with renal insufficiency and consequent hyperphosphatemia, leads to the deposition of calcium salts within the alveolar septa. Acute deposition can lead to diffuse alveolar damage and adult respiratory distress syndrome,44 whereas gradual deposition causes interstitial thickening and fibrosis and a restrictive pattern on pulmonary function testing. On the chest radiograph, an interstitial process may be seen that is commonly misinterpreted as pulmonary edema or idiopathic pulmonary fibrosis because the microdeposits of calcium usually are not apparent on radiographs.45 Radionuclides used for bone imaging may localize in the lungs, and may suggest the diagnosis of metastatic pulmonary calcification before any radiographic changes can be recognized. Metastatic pulmonary calcification may revert with normalization of the hypercalcemia.46 Severe chest deformity may occur in primary hyperparathyroidism, particularly if azotemia supervenes (Fig. 202-5). An association between hypercalcemia and apnea in infants has been reported.47

FIGURE 202-5. Forty-year-old man with long-lasting primary hyperparathyroidism that culminated in azotemia and osteodystrophy. Note the closeness of the ribs to the pelvis, and the effect of marked resorptive deformation of the clavicles.

HYPOPARATHYROIDISM AND OTHER HYPOCALCEMIC DISORDERS Hypocalcemia can affect the respiratory system by inducing spontaneous electrical activity (a) in the motor and sensory nerves, which produces tetanic laryngeal spasm, in which the vocal cords are fixed in the midline and cause stridor, crowing respiration, and sometimes asphyxiation,48 and (b) in the autonomic ganglia, which may increase bronchomotor tone and lead to wheezing.49 HYPERPHOSPHATEMIA AND HYPOPHOSPHATEMIA Hyperphosphatemia has no respiratory consequences. Conversely, severe hypophosphatemia, with serum levels less than 1 mg/dL, produces generalized muscle weakness, including respiratory muscles, and can lead to acute respiratory failure.50 RICKETS, OSTEOMALACIA, AND OSTEOPOROSIS Rickets, osteomalacia, and osteoporosis are associated with defective mineralization of bones, which can affect the thorax. Their manifestations are listed in Table 202-2 (Fig. 202-6).

FIGURE 202-6. Osteoporotic 75-year-old woman with severe chest deformity and kyphosis.B, The chest radiograph demonstrates the markedly diminished intrathoracic volume. Note the inward curvature of the midportions of the lower ribs.

ADRENAL CORTEX HYPERCORTISOLISM Chronic excess secretion of cortisol, as in Cushing syndrome or from exogenous corticosteroid therapy, influences the respiratory system because of (a) a demineralization of bones with collapse of vertebral bodies; (b) a redistribution of subcutaneous fat causing truncal obesity, increased supraclavicular fat pads, and buffalo hump; (c) a modulation of allergic reactions with striking remissions of bronchial asthma and allergic rhinitis; and (d) an impairment of host defense ability to localize infections. As a result, patients with hypercortisolism may develop miliary tuberculosis, invasive pulmonary aspergillosis, pulmonary cryptococcosis, or Pneumocystis carinii pneumonia. Moreover, deposition of fat in the paracardiac areas and upper mediastinum in patients with hypercortisolism can be mistaken on a chest radiograph for lymphoma or a mediastinal disease process. Polysomnographic studies performed in patients with hypercortisolism, which is associated with the obesity and disordered sleeping pattern described earlier as a part of the depressive syndrome, also have demonstrated abnormal sleep architecture and obstructive sleep apnea.51 Respiratory depression resulting in chronic respiratory failure has been reported in a patient with severe metabolic alkalosis associated with Cushing syndrome,52 and exogenous corticosteroid therapy in patients with chronic obstructive pulmonary disease or asthma has been shown to contribute to respiratory muscle weakness.53 HYPOCORTISOLISM Increased production of reaginic antibody and increased sensitivity to bronchial anaphylactic reactions occur in adrenalectomized animals. In humans, atopic manifestations, such as bronchial asthma or allergic rhinitis, may be an uncommon presenting sign of Addison disease.54,55 ADRENAL MEDULLA PHEOCHROMOCYTOMA Functional Hyperventilation Syndrome. Catecholamine excess in pheochromocytoma can produce anxiety, tremulousness, changes in sleep patterns, paresthesias, and hyperventilation. These symptoms have led to a mistaken diagnosis of a functional hyperventilation syndrome.56 Pulmonary Edema. Epinephrine- and norepinephrine-induced pulmonary edema is a well-known phenomenon in experimental studies in animals. In humans, pulmonary edema has been reported with pheochromocytoma. This is thought to be due to left ventricular dysfunction, to postcapillary venoconstriction, or to the toxic effects of catecholamines on pulmonary capillary endothelial cells.57 The development of pulmonary edema that follows the administration of b-adrenergic blockers without the concomitant administration of an a-adrenergic blocker results from the combination of vasoconstriction, mediated by unopposed a-adrenergic stimulation, and myocardial depression caused by interruption of b-adrenergic drive.58 Amelioration of Bronchial Asthma. Disappearance of bronchial asthma with the development of noradrenaline-secreting pheochromocytoma and its recurrence after the removal of the tumor have been described.59 Excess circulating norepinephrine produces bronchodilation because airways have sparse, if any, a-adrenergic receptors, but have a profusion of b receptors; therefore, only the b-agonist effects of the hormone become evident clinically. In addition, norepinephrine inhibits cholinergic discharge from parasympathetic preganglionic fibers and, thereby, reduces cholinergic bronchoconstrictor tone. FEMALE ENDOCRINE SYSTEM HYPERGONADOTROPISM Hypergonadotropism, either caused by the exogenous administration of human menopausal gonadotropins for treating infertility or caused endogenously by a hydatidiform mole or chorioepithelioma, can lead to the formation of large theca lutein cysts and the ovarian hyperstimulation syndrome.60 It is postulated that a

metabolite elaborated from these overstimulated ovaries increases the permeability of the peritoneum, and possibly the pleura, and can lead to the development of massive ascites and pleural effusions with hypovolemia. HYPOGONADOTROPIC HYPOGONADISM The prepubertal development of hypogonadotropic hypogonadism is associated with eunuchoid skeletal features. This causes a smaller thorax in proportion to height; hence, all the predicted lung volumes are overestimated because they are based on height, as well as age and sex. Thus, restrictive lung disease may be diagnosed mistakenly in these patients. ESTROGEN EFFECTS No known respiratory abnormalities are associated with endogenous excess or deficiency of estrogen. However, the exogenous administration of estrogens in contraceptives is associated with an increased frequency of hay fever and a hypercoagulable state, manifested as pulmonary thromboembolism. PROGESTERONE Hyperventilation during pregnancy results from the mechanical effects of the gravid uterus on the position and configuration of the diaphragm, increased metabolic demands, and the stimulatory effects of progesterone. That progesterone has stimulatory effects on respiration is suggested by the hyperventilation observed during the luteal phase of the menstrual cycle and the lack of this cyclic hyperventilation in postmenopausal women, and also by the stimulation of respiration in men with the administration of progesterone. Progesterone enhances the respiratory response to hypoxia and hypercapnia.61 Thus, progesterone has been used therapeutically in treating the sleep apnea syndrome. When this drug is used in combination with estrogen, it reduces sleep-disordered breathing in postmenopausal women.62 Premenstrual exacerbation of asthma is believed to be caused by a cyclic decrease in progesterone and, perhaps, estrogen before menses.63 These exacerbations can be prevented by the administration of progesterone.64 The mechanism by which progesterone produces this effect is unknown. It may be related to a decrease in contractility of smooth muscle in the airways, regulation of microvascular leakage and edema in the bronchial mucosa, or the drug's immunosuppressive effects. MEIG SYNDROME AND PSEUDO-MEIG SYNDROME Meig syndrome is characterized by the presence of ascites and hydrothorax in association with fibroma of the ovary, and their spontaneous resolution with removal of the tumor.65 When observed with other ovarian neoplasms, the same phenomenon is termed pseudo-Meig syndrome. It is thought that fluid transudes from the ovarian lesion into the peritoneal cavity, from which it is transported to the thorax through transdiaphragmatic lymphatic channels. The fluid is a transudate, but it can be exudative, hemorrhagic, or rich in amylase if the tumor contains a high concentration of this enzyme and is undergoing hemorrhagic necrosis. OVARIAN DYSGENESIS SYNDROME AND ULLRICH-NOONAN SYNDROME See Table 202-3 for a summary of phenotypic features relating to the respiratory system in patients with X-chromosome abnormalities and those with Ullrich-Noonan syndrome.

TABLE 202-3. Phenotypic Features Pertaining to the Respiratory System in Sex Chromosomal and Related Disorders Associated with Hypogonadism

MALE ENDOCRINE SYSTEM DISORDERS OF SEMINIFEROUS TUBULES Testicular tubular disease has no effect on the respiratory tract, except when the same disease affects both of the organ systems, as in Kartagener syndrome.66 This condition results from lack of dynein arms in the outer microtubules of the cilia and flagella. This ultrastructural defect reduces the motility of (a) embryonal cells, predisposing to the development of situs inversus; (b) spermatozoa, causing infertility; and (c) mucociliary transport in the paranasal sinuses and airways, leading to sinusitis and bronchiectasis. TESTOSTERONE EFFECT A relationship between testosterone and respiratory disturbances during sleep is suggested by the higher frequency of obstructive sleep apnea in men than in premenopausal women, the higher frequency of disordered breathing during sleep in postmenopausal women compared with premenopausal women, the report of the development of central sleep apnea and blunted ventilatory response to CO2in a patient with hypogonadotropic hypogonadism after the administration of testosterone, and the report of the development of obstructive sleep apnea after therapy with testosterone in patients with a history of snoring.67,68 Development of obstructive sleep apnea was related to an increase in supraglottic airway resistance resulting from hormonally induced hypertrophy of the soft tissues of the oronasopharynx. Furthermore, men with the sleep apnea syndrome often are impotent and have low concentrations of free and total testosterone that may be related to decreased steroidogenic activity of luteinizing hormone.69 HYPOGONADAL DISORDERS The prepubertal development of hypogonadotropic hypogonadism leads to testosterone deficiency and the somatic features of eunuchoidism. Kallmann syndrome is a familial hypogonadotropic hypogonadism associated with features of eunuchoidism, anosmia, or hyposmia caused by agenesis of the olfactory bulbs and tracts, midline defects such as cleft lip and cleft palate, and hypoplasia of the first rib.70 Other hypogonadal disorders such as Klinefelter syndrome (Fig. 202-7)71,72 and the Ullrich-Noonan syndrome73 associated with thoracopulmonary abnormalities are detailed in Table 202-3. Although they do not have hypogonadism, eunuchoidappearing patients with the inborn disease of collagen metabolism, Marfan syndrome, commonly have chest deformities as well as cystic lung disease.

FIGURE 202-7. A chest deformity in a 25-year-old man with XXY Klinefelter disease. This patient also had a “straight-back syndrome”(loss of the normal mild thoracic kyphosis), which resulted in a reduced anteroposterior diameter of the chest and a markedly diminished total lung capacity. There was a loud systolic murmur heard best at the base of the heart.

DIABETES MELLITUS Diabetes mellitus is associated with widespread hormonal, metabolic, microvascular, and neuropathic abnormalities, leading to dysfunction of many organ systems, including the respiratory system.74 PULMONARY INFECTIONS Infection may be an immense problem in poorly controlled diabetes mellitus because of impaired chemotaxis, phagocytosis, and intracellular killing by granulocytes and mononuclear phagocytes. The diminished killing activity by alveolar macrophages is related to a depressed respiratory burst and a diminished superoxide anion production, probably because of decreased availability of NADPH (reduced form of nicotinamide adenine dinucleotide phosphate).75 Poorly controlled diabetes predisposes to mycobacterial and fungal pulmonary infections, which also tend to be rapidly progressive. The increased susceptibility of patients with diabetes to zygomycosis (mucormycosis) may be related to an impaired ability of the alveolar macrophages to inhibit the germination of spores of this fungus.76 ALTERATIONS IN PULMONARY MECHANICS Abnormal pulmonary function has been found in 60% of a cross section of patients with diabetes. In some of these patients, the abnormalities are related to the obesity, congestive heart failure, and muscle weakness that are common in diabetes; in other patients, abnormal pulmonary function is noted in the absence of such confounding variables. The most consistent abnormalities are reduced lung volumes in young patients with insulin-dependent diabetes, reduced pulmonary elastic recoil and dynamic compliance in both young and adult patients with diabetes, and impaired diffusion.77,78 Reduced lung volumes result from arthropathy affecting costovertebral joints,79 or from nonenzymatic, glycosylation-induced alterations and the accumulation of connective tissue, including collagen, in the lungs. Impaired pulmonary diffusion is related to pulmonary microangiopathy that is characterized by thickening of epithelial and capillary basement membranes in the alveolar walls.80 Impaired lung epithelial permeability has been documented using aerosol scintigraphy.81 SUDDEN DEATH Unexplained cardiorespiratory arrests in patients with diabetes and autonomic neuropathy may result from abnormal ventilatory responses to hypoxia, either because of reduced sensitivity of the central chemoreceptors or, more likely, because of neuropathy involving the vagus and glossopharyngeal nerves, which transmit afferent impulses from the carotid body and aortic arch chemoreceptors to the respiratory center. Studies evaluating the control of respiration in patients with diabetes have provided conflicting findings, with some investigators showing abnormal ventilatory responses to exercise, hypoxia, or hypercarbia, and others finding normal function.82 SLEEP APNEA SYNDROME Like patients with primary autonomic neuropathy (Shy-Drager syndrome), patients with diabetes who have autonomic neuropathy may have an increased prevalence of sleep-related breathing disorders that may be the basis for the unexplained cardiorespiratory deaths.83,84 Although earlier studies in elderly patients with diabetic autonomic neuropathy had suggested a higher prevalence of hypopnea and sleep apnea, subsequent studies in younger diabetic patients, with and without autonomic neuropathy, have failed to confirm these findings. ADULT RESPIRATORY DISTRESS SYNDROME Noncardiogenic pulmonary edema occurs during the treatment of diabetic ketoacidosis.85 It may result from thickening of the basal lamina of alveolar capillaries, rendering them more permeable, especially during acidosis. Other factors that are implicated are development of cerebral edema causing neurogenic pulmonary edema or a decrease in the pressure gradient between the serum colloid oncotic pressure and the pulmonary capillary wedge pressure, leading to accumulation of fluid in the lungs. ALVEOLAR HYPOVENTILATION Severe muscular weakness, leading to alveolar hypoventilation, has been reported when severe hypophosphatemia or hypokalemia develops during the treatment of diabetic ketoacidosis. This can be prevented and treated by repletion of phosphate or potassium.86 BRONCHIAL ASTHMA Asthma and diabetes rarely occur in the same patient; however, the development of hypoglycemia caused by either exogenous or endogenous insulin excess has been implicated in the exacerbation of bronchospasm in patients with asthma. The mechanism for this apparent exclusion between the two diseases is not understood but probably is related to genetic factors; to alterations in both insulin release and its hypoglycemic effect in atopic patients with asthma; to effects on the metabolism of cyclic nucleotides, which mediate smooth muscle contraction and relaxation in patients with diabetes; and to the effect of diabetic neuropathy on bronchomotor tone. Patients with diabetes who have autonomic neuropathy have reduced bronchial reactivity because of depression of cholinergic bronchomotor tone.87,88 MISCELLANEOUS Transient vocal cord paralysis,89 apparently due to neuropathy, and persistent transudative pleural effusions related to left ventricular dysfunction and possibly other unknown factors, have been reported in patients with long-standing diabetes mellitus.90 Massive pulmonary thromboembolism may develop secondary to the dehydration that occurs in diabetic ketoacidosis.91 CHAPTER REFERENCES 1. 2. 3. 4. 5.

Thurlbeck WM. Postnatal growth and development of the lung. Am Rev Respir Dis 1975; 111:803. Farrell PM, Avery ME. Hyaline membrane disease. Am Rev Respir Dis 1975; 111:657. Ballard PL. Hormonal regulation of pulmonary surfactant. Endocr Rev 1989; 10:165. Floros J, Kala P. Surfactant proteins: molecular genetics of neonatal pulmonary diseases. Ann Rev Physiol 1998; 60:365. Rooney SA. The surfactant system and lung phospholipid biochemistry. Am Rev Respir Dis 1985; 131:439.

5a. Spencer C, Heales K. Antenatal corticosteroids to prevent neonatal respiratory distress syndrome. BMJ 2000; 320:325. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Piper JM, Langer O. Does maternal diabetes delay pulmonary maturity? Am J Obstet Gynecol 1993; 168:783. Morales WJ, O'Brien WF, Angel JL, et al. Fetal lung maturation: the combined use of corticosteroids and thyrotropin-releasing hormone. Obstet Gynecol 1989; 73:111. Miettinen PJ, Warburton D, Bu D, et al. Impaired lung branching morphogenesis in the absence of functional EGF receptor. Dev Biol 1997; 186:224. Galioto G, Mevio E, Galioto P, et al. Modifications of the nasal cycle in patients with hypothalamic disorders. Kallmann's syndrome. Ann Otol Rhinol Laryngol 1991; 100:559. Brambrink AM, Dick WF. Neurogenic pulmonary edema. Pathogenesis, clinical picture and therapy. Anaesthesist 1997; 46:953. Moskowitz MA, Fisher JN, Simpser MD, Strieder DJ. Periodic apnea, exercise hypoventilation and hypothalamic dysfunction. Ann Intern Med 1976; 84:171. Sforza E, Krieger J, Geisert J, et al. Sleep and breathing abnormalities in a case of Prader-Willi syndrome. Acta Paediatr Scand 1991; 80:80. Nink M, Salomon E, Coutinho M, et al. Corticotropin-releasing hormone (CRH) is a respiratory stimulant in humans: a comparative study of human and ovine CRH. Life Sci 1994; 54:1793. Jain BP, Brody JS, Fisher AB. The small lung of hypopituitarism. Am Rev Respir Dis 1975; 108:49. Hampal S, Bingham B, Desai P, et al. Recurrent simple nasal polyps associated with Sheehan's syndrome. J Otolaryngol 1992; 21:368.

15a. Kinhult J, Andersson JA, Uddman R, et al. Pituitary adenylate cyclase-activating peptide 38, a potent endogenously produced dilator of human airways. Eur Respir J 2000; 15:243.

16. 17. 18. 19. 20. 21.

Harrison BDW, Millhouse KA, Harrington M, Nabarro JDN. Lung function in acromegaly. Q J Med 1978; 47:517. Brody JS, Fisher AB, Gocmen A, Dubois AB. Acromegalic pneumomegaly: lung growth in the adult. J Clin Invest 1970; 49:1051. Donnelly PM, Grunstein RR, Peat JK, et al. Large lungs and growth hormone: an increased alveolar number? Eur Respir J 1995; 8:938. Iandelli I, Gorini M, Duranti R, et al. Respiratory muscle function and control of breathing in patients with acromegaly. Eur Respir J 1997; 10:977. Murrant NJ, Gatland DJ. Respiratory problems in acromegaly. J Laryngol Otol 1990; 104:52. Mazon BJ, West P, Maclean JP, Kryger HG. Sleep apnea in acromegaly. Am J Med 1980; 69:615.

21a. Saeki N, Isono S, Nishino T, et al. Sleep disordered breathing in acromegalics—relation of hormonal levels and quantitative sleep study by means of bedside oximeter. Endocr J 1999; 46:585. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Cadieux RJ, Kales A, Senten RJ, et al. Endoscopic findings in sleep apnea associated with acromegaly. J Clin Endocrinol Metab 1982; 55:18. Zlemer DC, Dunlap DB. Case report: relief of sleep apnea in acromegaly by bromocriptine. Am J Med Sci 1988; 295:49. Buyse B, Michiels E, Bouillon R, et al. Relief of sleep apnoea after treatment of acromegaly: report of three cases and review of the literature. Eur Respir J 1997; 10:1401. Grunstein RR, Ho KY, Sullivan CE. Sleep apnea in acromegaly. Ann Intern Med 1991; 115:527. Grunstein RR, Ho KY, Berthon-Jones M, et al. Central sleep apnea is associated with increased ventilatory response to carbon dioxide and hypersecretion of growth hormone in patients with acromegaly. Am J Respir Crit Care Med 1994; 150:496. Leibowitz G, Shapiro MS, Salameh M, Glaser B. Improvement of sleep apnea due to acromegaly during short-term treatment with octreotide. J Intern Med 1994; 236:231. McElvaney GN, Wilcox PG, Fairbarn MS, et al. Respiratory muscle weakness and dyspnea in thyrotoxic patients. Am Rev Respir Dis 1990; 141:1221. Massey DG, Becklace MR, McKenzie JM, Bates DV. Circulatory and ventilatory response to exercise in thyrotoxicosis. N Engl J Med 1967; 276:1104. Kahaly G, Hellerman J, Mohr-Kahaly S, Treese N. Impaired cardiopulmonary exercise capacity in patients with hyperthyroidism. Chest 1996; 109:57. Cockroft DW, Silverberg JDH, Dosman JA. Decrease in non-specific bronchial reactivity in an asthmatic following treatment of hyperthyroidism.Ann Allergy 1978; 41:160. Hoult JRS, Moore P. Thyroid disease, asthma, and prostaglandin. Br Med J 1978; 1:366. Thurnheer R, Jenni R, Russi EW, et al. Hyperthyroidism and pulmonary hypertension. J Intern Med 1997; 242:185. O'Donovan D, McMahon C, Costigan C. Reversible pulmonary hypertension in neonatal Graves disease. Ir Med J 1997; 90:147. Martinez FJ, Gomez MB, Celli BR. Hypothyroidism: a reversible cause of diaphragmatic dysfunction. Chest 1989; 96:1059. Wieshammer S, Keck FS, Schauffelen AC, et al. Effects of hypothyroidism on bronchial reactivity in non-asthmatic subjects. Thorax 1990; 45:947. Ladenson PW, Goldenheim PD, Ridgway EC. Prediction and reversal of blunted ventilatory responsiveness in patients with hypothyroidism. Am J Med 1988; 84:877. Zwillich CW, Pierson DJ, Hofeldt FD, et al. Ventilatory control in myxedema and hypothyroidism. N Engl J Med 1975; 292:662. Skatrud J, Iber C, Ewart R, et al. Disordered breathing during sleep in hypothyroidism. Am Rev Respir Dis 1981; 124:325. Kapur VK, Koepsell TD, deMaine J, et al. Association of hypothyroidism and obstructive sleep apnea. Am J Respir Crit Care Med 1998; 158:1379. Rajgopal KR, Abbrecht PH, Derderian SS, et al. Obstructive sleep apnea in hypothyroidism. Ann Intern Med 1984; 101:491. Gottehrer A, Roa J, Stanford GG. Hypothyroidism and pleural effusions.Chest 1990; 98:1130. Lopez A, Lorente JA, Jerez V, et al. Sleep apnea, hypothyroidism and pulmonary edema. Chest 1990; 97:763.

43a. Folkesson HG, Marlin A, Wang Y, et al. Dexamethasone and thyroid hormone pretreatment upregulate alveolar epithelial fluid clearance in adult rate. J Appl Physiol 2000; 88:416. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91.

Khafif RA, Delima C, Silverberg A, et al. Acute hyperparathyroidism with systemic calcinosis. Arch Intern Med 1989; 149:681. Hartman TE, Muller NL, Primack SL, et al. Metastatic pulmonary calcification in patients with hypercalcemia: findings on chest radiographs and CT scans. AJR Am J Roentgenol 1994; 162:799. Weber CK, Friedrich JM, Merkle E, et al. Reversible metastatic pulmonary calcification in a patient with multiple myeloma. Ann Hematol 1996; 72:329. Kooh S, Binet A. Hypercalcemia in infants presenting with apnea. Can Med Assoc J 1990; 143:509. Williams GT, Brown M. Laryngospasm in hypoparathyroidism. J Laryngol Otol 1974; 88:369. Aberg H, Johansson H, Werner I. Hyperparathyroidism and asthma. Lancet 1972; 2:381. Newman JH, Neff TA, Ziporin P. Acute respiratory failure associated with hypophosphatemia. N Engl J Med 1977; 296:1101. Shipley JE, Schteingart DE, Tandon R, et al. Sleep architecture and sleep apnea in patients with Cushing's disease. Sleep 1992; 15:514. Tanaka M, Yano T, Ichikawa Y, et al. A case of Cushing's syndrome associated with chronic respiratory failure due to metabolic alkalosis. Intern Med 1992; 31:385. Decramer M, Laquet LM, Fagard R, Ragiers P. Corticosteroids contribute to muscle weakness in chronic airflow obstruction. Am J Respir Crit Care Med 1994; 150:11. Green M, Lim KH. Bronchial asthma with Addison's disease. Lancet 1971; 1:1159. Saraclar Y, Turktas I, Adalioglu G, Tuncer A. Bronchial asthma with Addison's disease. Respiration 1993; 60:241. Magarian GJ. Hyperventilation syndromes: infrequently recognized common expressions of anxiety and stress. Medicine (Baltimore) 1982; 61:219. Fahmy N, Assaad M, Bathija P, Whittier FC. Postoperative acute pulmonary edema: a rare presentation of pheochromocytoma. Clin Nephrol 1997; 48:122. Wark JD, Larkins RG. Pulmonary edema after propranolol therapy in two cases of phaeochromocytoma. Br Med J 1978; 1:1395. Harvey JN, Dean HG, Lee MR. Recurrence of asthma following removal of a noradrenaline-secreting phaeochromocytoma. Postgrad Med J 1984; 60:364. Shapiro AG, Thomas T, Epstein M. Management of hyperstimulation syndrome. Fertil Steril 1977; 28:238. Bayliss DA, Millhorn DE. Central neural mechanisms of progesterone action. Application to the respiratory system. J Appl Physiol 1992; 73:393. Pickett CK, Regensteiner JG, Woodard WD, et al. Progestin and estrogen reduce sleep-disordered breathing in post menopausal women. J Appl Physiol 1989; 66:1656. Chandler MH, Schuldheisz S, Phillips BA, Muse KN. Premenstrual asthma: effect of estrogen on symptoms, pulmonary function, and beta 2 receptors.Pharmacotherapy 1997; 17: 224. Beynon HLC, Garbett ND, Barnes PJ. Severe premenstrual exacerbations of asthma. Effect of intramuscular progesterone. Lancet 1988; 1:370. Lemming R. Meigs syndrome and pathogenesis of pleurisy and polyserositis. Acta Med Scand 1960; 168:197. Afzelius BA. A human syndrome caused by immotile cilia. Science 1976; 193:317. Sandblom RE, Matsumoto AM, Schoene RB, et al. Obstructive sleep apnea syndrome induced by testosterone administration. N Engl J Med 1983; 308:508. Johnson MW, Anch AM, Remmers JE. Induction of the obstructive sleep apnea syndrome in a woman by exogenous androgen administration. Am Rev Respir Dis 1984; 129:1023. Grunstein RR. Metabolic aspects of sleep apnea. Sleep 1996; 19:S218. Santen RJ, Paulsen A. Hypogonadotropic eunuchoidism: clinical study of the mode of inheritance. J Clin Endocrinol 1973; 36:47. Gluck MC, Becker KL, Katz S. Pulmonary function of hypogonadal men before and after testosterone. Am Rev Respir Dis 1966; 94:676. Morales P, Furest I, Marco V, et al. Pathogenesis of the lung in restrictive defects of Klinefelters syndrome. Chest 1992; 102:1550. Baltaxe HA, Lee JG, Ehlers KH, Engle MA. Pulmonary lymphangiectasia demonstrated by lymphangiography in two patients with Noonan's syndrome. Radiology 1975; 115:149. Strojek K, Ziora D, Sroczynski JW, et al. Pulmonary complications of type I(insulin dependent) diabetic patients. Diabetologia 1992; 35:1173. Mohsenin V, Latifpour J. Respiratory burst in alveolar macrophages of diabetic rats. J Appl Physiol 1990; 68:2384. Waldorf AR, Ruderman N, Diamond RD. Specific susceptibility to mucormycosis in murine diabetes and bronchoalveolar macrophage defense against Rhizopus. J Clin Invest 1984; 74:150. Sandler M. Is the lung a “target organ” in diabetes mellitus. Arch Intern Med 1990; 150:1385. Ljubic S, Metelko Z, Car N, et al. Reduction of diffusion capacity for carbon monoxide in diabetic patients. Chest 1998; 114:1033. Schnapf BM, Banks RA, Silverstein JH, et al. Pulmonary function in insulin-dependent diabetes mellitus with limited joint mobility. Am Rev Respir Dis 1984; 130:930. Vracko RD, Thorning D, Huang TW. Basal lamina of alveolar epithelium and capillaries. Quantitative changes with aging and in diabetes mellitus.Am Rev Respir Dis 1979; 120:973. Caner B, Ugur O, Byraktar M, et al. Impaired lung epithelial permeability in diabetics detected by technetium-99m-DTPA aerosol scintigraphy. J Nucl Med 1994; 35:204. Tantucci C, Scionti L, Bottini P, et al. Influence of autonomic neuropathy of different severities on the hypercapnic drive to breathing in diabetic patients. Chest 1997; 112:145. Catterall JR, Calverley PMA, Ewing DJ, et al. Breathing, sleep, and diabetic autonomic neuropathy. Diabetes 1984; 33:1025. Strohl KP. Diabetes and sleep apnea. Sleep 1996; 19:S225. Hillerdal G, Wibell L. Adult respiratory distress syndrome and diabetes.Acta Med Scand 1982; 211:221. Tillman CR. Hypokalemic hypoventilation complicating severe diabetic ketoacidosis. South Med J 1980; 73:231. Villa MP, Cacciari E, Bernardi F, et al. Bronchial reactivity in diabetic patients. Relationship of duration of diabetes and degree of glycemic control. Am J Dis Child 1988; 142:726. Santos e Fonesca CM, Manco JC, Gallo J, et al. Cholinergic bronchomotor tone and airway caliber in insulin-dependent diabetes mellitus. Chest 1992; 101:1038. Kabadi UM. Unilateral vocal cord palsy in diabetic patient. Postgrad Med 1988; 84:53. Chertow BS, Kadzielowa R, Burger AJ. Benign pleural effusions in longstanding diabetes mellitus. Chest 1991; 99:1108. Quigley RL, Curran RD, Stagl RD, Alexander JC Jr. Management of massive pulmonary thromboembolism complicating diabetic ketoacidosis. Ann Thorac Surg 1994; 57:1322.

CHAPTER 203 THE CARDIOVASCULAR SYSTEM AND ENDOCRINE DISEASE Principles and Practice of Endocrinology and Metabolism

CHAPTER 203 THE CARDIOVASCULAR SYSTEM AND ENDOCRINE DISEASE ELLEN W. SEELY AND GORDON H. WILLIAMS Acromegaly Cardiomegaly Hypertension Atherosclerosis Acromegalic Cardiomyopathy Treatment of Cardiovascular Diseases in Acromegaly Growth Hormone's Effect on the Heart Diseases of the Thyroid The Sympathetic Nervous System and Thyroid Dysfunction Hyperthyroidism Hypothyroidism Parathyroid Diseases Hyperparathyroidism Hypoparathyroidism Diseases of the Adrenal Gland Cushing Syndrome Hyperaldosteronism Adrenal Insufficiency Pheochromocytoma Gonadal Hormones and the Heart Postmenopausal Hormone Replacement Oral Contraceptives Chapter References

Since the reports of Robert Graves and Thomas Addison in the mid-nineteenth century, deranged hormonal secretions have been known to alter substantially the function of the cardiovascular system. However, the magnitude of the interaction has begun to be appreciated only since the advent of precise measurement techniques for circulating hormone concentrations.1 In fact, the heart has been added to the list of endocrine organs, with the evidence that it produces a circulating hormone (atrial natriuretic hormone; see Chap. 178).

ACROMEGALY The cardiac manifestations of acromegaly include enlargement of the heart, hypertension, premature coronary artery disease, cardiac arrhythmias, and congestive heart failure.2 Indeed, a specific acromegalic cardiomyopathy has been suggested to account for patients with congestive heart failure3 and cardiac arrhythmia in whom predisposing factors cannot be demonstrated. CARDIOMEGALY After the fifth decade, nearly all patients with active acromegaly have cardiomegaly. In some, the dysfunction is so great that the ejection fraction is reduced. Because the enlargement of the heart is greater than the generalized organomegaly that is usually observed with this disease, the cause is probably more than the generalized effect of growth hormone on protein synthesis. Other contributing factors may include hypertension and atherosclerosis, both of which occur with increased frequency in acromegaly, and a cardiomyopathy. Focal cardiac interstitial fibrosis and a myocarditis also have occurred in most cases.2 Because these patients may also have small vessel disease of the myocardium, diabetes, and hypertension, it is probable that a combination of factors contributes to the cardiac hypertrophy observed in acromegalics.3a Importantly, even short-term acromegaly affects the heart.3b HYPERTENSION Hypertension is probably the most common cardiovascular manifestation, occurring in 15% to 50% of acromegalic patients.4 The frequency varies according to whether office or ambulatory blood pressures are assessed. Hypertensive acromegalics tend to be older and to have had their acromegaly longer than non-hypertensive patients. The underlying functional abnormality remains uncertain, but growth hormone, abnormal aldosterone secretion, and a pressor substance in the urine of acromegalics have been considered. Several studies have suggested that the elevated growth hormone may be responsible for the hypertension: Pituitary irradiation and hypophysectomy reduce arterial pressure in hypertensive acromegalics; also, the administration of growth hormone can produce sodium retention and extracellular fluid volume expansion in normal persons. An abnormality in the secretion of aldosterone may produce volume expansion and, secondarily, hypertension. Early studies suggested that aldosterone secretion is increased in many patients with acromegaly. However, more recent studies suggest that this is uncommon.5 What frequently does occur is a change in the responsiveness of the adrenal gland and the peripheral vasculature to angiotensin II (AII). Thus, with sodium restriction, the aldosterone response to AII is decreased, whereas the vascular response is increased, compared with that of normal persons (see Chap. 79). These abnormalities are present in both hypertensive and normotensive acromegalics, although, perhaps, at a greater frequency in the hypertensives.6 Whether this is related to the pathogenesis of the elevated arterial pressure or is a reflection of the expanded extracellular fluid volume is unclear. ATHEROSCLEROSIS Because of growth hormone's major effect on carbohydrate and lipid metabolism, it is not surprising that premature atherosclerosis occurs in patients with acromegaly. It is uncertain, however, how frequently this tendency toward atherosclerosis occurs. One report2 suggests that only 10% of acromegalics have major coronary artery disease. ACROMEGALIC CARDIOMYOPATHY Several lines of evidence suggest that not all heart disease in acromegalics can be attributed to the increased prevalence of atherosclerosis and hypertension, leading some investigators to propose a specific acromegalic cardiomyopathy.7,8,9 and 10 The evidence includes the following: (a) Ten percent to 20% of acromegalics have overt congestive heart failure and, in perhaps 25% of these, there is no known predisposing factor. (b) Approximately half of all patients with acromegaly, including patients without hypertension, have echocardiographic evidence of left ventricular hypertrophy (LVH),7,8,10 the degree of which correlates reasonably well with the level of growth hormone. (c) Most patients with acromegaly who do not have hypertension or atherosclerosis have some clinical evidence of cardiac dysfunction, manifested by a shortening of the left ventricular (LV) ejection time and prolongation of the preejection period. (d) Nearly 50% of acromegalics have electrocardiographic abnormalities,2 of which hypertension or atherosclerosis may account for 10% to 20%, but the remaining 30% remain unexplained. (e) Histologic studies of the heart show cellular hypertrophy, patchy fibrosis, and myofibrillar degeneration.2 (f) Sudden death has been associated with inflammatory and degenerative changes in the sinoatrial and perinodal nerve plexus, as well as degeneration of the atrioventricular node.5 (g) Finally, the acromegalic heart has a higher collagen content per gram of tissue than the normal myocardium. Thus, the evidence generally supports the hypothesis that there is a specific acromegalic cardiomyopathy. The presence of this cardiomyopathy, in turn, provides an explanation for the increased frequency of cardiac dysrhythmia and the difficulty in treating congestive heart failure with conventional therapy. TREATMENT OF CARDIOVASCULAR DISEASES IN ACROMEGALY Acromegalics with cardiovascular abnormalities usually respond to conventional therapy for hypertension, heart failure, or arrhythmias. Those with hypertension, however, are more responsive to volume-depleting maneuvers (i.e., diuretics and sodium restriction) than the average person with essential hypertension. Also, acromegalics with congestive heart failure, without evidence of underlying hypertensive heart disease, often appear particularly resistant to conventional therapy.

Absence of a clinical response in such patients suggests that vigorous efforts are needed to lower growth hormone levels (see Chap. 12). The use of somatostatin analogs is effective in reversing cardiac abnormalities in parallel with the fall in growth hormone.10,11 GROWTH HORMONE'S EFFECT ON THE HEART Increasing evidence suggests that while growth hormone excess (acromegaly) has profound adverse effects on the heart, growth hormone deficiency also is deleterious. Growth hormone replacement in deficient individuals improves cardiac performance. Furthermore, since growth hormone increases cardiac muscle mass, it may be useful in treating individuals with idiopathic dilated cardiomyopathy by reducing the size of the LV chamber size.12

DISEASES OF THE THYROID The relationship between thyroid and cardiac function is probably the most widely known interaction of an endocrine organ with the heart. These effects often are separated into two classes: (a) those that are indirect and appear to be mediated by the sympathetic nervous system, and (b) those that are directly mediated by thyroid hormone.13 THE SYMPATHETIC NERVOUS SYSTEM AND THYROID DYSFUNCTION Various theories have been proposed to explain altered sympathetic nervous system function in hyperthyroidism and hypothyroidism. Thyroid hormone has been suggested as a cause of an altered interrelationship between the sympathetic nervous system and the heart by increasing the activity of the sympathetic nervous system and by increasing the sensitivity of cardiac tissue to an unchanged level of sympathetic nervous system activity. However, investigations seeking to show an increased activity of the sympathetic nervous system in hyperthyroidism have been conflicting and inconclusive. Studies of catecholamine levels in hyperthyroidism and hypothyroidism have shown that plasma and urinary levels of norepinephrine, epinephrine, and dopamine b-hydroxylase are low or normal in hyperthyroidism and normal or elevated in hypothyroidism.14 Therefore, the direction of alteration of catecholamine levels does not appear to account for the clinical manifestations in these two conditions. Despite these conflicting data, it has been observed that b-adrenergic blockers, such as propranolol, improve or even eliminate many of the cardiac manifestations of hyperthyroidism. This observation, plus the fact that the clinical manifestations of hyperthyroidism parallel those of catecholamine excess, have led to the proposal that thyroid hormone enhances the sensitivity of cardiac tissue to catecholamines. It has been clearly documented that the enhancement of catecholamine responsiveness induced by thyroid hormone results from an alteration in the activity of the b-adrenergic receptor– adenylate cyclase system.15 Exogenous triiodothyronine (T3) or thyroxine (T4) increases the number of b-adrenergic receptors.Moreover, an increase in b-adrenergic receptor affinity has been demonstrated in the presence of thyroid hormone. The observation that after b-adrenergic blockade dogs that are hyperthyroid have higher heart rates and greater myocardial contractility than when they were euthyroid supports the theory that thyroid hormone has an independent direct effect on the heart apart from that modulated by the sympathetic nervous system.16 Additional support for this theory comes from experiments showing that chick embryonic heart cells increase their rate of beating with the addition of thyroid hormone. There is an augmentation of myocardial contractility of the right ventricular papillary muscle from hyperthyroid cats, manifested by an upward shift of the myocardial force velocity curve, increased velocity of myocardial fiber shortening, a decreased time to peak ejection during isometric contraction, and an augmentation of peak tension. Pretreatment of these hyperthyroid cats with reserpine to deplete catecholamines did not alter the effects of hyperthyroidism. The direct effect of thyroid hormone on the heart appears to be mediated by changes in the level of messenger RNA (mRNA) for specific proteins.13 As in other tissues, thyroid hormone increases the activity of the sodium pump in cardiac cells. In hypothyroid rats administered T3, the activity of Na + , K+– adenosine triphosphatase (ATPase) and potassium-dependent p-nitrophenyl phosphatase in the heart increases by more than 50%.17 Thyroid hormone also increases the synthesis of myosin, as well as increasing its contractile properties by increasing the more mobile myosin isoenzymes.18 In thyrotoxicosis, there is an increased synthesis of a myosin isoenzyme with fast ATPase activity. This additional myosin ATPase may contribute to the increased cardiac contractility of the hyperthyroid heart, but it is unlikely to be the primary causative factor, because the administration of exogenous thyroid hormone increases contractility before an alteration of myosin ATPase activity occurs. Thus, an increasing body of evidence suggests that thyroid hormone also has extranuclear nonadrenergic-mediated effects. A specific effect on the activity of Ca2+-ATPases has been proposed.19 HYPERTHYROIDISM PHYSICAL FINDINGS Many of the prominent clinical manifestations of hyperthyroidism are those of the cardiovascular system, such as tachycardia, palpitations, and systolic hypertension (see Chap. 42). Diastolic hypertension also may be seen, but it is not as typical as systolic hypertension. Additional cardiac findings on physical examination include a hyperactive precordium, a loud first heart sound, an accentuated pulmonic component to the second heart sound, and a third heart sound; the Means-Lerman scratch, a systolic scratch thought to result from the rubbing together of the pericardial and pleural surfaces by a hyperdynamic heart, may be heard at the second left intercostal space during expiration. A systolic flow murmur, representative of the hyperdynamic state, may be heard along the left sternal border. Moreover, a murmur of mitral valve prolapse, a click alone, or a click and a murmur may be heard. There appears to be an increased prevalence of mitral valve prolapse in patients with hyperthyroidism. A series20 of 40 patients with active or previous hyperthyroidism had a 43% prevalence of mitral valve prolapse on echocardiogram compared with 18% of controls. HEMODYNAMIC EFFECTS Patients with hyperthyroidism may have angina pectoris and congestive heart failure. Initially, it was thought that congestive heart failure occurred only in patients with underlying heart disease. However, in infants with neonatal hyperthyroidism (see Chap. 47) congestive heart failure has been seen without underlying cardiac disease. Also, congestive heart failure may be induced in experimental animals made hyperthyroid. Therefore, although congestive heart failure is a more common clinical manifestation in patients with underlying heart disease, hyperthyroidism may overstress even a normal heart. Hemodynamic changes in hyperthyroidism include an increase in cardiac and stroke volume, mean systolic ejection rate, and coronary blood flow. The systolic ejection and/or ejection periods decrease, the pulse pressure widens, and systemic venous resistance decreases. The increase in cardiac output is greater than that expected by the increase in total body oxygen consumption; hence, it appears to be directly mediated by thyroid hormone effect on the heart independent of the effect of the thyroid hormone on general tissue metabolism. ELECTROCARDIOGRAPHIC FINDINGS Electrocardiographic changes are nonspecific in hyperthyroidism. Approximately 40% of these patients have sinus tachycardia.21 Also, atrial fibrillation occurs in ~15%. Intraatrial conductance disturbances may occur and be manifested by notching of the P wave in ~15% or by a prolonged PR interval in ~5% of these patients. Secondand third-degree heart block have also been observed, as well as intraventricular conductance defects, most commonly a right bundle branch block. THERAPY The most effective treatment for cardiac dysfunction in hyperthyroidism is treatment of the hyperthyroidism itself. Often, no further therapy is required. In patients with angina pectoris, ~15% have resolution of symptoms when they become euthyroid.22 A retrospective study of 163 patients with hyperthyroidism and atrial fibrillation reported that most revert to normal once they become euthyroid. The most commonly used treatment of the cardiac manifestations of hyperthyroidism is a b-adrenergic blocking drug. These drugs can decrease heart rate in patients with sinus tachycardia and slow the rate of atrial fibrillation. In some patients with congestive heart failure caused by a hyperdynamic state, b-adrenergic blockade may even improve the congestive heart failure, although it must be administered cautiously and under close supervision. When congestive heart failure is severe, cardiac glycosides can be used. Patients with hyperthyroidism require higher doses of these drugs because of an increased volume distribution and because the hyperthyroidism decreases the enhancement of myocardial contractility produced by digitalis. Atrial fibrillation reverts to normal sinus rhythm in most patients who become euthyroid. However, in patients who do not revert spontaneously, chemical or electrical cardioversion is indicated. Guidelines for the timing of cardioversion are based on the observations that 75% of hyperthyroid patients revert within 3 weeks of becoming

euthyroid, and no spontaneous reversion occurs when the patients remain in atrial fibrillation more than 4 months after being rendered euthyroid.23 Therefore, if a patient remains in atrial fibrillation for 16 weeks after achievement of the euthyroid state, cardioversion should be attempted. A series24 of 262 patients with hyperthyroidism and atrial fibrillation, reviewed in an 18-year retrospective study, showed a 10% incidence of arterial embolization. Patients with atrial fibrillation and mitral stenosis have a rate of embolization of 5% per year. Therefore, anticoagulation therapy should be used in patients with atrial fibrillation and hyperthyroidism unless there is a contraindication. Anticoagulation must be attempted cautiously, and with smaller doses than usual. The half-life of the clotting factors is shortened in hyperthyroidism; therefore, smaller-than-usual doses of warfarin derivatives can achieve adequate levels of anticoagulation. Finally, ipodate, used as a contrast agent for cholecystography, is beneficial in treating hyperthyroidism and its cardiac manifestations.25 HYPOTHYROIDISM CLINICAL MANIFESTATION The cardiac manifestations of hypothyroidism include cardiac enlargement with cardiac dilatation, sinus bradycardia, hypotension, distant heart sounds, edema, and evidence of congestive heart failure with ascites or orthopnea and paroxysmal nocturnal dyspnea. These full-blown clinical manifestations are seen infrequently, because hypothyroidism often is diagnosed at an earlier stage of the disease. Nevertheless, patients with hypothyroidism often complain of exertional dyspnea and, on physical examination, may have evidence of pleural effusions. HEMODYNAMIC EFFECTS The hemodynamic manifestations of hypothyroidism include a decrease in cardiac output, stroke volume, and blood and plasma volumes. Similar to the delay in the relaxation phase of skeletal muscle that is seen in hypothyroidism, there is also a prolongation of the isovolumetric relaxation time of the heart on echocardiography, which returns to normal with thyroid replacement.26 Contrary to the findings in hyperthyroidism, an increase in the preejection period and an increase in the ratio of preejection period to LV ejection time are observed. In severe hypothyroidism, there is increased capillary permeability leading to peripheral edema, with approximately one-third of all patients with myxedema developing pericardial effusions, increased interstitial edema, and pleural effusions seen on chest x-ray films. Thus, determining if these patients have congestive heart failure is often difficult. Invasive hemodynamic monitoring can differentiate these two conditions, because cardiac output rises with exercise in patients with hypothyroidism, whereas this rise does not occur if left heart failure is also present. Therefore, in most patients with hypothyroidism, although depressed myocardial contractility may be present, cardiac function remains sufficient to sustain the workload placed on the heart, because this workload itself is reduced. HYPERTENSION Patients with hypothyroidism have a higher incidence of hypertension than the normal population.27 In one series of 477 patients with hypothyroidism, 14.8% had hypertension (defined as a systolic/diastolic blood pressure above 160/95 mm Hg), as opposed to only 5.5% of 308 euthyroid patients who were age- and sex-matched. A significant correlation was observed between diastolic blood pressure and serum levels of T3 or T4. Of 14 patients who received thyroid replacement, 13 had normalization of their blood pressure.27 Because of changes in lipid metabolism observed in patients with hypothyroidism, the possibility of an increased risk of atherosclerosis has been raised. In hypothyroidism, there is elevation of cholesterol and triglycerides, and an impairment of free fatty acid mobilization, all of which are associated with premature coronary artery disease. Experimentally, atherosclerosis develops more readily in cholesterolfed hypothyroid animals, compared with euthyroid animals. Moreover, patients with hypothyroidism have about twice the frequency of coronary atherosclerosis as do age- and sex-matched controls. ELECTROCARDIOGRAPHIC FINDINGS The most common electrocardiographic finding in patients with hypothyroidism is sinus bradycardia. Also, there may be a low P-wave amplitude and prolongation of the QT interval.28 There also is an increased duration of the QRS interval. The prolongation of the QRS and QT intervals may predispose to reentrant rhythms and may be an explanation for the increased incidence of ventricular arrhythmias occurring in patients with hypothyroidism. Furthermore, in patients with pericardial effusions, the electrocardiogram may show decreased voltage. ENZYME ABNORMALITIES Laboratory abnormalities in patients with hypothyroidism may include an elevated creatine phosphokinase (CPK) concentration. This elevation may complicate the evaluation of patients with hypothyroidism and chest pain. The CPK level may be elevated in the absence of myocardial damage and, when levels are high, the MB band may be positive.29 THERAPY As with hyperthyroidism and cardiac dysfunction, the primary mode of treatment of cardiac dysfunction in the hypothyroid patient is correction of the thyroid abnormality itself (see Chap. 45). In patients who are hypothyroid and elderly, the thyroid hormone should be replaced cautiously because of the possibility of unmasking underlying organic heart disease. Depending on the clinical situation, an adequate starting dose of levothyroxine (L -thyroxine) is often 0.025 mg per day with an increase of 0.025 mg approximately every 2 weeks, with frequent monitoring of the patient's cardiovascular status. Serum thyroid-stimulating hormone (TSH) levels should be followed closely to determine when adequate replacement has been achieved. With the replacement of thyroid hormone, cardiovascular responses are rapid. In one series,30 where full replacement of thyroid hormone was achieved, improvement in the preejection period and in the preejection period/LV ejection time ratio was seen within 1 week of treatment. Moreover, CPK levels and serum cholesterol levels normalized within 1 week of therapy. In patients with true congestive heart failure, treatment is difficult because the heart's responsiveness to the cardiac glycosides is reduced substantially. The presence of ventricular arrhythmias should not be considered a contraindication for thyroid hormone replacement because these arrhythmias more often improve, rather than exacerbate, when thyroid hormone is given. Whether to treat patients with angina and hypothyroidism with thyroid hormone is uncertain because the angina may be exacerbated by hormone administration. However, there is also an increased risk during coronary artery bypass surgery in these patients if they are severely hypothyroid at the time of surgery. In most cases, at least partial thyroid hormone replacement is indicated, at which time, if angina persists, the patient may undergo coronary revascularization. After surgery, full thyroid replacement can be safely achieved.31 AMIODARONE AND THYROID FUNCTION The increasing use of amiodarone over the past 15 years for cardiac arrhythmia has substantially increased the number of cardiology patients with thyroid problems ranging from hypo- to hyperthyroidism. The most classic thyroid function test (TFT) findings in patients on amiodarone are upper range of normal T4, low range of normal T3, and a transient increase in TSH on initiation of treatment, with a subsequent fall to the low normal range.32,33 These changes result in the clinical manifestations of hypo- or hyperthyroidism in 2% to 24% of patients.32 Hypothyroidism is a more frequent manifestation in areas of adequate iodine intake such as the United States, whereas hyperthyroidism is more common in areas of low iodine intake such as Italy. The mechanisms of amiodarone-induced thyroid dysfunction remain unclear. Several possibilities that have been proposed include (a) disturbance in iodine autoregulation, (b) induction of thyroid autoimmunity, and (c) direct thyroid cytotoxicity. PATIENT EVALUATION AND DIAGNOSIS Because of the frequency of abnormalities in TFTs in patients on amiodarone and the potential impact of these abnormalities on cardiac function, routine testing is recommended. A suggested algorithim for management of patients receiving amiodarone is depicted in Figure 203-1.32 Clinical indications for TFTs include classic manifestations of hyper- and hypothyroidism as well as deterioration of the underlying cardiac disorder. It is unusual for hypothyroidism to present after the first 1½ years of treatment, whereas hyperthyroidism may present at any time. Diagnosis is made by the combination of corresponding clinical manifestations and an elevated or suppressed TSH.

FIGURE 203-1. An algorithm for following thyroid function in patients on amiodarone. *Test should be performed as a baseline reference for future test results.†Goal is to increase serum thyroxine (T4) to high normal or slightly above. (Do not attempt to normalize serum thyroid-stimulating hormone [TSH].) ‡Thionamides plus potassium perchlorate or prednisone for a short period of time might be helpful. b-Blockers can be added. (T3, triiodothyronine; AIH, amiodarone-induced hypothyroidism; AIT, amiodarone-induced hyperthyroidism.) (From Harjari KJ, Licata AA. Effects of amiodarone on thyroid function. Ann Intern Med 1997; 126:63.)

TREATMENT OF AMIODARONE-INDUCED THYROID DYSFUNCTION Treatment is difficult because of the long half-life of amiodarone and the necessity of continuing treatment for some patients. When hypothyroidism occurs, L -thyroxine can be used. However, the treatment of hyperthyroidism is more complicated. If the drug is stopped, hyperthyroidism may take more than 6 months to resolve. During this period or in those patients who continue amiodarone therapy, medical treatment options include the use of corticosteroids, propylthiouracil (PTU), and perchlorate. In some cases, near-total thyroidectomy is the best option to allow continuation of amiodarone treatment. Radioiodine thyroid ablation is usually not successful due to the high total body and thyroid iodine load and resultant low thyroid iodine uptake.

PARATHYROID DISEASES The effect of parathyroid disease on the heart has been thought to be primarily due to hypercalcemia or hypocalcemia. However, it has been documented that parathyroid hormone (PTH) has a direct effect on the heart.34 When PTH is added to isolated heart cells, there is an increase in chronotropy and inotropy.35 The direct effect of PTH on the heart is probably mediated by the binding of PTH to receptors, which increases entry of calcium into the myocardial cell. In rat heart cells, PTH causes early cell death; therefore, the direct effects of PTH in hyperparathyroidism may be harmful.35 PTH has a mixed effect on blood pressure and vascular reactivity. Intact PTH levels are closely related to blood pressure, particularly in elderly subjects.36 Subacute infusions of physiologic doses of PTH in humans modestly increase blood pressure.37 In contrast, some osteogenic fragments of PTH are hypotensive.38 HYPERPARATHYROIDISM In patients with chronic hypercalcemia, there may be deposition of calcium in the heart valve, coronary arteries, myocardial fibers, and fibrous skeleton of the heart.39 In hypercalcemic patients, the action potential plateau of cardiac fibers is shortened, thus decreasing the action potential duration that is reflected in the electrocardiographic finding of a shortened QT interval.28 Arrhythmias may occur.40,40a Patients with hyperparathyroidism and hypercalcemia have a higher prevalence of hypertension than normocalcemic matched controls. There are various mechanisms that have been proposed to account for the higher prevalence of hypertension. Whether PTH, itself, contributes to the pathogenesis of hypertension is unclear.41 A hypertensive parathyroid-like factor has been proposed as the etiologic factor,42 and it correlates with blood pressure in salt-sensitive essential hypertensive subjects.43 Hypercalcemia may cause hypertension by inducing renal failure from nephrocalcinosis. In addition, there is increased myocardial contractility, peripheral resistance, and vascular sensitivity to vasoconstrictor agents, such as AII and norepinephrine. The most likely cause of the hypertension in patients with hyperparathyroidism is an elevation of peripheral resistance. A significant number of patients have resolution of the hypertension after surgical cure of their hyperparathyroidism.44 HYPOPARATHYROIDISM The effects on cardiac muscle of hypocalcemia due to hypoparathyroidism are opposite those of hypercalcemia; there is action potential prolongation that may prolong the QT interval.28 Hypocalcemia also may have detrimental effects on cardiac function. In some patients with congestive heart failure and hypocalcemia, the heart failure did not respond to conventional therapy until the serum calcium concentration was normalized. Although experimentally a decrease in serum calcium concentration reduces cardiac contractility, heart failure is rarely seen in patients with hypocalcemia; therefore, it usually is the patient with an already compromised myocardial function who is susceptible to its effects. Besides the direct effects of the hypocalcemia on myocardial contractility, there is some evidence for a decrease in sodium excretion with hypocalcemia, which could be an additive factor in the development of congestive heart failure. Although most patients who develop myocardial dysfunction with hypocalcemia are patients with underlying cardiac disease, there have been several case reports of patients with no underlying cardiac disease who developed decreased myocardial function secondary to hypocalcemia resulting from hypoparathyroidism.45

DISEASES OF THE ADRENAL GLAND CUSHING SYNDROME Before the development of effective treatment for their disease, accelerated atherosclerosis was common in patients with Cushing syndrome; early death from either myocardial infarction, congestive heart failure, or stroke was the usual course. Although hypertension may partly contribute to the atherosclerosis, the accelerated atherosclerosis most likely is secondary to the lipid-mobilizing effects of cortisol. Excess cortisol production leads to hyperlipidemia and hypercholesterolemia, both of which may promote the development of atherosclerosis. The pathophysiology of the hypertension in Cushing syndrome is uncertain. Initially, it was thought to be secondary to volume expansion because of cortisol's mineralocorticoid properties. However, studies46 have been unable to support this hypothesis, except when the glucocorticoid levels exceed the capacity of the renal 11b-hydroxysteroid dehydrogenase. Other hypotheses include glucocorticoid potentiation of vascular smooth muscle response to the glucocorticoid itself47 or to vasoconstrictor agents,48 perhaps by suppressing nitric oxide production.49 Glucocorticoids can also increase renin substrate.50 With an increase in renin substrate, the elevated blood pressure would be secondary to the increased generation of AII. Finally, it has been proposed that glucocorticoids specifically increase renal vascular resistance, resulting in increased sodium retention.51 Thus, it is likely that the pathophysiology of the hypertension is multifactorial, with volume expansion, increased production of AII, and increased sensitivity of the vascular smooth muscle to glucocorticoids and vasoactive agents all having a role. In addition to efforts directed at lowering cortisol production, treatment of the cardiovascular manifestations of Cushing syndrome are aimed primarily at correcting the hypokalemia, if present, and lowering the blood pressure. Because these patients already have a tendency to lose potassium, treating the hypertension with potassium-losing diuretics must be done cautiously. Several studies have suggested that the hypertension may be more specifically treated with agents that block the release or action of renin (e.g., b-adrenergic blockers and converting enzyme inhibitors).50 Because of their tendency toward hypokalemia, patients with Cushing syndrome should be treated cautiously with cardiac glycosides (see Chap. 75). HYPERALDOSTERONISM Traditionally, aldosterone's effect on the cardiovascular system has been viewed as nonspecific and secondary to its effect on arterial pressure and potassium balance52 (see Chap. 80). Thus, T- or U-wave abnormalities on the electrocardiogram, and premature ventricular contractions, as well as other arrhythmias caused by the hypokalemia, often are observed.28 Evidence for LV hypertrophy, either on an electrocardiogram or on a chest x-ray film, also is common, particularly with long-standing primary aldosteronism. However, data suggest a much larger role for aldosterone in inducing cardiovascular injury.53 While LVH is common in primary aldosteronism, the LVH is greater than what would be expected for the degree of hypertension when compared to other secondary or

primary forms of hypertension.54,55 Arterial compliance is also inversely correlated to the level of aldosterone in subjects with hypertension.56 These data suggest that aldosterone may have a direct effect on collagen formation—a hypothesis strongly supported by experimental data.57 Aldosterone increases cardiac fibrosis independent of an AII effect in experimental animals.58 Experimental studies also suggest that aldosterone induces strokes and renal disease,59 which also occur with an increased frequency in the human hypertension that is associated with increased aldosterone production.60 Finally, blockade of mineralocorticoid receptors by agents such as spironolactone or epleranone (not available in the United States) inhibits the development of fibrosis and/or reverses it.61,62 ADRENAL INSUFFICIENCY The most common cardiovascular finding in adrenal insufficiency (see Chap. 76) is arterial hypotension. In severe cases, the pressure has been in the range of 80/50 mm Hg, with a substantial postural fall. Heart size and the peripheral pulse also may be decreased. Most patients with chronic adrenal insufficiency have an abnormal electrocardiogram, with low or inverted T waves, sinus bradycardia, and prolongation of the QTc interval.30 Twenty percent of these patients have conduction defects, with first-degree heart block being the most common. Paradoxically, changes secondary to hyperkalemia are not common, even though the serum potassium levels may be elevated. Interestingly, most of the cardiographic abnormalities do not respond to mineralocorticoid replacement, but do respond to glucocorticoid therapy. PHEOCHROMOCYTOMA The major cardiovascular manifestation of pheochromocytoma is hypertension. Episodic discharge of catecholamines, reduction in plasma volume, and impaired sympathetic reflexes all contribute to the lability of blood pressure in these patients. Although some studies indicate that an absolute reduction in plasma volume exists only in a few cases, a number of observations strongly support the hypothesis that chronic volume depletion, to some extent, is present in most untreated patients with pheochromocytoma. For example, severe hypotension occurs with a-adrenergic blockade or removal of the tumor and is correctable by volume expansion. Impaired peripheral vascular reflexes are suggested by (a) orthostatic hypotension, (b) an increase in heart rate accompanied by decreased stroke volume, and (c) inadequate adjustment in peripheral vascular resistance. In 75% of patients with pheochromocytoma, the electrocardiogram is abnormal.28 The most common abnormalities include LVH, T-wave inversion, sinus tachycardia, and other alterations in rhythm such as supraventricular ectopic beats or paroxysmal supraventricular tachycardia. Depending on the level of blood pressure, changes are observed that are suggestive of myocardial damage, including transient ST segment elevation, diffuse T-wave inversions, and ST segment depressions. Because not all of the electrocardiographic changes can be attributed to hypertension or myocardial ischemia, a specific catecholamine-induced myocarditis has been proposed.63 In one study, 50% of patients who died from pheochromocytoma had myocarditis, usually accompanied by LV failure and pulmonary edema.63 Pathologically, the myocarditis consisted of focal necrosis with infiltration of inflammatory cells, contraction band necrosis, perivascular inflammation, and fibrosis.63 Although there is evidence of atherosclerosis of the coronary arteries, medial thickening is the most common characteristic. Rats exposed to high levels of norepinephrine show similar changes in their coronary arteries.

GONADAL HORMONES AND THE HEART The effect of sex steroids on cardiovascular disease (CVD) has been investigated extensively, although there is still much debate over their role. The hypothesis that estrogens are protective against the development of atherosclerotic CVD is based on two observations. First, middle-aged men are at a higher risk for developing CVD than age-matched women. Second, this difference between men and women equalizes when postmenopausal women are compared with men of the same age. The mechanism whereby estrogens exert their protective effect is unclear and may involve effects on lipids, endothelial-dependent vasodilation, and blood pressure. POSTMENOPAUSAL HORMONE REPLACEMENT Lower total cholesterol and higher high-density lipoprotein (HDL) cholesterol levels are seen in premenopausal women and postmenopausal women on estrogen. In the Postmenopausal Estrogen/Progestin Intervention (PEPI) study64 of 875 women, higher HDL levels were seen in those taking unopposed estrogen (conjugated equine estrogen [CEE] 0.625 mg). However, the concomitant use of medroxyprogesterone, commonly used in clinical practice, tended to negate this benefit, while micronized progesterone did not. An improvement in endothelial-dependent vasodilation has also been reported with unopposed estrogen therapy. This effect may be blunted or negated when medroxyprogesterone is added, but is maintained with the addition of natural progesterone.65 Although early oral contraceptives, containing higher doses of estrogens and varying progestins, were reported to cause hypertension, the doses of estrogen currently used for hormone-replacement therapy (HRT) have not been associated with hypertension. CEE appears to have a neutral effect on blood pressure, whereas estradiol may actually lower blood pressure.66 Since estradiol is the primary estrogen of the premenopausal state, this may explain why hypertension is less common in prethan in postmenopausal women. The rise in blood pressure following menopause may be one of the several contributors to the increase in CVD. The potential benefit of estradiol for HRT in this regard has not been evaluated. Although epidemiologic studies had suggested a benefit to women using HRT on CVD, the Heart and Estrogen/Progestin Replacement Study (HERS)67 did not. In a study of 2673 postmenopausal women with already established coronary heart disease, the use of CEE, 0.625 mg per day, and medroxyprogesterone, 2.5 mg per day, led to no improvement in future cardiac events. Indeed, in the first year of study, there was an increase in venous thromboembolism in the treatment group. Whether a benefit would have been seen in a primary prevention study remains to be seen, and should be answered by the Women's Health Initiative, scheduled for completion in 2006. Perhaps, if CEE does not give the CVD benefit of the premenopausal state, other estrogens (e.g., estradiol) might, and further studies of the impact of different estrogens on CVD will need to be carried out. Furthermore, the selective estrogen-receptor modulators will need to be evaluated in this regard. ORAL CONTRACEPTIVES The use of oral contraceptive agents in the premenopausal woman has been associated with an increased risk of cardiovascular morbidity and mortality (see Chap. 104 and Chap. 105). A three- to four-fold increase in incidence of deep venous thrombosis (DVT) has been reported.68 This risk appears to be limited to women with other cardiovascular risk factors, such as smoking and hypertension. There also is a higher incidence of myocardial infarctions. These also may represent a thrombotic process, because in most myocardial infarctions a thrombosis forms on a preexisting substrate of a narrowing, which is secondary to atherosclerosis. This theory is supported by autopsy studies in women who have died from myocardial infarction while using oral contraceptives. They usually have thrombosis as a cause of their myocardial infarction.69 As with DVT, women without underlying CVD do not appear to have increased risk. Changes in intrinsic coagulability induced by the oral contraceptives may account for the increased risk of thrombosis. Women with factor V Leiden mutations have a five- to ten-fold increased risk of thromboembolism. In many women taking oral contraceptives, there is an elevation of blood pressure. The rise in blood pressure that develops in these patients is ascribed to the estrogen component, which increases the production of renin substrate by the liver. Although there is a small increase in blood pressure in many patients taking oral contraceptives, most do not develop clinical hypertension, suggesting that there are counterregulatory mechanisms that are brought into play in most women to compensate for the increase in renin substrate. It also has been theorized that women who develop clinical hypertension are those who are predisposed (e.g., family history of hypertension, renal disease). CHAPTER REFERENCES 1. Klee GG. Biochemical thyroid function testing. Mayo Clin Proc 1994; 69:469. 2. Ezzat S, Forster MJ, Berchtold P, et al. Acromegaly. Clinical and biochemical features in 500 patients. Medicine 1994; 73:233. 3. Aono J, Nobuoka S, Nagashima J, et al. Heart failure in three patients with acromegaly: echocardiographic assessment. Intern Med 1998; 37:599. 3a. Colao A, Baldelli R, Marzullo P, et al. Systemic hypertension and impaired glucose tolerance are independently correlated to the severity of the acromegalic cardiomyopathy. J Clin Endocrinol Metab 2000; 85:193. 3b. Fazio S, Cittadini A, Biondi B, et al. Cardiovascular effects of short-term growth hormone hypersecretion. Clin Endocrinol Metab 2000; 85:179. 4. Minniti G, Moroni C, Jaffrain-Rea ML, et al. Prevalence of hypertension in acromegalic patients: clinical measurement versus 24-hour ambulatory blood pressure monitoring. Clin Endocrinol 1998; 48:149. 5. Strauch G, Valloton MB, Touitou Y. The renin-angiotensin-aldosterone system in normotensive and hypertensive patients with acromegaly. N Engl J Med 1972; 287:795. 6. Moore TJ, Thein-Wai W, Dluhy RG, et al. Abnormal adrenal and vascular responses to angiotensin II and an angiotensin antagonist in acromegaly. J Clin Endocrinol Metab 1980; 51:215. 7. Lopez-Velasco R, Escobar-Morreale HF, Vega B, et al. Cardiac involvement in acromegaly: specific myocardiopathy or consequence of systemic hypertension? J Clin Endocrinol Metab 1997; 82:1047. 8. Fazio S, Cittadini A, Cuocolo A, et al. Impaired cardiac performance is a distinct feature of umcomplicated acromegaly. J Clin Endocrinol Metab 1994; 79:441.

9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40.

Ozbey N, Oncul A, Bugra Z, et al. Acromegalic cardiomyopathy: evaluation of the left ventricular diastolic function in the subclinical stage. J Endocrinol Invest 1997; 20:305. Baldelli R, Ferretti E, Jaffrain-Rea ML, et al. Cardiac effects of slow-release lanreotide, a slow-release somatostatin analog, in acromegalic patients. J Clin Endocrinol Metab 1999; 84:527. Nishiki M, Murakami Y, Sohmiya M, et al. Histopathological improvement of acromegalic cardiomyopathy by intermittent subcutaneous infusion of octreotide. Endocrinol J 1997; 44:655. Lombardi G, Colao A, Ferone D, et al. Effect of growth hormone on cardiac function. Horm Res 1997; 48:38. Polikar R, Burger AG, Scherrer U, Nicod P. The thyroid and the heart. Circulation 1993; 87:1435. Nishizawa Y, Hamada N, Fujii S, et al. Serum dopamine-beta-hydroxylase activity in thyroid disorders. J Clin Endocrinol Metab 1974; 39:599. Hammond HK, White FC, Buxton ILO, et al. Increased myocardial beta receptors and adrenergic responses in hyperthyroid pigs. Am J Physiol 1987; 252:H283. Rutherford JP, Vatner SF, Braunwald E. Adrenergic control of myocardial contractility in conscious hyperthyroid dogs. Am J Physiol 1980; 237:590. Philipson KD, Edelman IS. Thyroid hormone control of Na+/K+-ATPase and K+ dependent phosphatase in rat heart. Am J Physiol 1977; 232:C196. Chizzonite RA, Everett AW, Clark WA, et al. Isolation and characterization of two molecular variants of myosin heavy chain from rabbit ventricle.Change in their content during normal growth and after treatment with thyroid hormone. J Biol Chem 1982; 257:2056. Davis PJ, Davis FB. Acute cellular actions of thyroid hormone and myocardial function. Ann Thorac Surg 1993; 56(Suppl 1):S16. Channick BJ, Adlin EV, Marks AD, et al. Hyperthyroidism and mitral valve prolapse. N Engl J Med 1981; 305:497. Olshaasen K, Bischoff S, Kahaly G, et al. Cardiac arrhythmias and heart rate in hyperthyroidism. Am J Cardiol 1989; 63:930. Sandler G, Wilson GM. The nature and prognosis of heart disease in thyrotoxicosis. A review of 150 patients treated with 131 I. Q J Med 1959; 28:347. Nakazawa HK, Sakurai K, Hamada N, et al. Management of atrial fibrillation in the post-thyrotoxic state. Am J Med 1982; 72:903. Staffurth JS, Gibberd MC, Fui ST. Arterial embolism in thyrotoxicosis with atrial fibrillation. Br Med J 1977; 2:688. Chopra IJ, Huang T-S, Hurd RE, Solomon DH. A study of cardiac effects of thyroid hormones: evidence for amelioration of the effects of thyroxine by sodium ipodate. Endocrinology 1984; 114:2039. Shenoy MM, Goldman JM. Hypothyroid cardiomyopathy: echocardiographic documentation of reversibility. Am J Med Sci 1987; 294:1. Saito I, Ito K, Saruta T. Hypothyroidism as a cause of hypertension. Hypertension 1983; 5:112. Surawicz B, Manigiardi ML. Electrocardiogram in endocrine and metabolic disorders. In: Rios JC, ed. Clinical electrocardiographic correlation. Philadelphia: FA Davis Co, 1977:243. Goldman J, Matz R, Mortimer R, Freeman R. High elevation of creatine phosphokinase in hypothyroidism. An isoenzyme analysis. JAMA 1977; 283:325. Ladenson PW, Goldenheim PD, Cooper DS, et al. Early peripheral responses to intravenous L-thyroxine in primary hypothyroidism. Am J Med 1982; 73:467. Drucker DJ, Burrow GN. Cardiovascular surgery in the hypothyroid patient. Arch Intern Med 1985; 145:1585. Harjari KJ, Licata AA. Effects of amiodarone on thyroid function. Ann Intern Med 1997; 126:63. Newman CM, Price A, Davies DW, et al. Amiodarone and the thyroid: a practical guide to the management of thyroid dysfunction induced by amiodarone therapy. Heart 1998; 79:121. Stefenelli T, Mayr H, Berler-Klein J, et al. Primary hyperparathyroidism: incident of cardiac abnormalities and partial reversibility after successful parathyroidectomy. Am J Med 1993; 95:197. Bogin E, Massry SG, Harary I. Effect of parathyroid hormone on rat heart cells. J Clin Invest 1981; 67:1215. Morfis L, Smerdely P, Howes LG. Relationship between serum parathyroid hormone levels in the elderly and 24-h ambulatory blood pressures. J Hypertens 1997; 15:1271. Fliser D, Franek E, Fode P, et al. Subacute infusion of physiological doses of parathyroid hormone raises blood pressure in humans. Nephrol Dial Transplant 1997; 12:933. Whitfield JF, Morley P, Ross V, et al. The hypotensive actions of osteogenic and non-osteogenic parathyroid hormone fragments. Calcif Tissue Int 1997; 60:302. Roberts WC, Waller BF. Effect of chronic hypercalcemia on the heart: an analysis of 18 necropsy patients. Am J Med 1981; 71:371. Carpenter C, May ME. Case report: cardiotoxic calcemia. Am J Med Sci 1994; 307:43.

40a. Chang CJ, Chen SA, Tai CT, et al. Ventricular tachycardia in a patient with primary hyperparathyroidism. Pacing Clin Electro Physiol 2000; 23:534. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69.

Hulter HN, Melby JC, Peterson JC, Cooke CR. Chronic continuous PTH infusion results in hypertension in normal subjects. J Clin Hypertens 1986; 2:360. Pang PK, Benishin CG, Lewanczuk RZ. Parathyroid hypertensive factor: a circulating factor in animal and human hypertension. Am J Hypertens 1991; 4:472. Resnick LM. Calciotropic hormones in salt-sensitive essential hypertension: 1,25-dihydroxyvitamin D and parathyroid hypertensive factor. J Hypertens 1994; 12(Suppl):S3. Resnick LM. Calcium, parathyroid disease, and hypertension. Cardiovasc Rev Rep 1982; 3:1341. Suzuki T, Ikeda U, Fujikawa H, et al. Hypocalcemic heart failure: a reversible form of heart muscle disease. Clin Cardiol 1998; 21:227. Williamson PM, Kelly JJ, Whitworth JA. Dose-response relationships and mineralocorticoid activity in cortisol-induced hypertension in humans. J Hypertens 1996; 14(Suppl):S37. Walker BR, Best R, Shackleton CH, et al. Increased vasoconstrictor sensitivity to glucocorticoids in essential hypertension. Hypertension 1996; 27:190. Kalsner S. Mechanism of hydrocortisone potentiation of response to epinephrine and norepinephrine in rabbit aorta. Circ Res 1969; 24:383. Kelly JJ, Mangos G, Williamson PM, Whitworth JA. Cortisol and hypertension. Clin Exp Pharmacol Physiol 1998; 25(Suppl):S51. Krakoff L, Nicolis G, Amsel B. Pathogenesis of hypertension in Cushing's syndrome. Am J Med 1975; 58:216. Gordon RD. Mineralocorticoid hypertension. Lancet 1994; 344:240. Whitworth JA, Kelly JJ, Brown MA, et al. Glucocorticoids and hypertension in man. Clin Exp Hypertens 1997; 19:871. Pessina AC, Sacchetto A, Rossi GP. Left ventricular anatomy and function in primary aldosteronism and renovascular hypertension. Adv Exp Med Biol 1997; 432:63. Rossi GP, Sacchetto A, Pavan E, et al. Remodeling of the left ventricle in primary aldosteronism due to Conn's adenoma. Circulation 1997; 95: 1471. Shigematsu Y, Hamada M, Okayama H, et al. Left ventricular hypertrophy precedes other target-organ damage in primary aldosteronism. Hypertension 1997; 29:723. Blacher J, Amah G, Girerd X, et al. Association between increased plasma levels of aldosterone and decreased systemic arterial compliance in subjects with essential hypertension. Am J Hypertens 1997; 10:1326. Funck RC, Wilke A, Rupp H, Brilla CG. Regulation and role of myocardial collagen matrix remodeling in hypertensive heart disease. Adv Exp Med Biol 1997; 432:35. Sun Y, Ramires FJ, Weber KT. Fibrosis of atria and great vessels in response to angiotensin II or aldosterone infusion. Cardiovasc Res 1997; 35:138. Funder JW, Krozowski Z, Myles K, et al. Mineralocorticoid receptors, salt and hypertension. Recent Prog Horm Res 1997; 52:247. Litchfield WR, Anderson BF, Weiss RJ, et al. Intracranial aneurysm and hemorrhagic stroke in glucocorticoid remediable aldosteronism. Hypertension 1998; 31:445. Benetos A, Lacolley P, Safar ME. Prevention of aortic fibrosis by spironolactone in spontaneously hypertensive rats. Arterioscler Thromb Vasc Biol 1997; 17:1152. Stier CT Jr, Chander PN, Zuckerman A, Rocha A. Vascular protective effect of a selective aldosterone receptor antagonist in stroke-prone spontaneously hypertensive rats. Am J Cardiol 2000; in press. McManus BM, Fleury TA, Roberts WC. Fatal catecholamine crisis in pheochromocytoma: curable form of cardiac arrest. Am Heart J 1981; 102:930. Writing Group for PEPI Trial. Effects of estrogen or estrogen/progestin regimens on heart disease risk factors in postmenopausal women: the post-menopausal estrogen/progestin interventions (PEPI) trial. JAMA 1995; 273:199. Gerhard M, Walsh BW, Tawakol A, et al. Estradiol therapy combined with progesterone and endothelium-dependent vasodilation in postmenopausal women. Circulation 1998; 98:1158. Seely EW, Walsh BW, Gerhard MD, Williams GH. Estradiol with or without progesterone and ambulatory blood pressure in postmenopausal women. Hypertension 1999; 33:1190. Hulley S, Grady D, Bush T, et al. Randomized trial of estrogen plus progestin for secondary prevention of coronary heart disease in postmenopausal women. Heart and Estrogen/Progestin Replacement Study (HERS) Research Group. JAMA 1998; 280:605. Mosca L, Manson JE, Sutherland SE, et al. Cardiovascular disease in women. A statement for healthcare professionals from the American Heart Association. Circulation 1997; 96:2468. Psaty BM, Heckbert SR, Atkins D, et al. A review of the association of estrogens and progestins with cardiovascular disease in postmenopausal women. Arch Intern Med 1993; 153:1421.

CHAPTER 204 GASTROINTESTINAL MANIFESTATIONS OF ENDOCRINE DISEASE Principles and Practice of Endocrinology and Metabolism

CHAPTER 204 GASTROINTESTINAL MANIFESTATIONS OF ENDOCRINE DISEASE ALLAN G. HALLINE Hormones and Gastrointestinal Function Hormone-Secreting Tumors of the Gut Impact of Endocrine Disease on the Gut Hypothalamus-Pituitary Conditions Thyroid Disorders Goiter and Compression Syndromes Hyperthyroidism Hypothyroidism Physiologic/Laboratory Studies Medullary Thyroid Cancer and Multiple Endocrine Neoplasia Type 2B Syndrome Parathyroid Gland Disorders Hyperparathyroidism Hypoparathyroidism Adrenal Gland Disorders Adrenal Insufficiency Cushing Syndrome Pheochromocytoma Sex Hormone–Related Conditions Chapter References

HORMONES AND GASTROINTESTINAL FUNCTION Gastrointestinal function involves a complex process of food transport, digestion, and absorption, secretion, and excretion. The normal function of the alimentary tract depends on a coordinated interaction of many different hormones and neuroendocrine substances; consequently, disorders of the endocrine system may have a profound effect on the gut. The gastrointestinal manifestations of endocrine disease are discussed here. The gastrointestinal symptomatology associated with diabetes is discussed in Chapter 149. Moreover, various hormones and other chemical messengers that are synthesized within the gut serve to coordinate such important activities as gastric emptying, intestinal motility, biliary and pancreatic secretion, and mucosal absorption and secretion. Known collectively as the diffuse neuroendocrine system of the gut (see Chap. 175), these chemical messengers include peptides, prostaglandins, leukotrienes, serotonin, and histamine working in concert with the autonomic nervous system. The distinction between hormonal and neurologic influences becomes even more blurred with the localization of several of these gastrointestinal peptides (e.g., cholecystokinin) both within mucosal endocrine cells and the nerves of the gut (see Chap. 182). Other messengers such as vasoactive intestinal peptide (VIP) are located primarily in nerves and function as neurotransmitters, but they are released into local tissues or the bloodstream and also may serve a paracrine or hemocrine function.

HORMONE-SECRETING TUMORS OF THE GUT Certain hormone-secreting tumors of the gut are associated with profound gastrointestinal manifestations and may be part of, or even the presenting manifestation of, the multiple endocrine neoplasia (MEN) syndromes (see Chap. 188). The gut-related hormone-producing tumors (gastrinoma, vasoactive intestinal peptide–producing carcinoma [VIPoma], somatostatinoma, glucagonoma, carcinoid) and their clinical manifestations are discussed in Chapter 220 and Chapter 221. The Zollinger-Ellison (ZE) syndrome results from excessive neoplastic gastrin production, capable of inducing acid hypersecretion and severe ulceration of the duodenum and jejunum. This ulceration is often refractory to antacid or histamine H2-receptor blocker therapy and may require high-dose proton-pump inhibitors such as omeprazole. One-third of ZE patients have diarrhea, which may be due to excessive gastric acid production, mucosal damage of the small bowel, and increased pancreatic and small bowel secretions. Inactivation of pancreatic lipase by the low pH in the intestinal lumen or precipitation of bile salts may cause steatorrhea. Moreover, the rugal folds of the stomach may become thickened because of the trophic influences of gastrin on the mucosa. VIP is a known secretagogue that may affect the small bowel. An excess production of this hormone, the so-called VIPoma, may produce watery diarrhea, hypokalemia, and achlorhydria (WDHA syndrome)1 Other tumors producing glucagon, somatostatin, or serotonin also lead to well-characterized syndromes that may include effects on the gut. Tumor production of other chemical messengers, such as neurotensin, mammalian bombesin, and substance P, has not been associated with distinct clinical manifestations.

IMPACT OF ENDOCRINE DISEASE ON THE GUT HYPOTHALAMUS-PITUITARY CONDITIONS Diseases of the hypothalamus and pituitary gland usually are not associated with significant gastrointestinal complaints, except in relation to their effects on other hormones that likewise may alter digestive tract functions. The role of hypothalamic disease in disruption of human gastrointestinal physiology has yet to be clearly defined2 (see Chap. 9). Normal pituitary function is required to maintain gastric acid secretion, perhaps related to the production of growth hormone. An increased frequency of colonic adenomatous polyps has been reported in patients with acromegaly3 although others have found no increased prevalence of colonic polyps or other gastrointestinal neoplasms. Generally, in patients with diseases of the hypothalamus and pituitary gland, neurologic and other systemic manifestations are more significant than are gastrointestinal symptoms. THYROID DISORDERS Abnormal gastrointestinal function frequently is encountered in patients with thyroid disorders and may be the initial or sole manifestation of either the hyperthyroid or hypothyroid state. This may result either from a direct mechanical effect of an enlarged thyroid gland or through the effects of increased or decreased thyroid hormone on the gut. In addition, disease affecting both the thyroid gland and the gut is well documented. For example, many reports have demonstrated concomitant thyroid conditions and gastrointestinal dysfunction,4 including thyroid abnormalities in patients with Sjögren syndrome, pernicious anemia, gluten-sensitive enteropathy, nonspecific inflammatory bowel disease, primary biliary cirrhosis, and chronic active hepatitis—disorders commonly demonstrating immunologic features (see Chap. 42 and Chap. 197). GOITER AND COMPRESSION SYNDROMES Esophageal compression from thyroid enlargement is common, and dysphagia may be seen in up to 30% of cases referred for thyroidectomy.5,6 Dysphagia and symptoms of esophageal dysfunction occasionally may also be secondary to altered esophageal motility; in severe myxedema, megaesophagus rarely may be present. Patients with thyrotoxicosis also may develop dysphagia and evidence of altered esophageal motility7 In patients with large goiters, superior vena cava compression may lead to formation of “downhill” esophageal varices in up to 12%8 Rarely, upper gastrointestinal hemorrhage from esophageal varices may be the initial presentation of a substernal goiter9 HYPERTHYROIDISM Hyperthyroidism (see Chap. 42) commonly presents with increased appetite, weight loss, and more frequent bowel movements. Less often, thyrotoxic patients complain of watery stools and may even have steatorrhea; rarely, constipation is seen10 In elderly patients or in those with severe thyrotoxicosis, anorexia, nausea, and vomiting are sometimes seen and may contribute to weight loss11,12 and 13 Rarely, abdominal pain and functional duodenal obstruction have been reported14,15 The abdominal pain may be either chronic or acute in nature, at times simulating a “surgical” abdomen. Although the cause of this pain is unclear, it generally resolves promptly with correction of the hyperthyroid state. In the rare reported cases of functional duodenal obstruction, a history of significant weight loss is common and features similar to superior mesenteric artery syndrome are found14,15 When vomiting is significant, nasogastric suction may be required and treatment with rectal instillation of propylthiouracil has been successful15 Correction of hyperthyroidism results in gradual resolution of the obstruction. The degree of weight loss in hyperthyroidism is variable, but after successful treatment most patients return to their premorbid weight. Additional factors contributing to weight loss include increased

metabolic rate and a decreased absorption of nutrients from the gut. HYPOTHYROIDISM Hypothyroidism (see Chap. 45) may present with weight gain, diminished appetite, and clinical manifestations of altered intestinal motility, such as constipation, bloating, and flatulence. Usually, the weight gain is mild and is due to fluid retention that is associated with mucopolysaccharide deposition in various tissues and also occasionally to the presence of ascites (Fig. 204-1). However, weight loss and/or diarrhea may occur in hypothyroidism16 Intestinal hypomotility and pseudoobstruction are well described in hypothyroidism and may lead to unnecessary surgery17 Occasionally, intestinal atony may progress to massive bowel distention and result in perforation.17

FIGURE 204-1. A 52-year-old hypothyroid man with ascites. Bilateral pleural effusions also were present.

In hypothyroidism abdominal pain may also occur. In myxedema, ~10% of patients insidiously develop nonspecific abdominal discomfort. When accompanied by bowel distention and elevated serum carcinoembryonic antigen levels (frequently seen in this disorder), the clinical picture may mimic colonic malignancy. PHYSIOLOGIC/LABORATORY STUDIES Thyroid dysfunction can affect motility throughout the gastrointestinal tract18,19 and 20 In hyperthyroidism, oral-cecal transit time, evaluated by hydrogen breath tests, may be accelerated21 In hypothyroidism, gastric emptying is prolonged18 and delayed small bowel and colonic transit has been documented, occasionally presenting as paralytic ileus or even as megacolon22 The frequency of intestinal slow-wave electrical activity, a cyclical depolarization of smooth muscle and an important determinant of motility, is increased in hyperthyroidism and decreased in hypothyroidism23 Decreased salivary secretion has been described in both hyperthyroidism and hypothyroidism and may be secondary to coexisting Sjögren syndrome. Although decreased gastric acid secretion has been reported in many studies of both hyperthyroid and hypothyroid patients, a great deal of controversy remains, and these findings have not been confirmed19,24 In some cases, diminished acid production may be due to underlying pernicious anemia12 Thyroid diseases also have been associated with mild abnormalities in intestinal absorption. For example, steatorrhea is described in more than 25% of hyperthyroid patients and appears to be due to the ingestion of large quantities of fat, as well as to rapid intestinal transit and reduced contact time with the small bowel mucosa13 Although abnormalities in pancreatic and biliary secretion may be detected in hyperthyroidism, they probably are not responsible for fat malabsorption. Thyrotoxic patients tend to have decreased calcium absorption from the jejunum and a resultant negative calcium balance. This defect may be due to increased mobilization of bone calcium, decreased parathyroid hormone (PTH) activity, and lowered plasma 1,25-dihydroxyvitamin D levels; yet, other mechanisms may be involved25 There appears to be no relationship between this decreased calcium absorption and the presence or absence of steatorrhea. Elevated glucose tolerance curves are found in thyrotoxic patients, and flattened curves may be observed for hypothyroid patients. However, these findings may be accounted for by abnormal gastrointestinal transit, altered activity of the sympathetic nervous system, and changes in other carbohydrate-regulating hormones. 12 D-Xylose absorption is normal in most instances of hyperthyroidism or hypothyroidism, further suggesting the absorptive integrity of the small bowel Yet, increased incidences of lactose intolerance, gluten-sensitive enteropathy, and nonspecific inflammatory bowel disease are found in association with either of these thyroid disorders. Various histologic abnormalities of the gastrointestinal tract may be observed in patients with thyroid disorders, but no consistent abnormality has been confirmed. For example, there appears to be an increased frequency of atrophic gastritis in both hyperthyroidism and hypothyroidism, with an infiltration of lymphocytes and plasma cells occurring throughout the mucosa and submucosa of the entire gastrointestinal tract22,26 Furthermore, myxedematous infiltrations of the bowel wall occur in hypothyroidism, especially in the duodenum and colon2 However, these types of changes are not found universally, and their significance is uncertain. Ascites is seen in 2.5 g/dL) and a high serum-ascites albumin gradient (>1.1 g/dL)27 It is often accompanied by pleural and pericardial effusions, and generally responds well to hormonal replacement (see Fig. 204-1). MEDULLARY THYROID CANCER AND MULTIPLE ENDOCRINE NEOPLASIA TYPE 2B SYNDROME Up to 30% of patients with medullary thyroid carcinoma present with secretory diarrhea that is occasionally severe enough to cause dehydration and electrolyte abnormalities28 (see Chap. 40). Sometimes, it resolves on resection of the tumor but usually it occurs in patients with widely metastatic disease. The microscopic appearance of the small bowel is generally normal, and the etiology probably relates to altered secretory or absorptive function of the small intestinal mucosa. Others29 have demonstrated that rapid colonic transport and a concomitant decrease in absorption may be more important in the pathogenesis of diarrhea. Calcitonin probably plays a central role in the production of this diarrhea; indeed, the infusion of this hormone into normal patients has been shown to lead to decreased intestinal absorption of water and electrolytes30 There is, however, poor correlation between the presence of diarrhea and the level of calcitonin; therefore, other humoral substances may be active. Besides calcitonin, medullary carcinoma of the thyroid may produce several humoral substances, such as VIP, prostaglandins, serotonin, kallikrein, and gastrin, all of which may have an effect on small bowel secretion or motility. Treatment with somatostatin analogs and interferon-a may be effective in controlling diarrhea and flushing, since results with chemotherapy have been disappointing31 Medullary thyroid carcinoma is associated with mucosal and submucosal ganglioneuromatosis of the bowel in patients with the MEN2B syndrome (see Chap. 188), which may alter motility throughout the gut. More than 90% of patients with MEN2B have gastrointestinal complaints32 Symptoms of colonic dysmotility are often the earliest manifestation of this inherited disorder and may present within the first few months of life. Constipation is the most frequent complaint but diarrhea is also seen. Radiographic studies often demonstrate proximal colonic dilatation and rectosigmoid narrowing similar to that observed in Hirschsprung disease32 Colonic diverticulosis is also common and, rarely, patients may present with a perforated diverticulum. In the esophagus, motility disorders may occasionally be seen, and dysphagia has been described. PARATHYROID GLAND DISORDERS HYPERPARATHYROIDISM Symptomatic primary hyperparathyroidism (see Chap. 58) most commonly presents with renal or skeletal disease, but gastrointestinal manifestations may be seen in up to 35% of the patients. The reported frequency of abdominal symptoms varies according to the patient population and selection bias, the degree of hypercalcemia, and the duration of illness. With the advent of multichannel serum autoanalyzers, up to 50% of patients with hyperparathyroidism are being diagnosed on the basis of asymptomatic hypercalcemia33 The most dramatic presentation of gastroenterologic manifestations occurs in the syndrome of so-called acute hyperparathyroidism, when the serum calcium level often exceeds 17 mg/dL. Such patients often present with severe abdominal pain, nausea, vomiting, lethargy, dehydration, and altered mental status that may progress

to coma and seizures. In many of these cases, digestive tract symptoms predominate and peptic ulceration or pancreatitis is common34 Acute hyperparathyroidism can be life threatening and emergency parathyroidectomy may become necessary. Primary hyperparathyroidism also is associated with frequent gastrointestinal symptoms that include anorexia, weight loss, nausea, vomiting, abdominal pain, and constipation.34a Although early studies described an increased incidence of peptic ulceration in hyperparathyroidism, this association has been disputed35 Any true association between these disorders is most often the result of a concurrent gastrinoma as part of the MEN1 syndrome. In patients with MEN1 and coexisting ZE syndrome and hyperparathyroidism, resection of the parathyroids alone has resulted in a significant reduction in serum gastrin levels, acid secretion, and requirement for antiulcer medications, lending further support to the association between hyperparathyroidism and peptic ulceration. Hyperparathyroidism may be associated with acute or chronic pancreatitis in 3% to 7% of cases37 Elevated levels of both calcium and PTH have been implicated in the pathophysiology of this disease.37a Several examples of pancreatitis secondary to hypercalcemia from other causes, including administration of calcium chloride, vitamin D intoxication, and malignancy-associated hypercalcemia, have been reported37,38 PTH may also have direct toxic effects on the pancreas; however, none of these mechanisms has been unequivocally demonstrated in human subjects. In acute pancreatitis, hypocalcemia has been attributed to the formation of calcium–fatty acid soaps in areas of fat necrosis, and also, a recognized abnormality of calcium homeostasis characterized by an impaired mobilization of calcium from bone and inadequate PTH release. Clinically, severe pancreatitis presenting with normal calcium levels, when ordinarily some degree of hypocalcemia is expected, should raise the suspicion of hyperparathyroidism. Constipation may be seen in 30% of chronically hyperparathyroid patients and may relate to decreased smooth muscle tone or abnormal autonomic function of the gut. Moreover, hypercalcemia can lead to increased renal loss of free water, and to severe dehydration—a predisposing factor to constipation. An association between hyperparathyroidism and esophageal dysfunction has been suggested, with symptoms of reflux reported in up to 60% of patients39 Esophageal manometry demonstrated reduced lower esophageal sphincter pressure, which normalized in most of these patients after parathyroidectomy. Although calcium infusion reduces lower esophageal sphincter tone in normal subjects, this result is unconfirmed; therefore, its mechanism and significance remain uncertain. The prevalence of gallstone disease may be as high as 25% to 35% among patients with hyperparathyroidism, but a relationship between the two entities is controversial. Carcinoid tumors, especially of the foregut, occur in association with both the MEN1 and MEN2 syndromes40 Thus, hyperparathyroidism should be considered in patients with carcinoids41 HYPOPARATHYROIDISM Hypoparathyroidism (see Chap. 60) most commonly presents with neurologic and cardiac disease related to hypocalcemia; gastrointestinal symptoms are uncommon and a review42 failed to mention any symptoms referable to the digestive tract. Yet, an association between idiopathic hypoparathyroidism and pernicious anemia has been recognized (see Chap. 197). In one study,43 9% of patients with idiopathic hypoparathyroidism had a diagnosis of pernicious anemia, and a slightly higher percentage had circulating antibodies against parietal cells. Gastric acid secretion ceases with serum calcium levels below 7.0 mg/dL, and the restoration of normal acid output follows correction of the hypocalcemia. The clinical significance of this finding is uncertain. Several cases of idiopathic hypoparathyroidism in association with diarrhea and malabsorption have been reported. Usually, the small bowel architecture is normal, and the cause of the bowel dysfunction remains unclear. Correction of hypocalcemia generally improves the diarrhea44 However, one case report described a loss of small bowel villous architecture suggestive of celiac sprue; the patient responded to a gluten-free diet and not to correction of the hypoparathyroid state45 ADRENAL GLAND DISORDERS Many studies suggest that normal human gastric acid secretion depends on an adequate production of adrenal corticosteroids, and diseases both of the adrenal cortex and medulla frequently may be associated with gastrointestinal symptoms and dysfunction. Evidence for this comes from the documentation of achlorhydria in up to 50% of patients with Addison disease (see Chap. 76) as well as the frequent finding of gastric atrophy in these cases46 Repletion with corticosteroids restores normal gastric acid secretion in many of these patients, but contrary data also exist46 The most common form of adrenal insufficiency seen today, idiopathic adrenal gland atrophy, often is presumed to be due to autoimmune mechanisms, and immune-mediated disease of the stomach is also a possibility in this clinical entity. For example, an increased incidence of pernicious anemia occurs in patients with idiopathic adrenal insufficiency, and up to 30% of the latter may have circulating antibodies reactive against parietal cells or intrinsic factor47 Clearly, the precise role of normal adrenal function in the maintenance of gastric secretion requires continued study. ADRENAL INSUFFICIENCY Both acute and chronic adrenal insufficiency often are associated with gastrointestinal symptoms. Anorexia and weight loss are early manifestations in most cases and may progress to nausea and intractable vomiting48 Abdominal pain occurs in one-third of patients, although often mild and of a nonspecific character. In advanced cases, the abdominal pain occasionally may be severe and accompanied by tenderness and rigidity that simulate an acute “surgical” abdomen48 The etiology of these manifestations is unknown, but may be due, in part, to delayed gastric emptying49 or altered intestinal motility,50 and usually symptoms are relieved with correction of the hormone deficiencies. Nonetheless, adrenal insufficiency may be precipitated by the stress of intraabdominal sepsis and, thus, the clinical picture of an acute abdomen must be carefully assessed in these patients. Diarrhea and steatorrhea have been described in a few cases of adrenal insufficiency51 The loss of fecal fat, electrolytes, and other nutrients may contribute to weight loss and hypotension and generally is corrected upon supplementation with corticosteroids. CUSHING SYNDROME Cushing syndrome (see Chap. 75) is not associated with significant gastrointestinal disturbances, despite reports of increased serum gastrin levels52 and excessive gastric acid secretion in this condition; for example, the incidence of peptic ulceration is not increased53 In normal human volunteers, chronic corticosteroid administration can increase both basal and stimulated gastric acid output46; long-term corticosteroid or adrenocorticotropic hormone (ACTH) administration also increases basal serum gastrin levels and the serum gastrin response to an ingested meal54 Although there are some who believe that exogenously administered corticosteroids play no role in the development of peptic ulcer disease,55 others suggest that the incidence of ulcers actually may be slightly increased56 PHEOCHROMOCYTOMA Gastrointestinal manifestations generally are not a predominant part of the clinical presentation of pheochromocytoma (see Chap. 86). However, one study revealed a high incidence of nausea and vomiting (50%), weight loss (40%), and abdominal pain (35%)57 Several cases of adynamic ileus associated with pheochromocytoma have been reported, and constipation has been described in up to 8% of patients57 This may be due to high circulating levels of epinephrine and norepinephrine, both of which have an inhibitory effect on the smooth muscle of the gastrointestinal tract58 In fact, most cases have been associated with large adrenal tumors with the potential to release massive amounts of catecholamines and cause decreased alimentary motility. Spasm of the intestinal vasculature may lead to ischemia of the gut and, thus, cases with bowel infarction59 and gastrointestinal bleeding have been described58 Diarrhea and steatorrhea occasionally occur, with resolution of symptoms after resection of the pheochromocytoma. Although this may result from ischemic damage to the intestinal mucosa, occasionally diarrhea may be due to secretion by the tumor of other hormones, including VIP, somatostatin, calcitonin, or gastrin.60,60a Rare reports of hyperamylasemia and abdominal pain mimicking pancreatitis have been described in pheochromocytoma; however, there have been no documented cases of pancreatitis, and the amylase has been shown to arise from nonpancreatic sources61 SEX HORMONE–RELATED CONDITIONS Various gastrointestinal symptoms caused by increased levels of female sex hormones have been suggested, based on observations made during pregnancy and in patients receiving hormonal therapy or taking oral contraceptives. Heartburn is a frequent complaint in up to 50% of pregnant women, occurring most often during the third trimester62 During pregnancy, there is a progressive fall in lower esophageal sphincter pressure and a decrease in esophageal peristalsis that returns to normal after delivery63 This appears to be due to raised levels of progesterone, a known smooth muscle relaxant, but other hormones including estrogen also may be important. Studies in normal women taking oral contraceptives have shown a reversible decrease in lower esophageal sphincter pressure when they are given a

combination of the progestational agent, dimethisterone, and ethinyl estradiol in physiologic doses64 A great deal of controversy exists over the effect of pregnancy on preexisting inflammatory bowel disease; however, most authors agree that pregnancy does not influence the severity or the frequency of exacerbation of Crohn disease or ulcerative colitis65 The increased incidence of reversible ischemic colitis in young women using oral contraceptives, although rare, is well recognized66 Symptoms include abdominal pain, nausea, vomiting, and hematochezia. In mild cases, mucosal ulceration, erythema, or friability, often limited to the left colon, may be seen at colonoscopy. Usually, these findings resolve upon stopping the oral contraceptive; rarely, they may progress to mesenteric thrombosis and bowel infarction. This may be caused by an acquired resistance to activated protein C67 Estrogens are a rare cause of pancreatitis, but when this does occur, it is often seen in patients with underlying lipid disorders68 In such cases, severe hypertriglyceridemia in response to estrogen administration may develop with levels often exceeding 1000 to 2000 mg/dL68 Pancreatitis during pregnancy has been related to the presence of gallstones in up to 90% of cases69 In a small percentage of cases, pancreatitis may be secondary to gestational hypertriglyceridemia70 Small intestinal transit is slowed during the second and third trimesters of pregnancy,71 but the clinical significance of this is unclear. Vomiting during pregnancy is common, and severe hyperemesis gravidarum may lead to dehydration and electrolyte and acid-base disturbances. Several hormones have been implicated, including human chorionic gonadotropin, estrogen, progesterone, thyroxine, and others; however, studies remain inconclusive72 Gastric slow-wave activity, a determinant of stomach emptying, has been shown to be frequently disrupted during pregnancy73 These effects have been duplicated in nonpregnant women with administration of progesterone alone, or with progesterone and estradiol at levels comparable to those found during pregnancy73 CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34.

O'Dorisio TM, Mekhjian HS, Gaginella TS. Medical therapy of VIPomas. Endocrinol Metab Clin North Am 1989; 18:545. Grijalva CV, Novin D. The role of the hypothalamus and dorsal vagal complex in gastrointestinal function and pathophysiology. Ann N Y Acad Sci 1990; 597:207. Ladas SD, Thalassinos NC, Ioannides G, Raptis SA. Does acromegaly really predispose to an increased prevalence of gastrointestinal tumors? Clin Endocrinol 1994; 41:597. Counsell CE, Taha A, Ruddell WSJ. Coeliac disease and autoimmune thyroid disease. Gut 1994; 35:844. Anders HJ. Compression syndromes caused by substernal goitres. Postgrad Med J 1998; 74:327. Newman E, Shaha AR. Substernal goiter. J Surg Oncol 1995; 60:207. Branski D, Levy J, Globus M, et al. Dysphagia as a primary manifestation of hyperthyroidism. J Clin Gastroenterol 1984; 6:437. Schmidt KJ, Lindner H, Bungartz A, et al. Mechanical and functional complications in endemic struma. Munch Med Wochenschr 1976; 118:7. Glanz S, Koser MW, Dallemand S, Gordon DH. Upper esophageal varices: report of three cases and review of the literature. Am J Gastroenterol 1982; 77:194. Middleton WRJ. Thyroid hormones and the gut. Gut 1971; 12:172. Scarf M. Gastrointestinal manifestations of hyperthyroidism: an analysis of 80 cases of hyperthyroidism with a report of 4 cases masked by digestive symptoms. J Lab Clin Med 1936; 21:1253. Miller LJ, Gorman CA, Go VLW. Gut-thyroid interrelationships. Gastroenterology 1978; 75:901. Thomas FB, Caldwell JH, Greenberger NJ. Steatorrhea in thyrotoxicosis: relation to hypermotility and excessive dietary fat. Ann Intern Med 1973; 78:669. McClenathan JH, Wood BP. Hyperthyroidism as a cause of superior mesenteric artery syndrome. Am J Dis Child 1988; 142:685. Cansler CL, Latham JA, Browne PM, et al. Duodenal obstruction in thyroid storm. South Med J 1997; 90:1143. Stagias JG, Marignani P. Diarrhea, constipation, and hypothyroidism. J Clin Gastroenterol 1994; 18:347. Bergeron E, Mitchell A, Heyen F, Dube S. Acute colonic surgery and unrecognized hypothyroidism: a warning. Dis Colon Rectum 1997; 40:859. Kahraman H, Kaya N, Demircali A, et al. Gastric emptying time in patients with primary hypothyroidism. Eur J Gastroenterol Hepatol 1997; 9:901. Dubois A, Goldman JM. Gastric secretion and emptying in hypothyroidism. Dig Dis Sci 1984; 29:407. Wegener M, Wedmann B, Langhoff T, et al. Effect of hyperthyroidism on the transit of a caloric solid-liquid meal through the stomach, the small intestine, and the colon in man. J Clin Endocrinol Metab 1992; 75:745. Papa A, Cammarota G, Tursi A, et al. Effects of propylthiouracil on intestinal transit time and symptoms in hyperthyroid patients. Hepatogastroenterology 1997; 44:426. Duret RL, Bastenie PA. Intestinal disorders in hypothyroidism: clinical and manometric study. Dig Dis Sci 1971; 16:723. Christensen J, Schedl HP, Clifton JA. The basic electrical rhythm of the duodenum in normal human subjects and in patients with thyroid disease. J Clin Invest 1964; 43:1659. Miller LJ, Owyang C, Malagelada JR, et al. Gastric, pancreatic, and biliary responses to meals in hyperthyroidism. Gut 1980; 21:695. Peerenboom H, Keck E, Kruskemper HL, Strohmeyer G. The defect of intestinal calcium transport in hyperthyroidism and its response to therapy. J Clin Endocrinol Metab 1984; 59:936. Hellesen C, Friis T, Larsen E, Pock-Steen OC. Small intestinal histology, radiology and absorption in hyperthyroidism. Scand J Gastroenterol 1969; 4:169. de Castro F, Bonacini M, Walden JM, Schubert TT. Myxedema ascites: report of two cases and review of the literature. J Clin Gastroenterol 1991; 13:411. Hanna FW, Ardill JE, Johnston CF, et al. Regulatory peptides and other neuroendocrine markers in medullary carcinoma of the thyroid. J Endocrinol 1997; 152:275. Rambaud JC, Jian R, Fluorie B, et al. Pathophysiological study of diarrhea in a patient with medullary thyroid carcinoma: evidence against a secretory mechanism and for the role of shortened colonic transit time. Gut 1988; 29:537. Gray TK, Bieberdorf FA, Fordtran JS. Thyrocalcitonin and jejunal absorption of calcium, water, and electrolytes in normal subjects. J Clin Invest 1973; 52:3084. Lupoli G, Cascone E, Arlotta F, et al. Treatment of advanced medullary thyroid carcinoma with a combination of recombinant interferon alpha-2b and octreotide. Cancer 1996; 78:1114. O'Riordain DS, O'Brien T, Crotty TB, et al. Multiple endocrine neoplasia type 2B: more than an endocrine disorder. Surgery 1995; 118:936. St. Goar WT. Gastrointestinal symptoms as a clue to the diagnosis of primary hyperparathyroidism: a review of 45 cases. Ann Intern Med 1957; 46:102. Anglem TJ. Acute hyperparathyroidism: a surgical emergency. Surg Clin North Am 1966; 46:727.

34a. Lumachi F, Zucchetta A, Angelini F, et al. Tumors of the parathyroid glands. J Exp Clin Cancer Res 2000; 19:7. 35. Akerstrom G. Non-familial primary hyperparathyroidism. Semin Surg Oncol 1997; 13:104. 36. Norton JA, Cornelius MJ, Doppman JL, et al. Effect of parathyroidectomy in patients with hyperparathyroidism, Zollinger-Ellison syndrome, and multiple endocrine neoplasia type I: a prospective study. Surgery 1987; 102:958. 37. Carnaille B, Oudar C, Pattou F, et al. Pancreatitis and primary hyperparathyroidism: forty cases. Aust N Z J Surg 1998; 68:117. 37a. Mjaland O, Normann E. Severe pancreatitis after parathyroidectomy. Scand J Gastroenterol 2000; 35:446. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.

Fernandez del Castillo C, Harringer W, Warshaw A, et al. Risk factor for pancreatic injury after cardiopulmonary bypass. N Engl J Med 1991; 325:382, Mowschenson PM, Rosenberg S, Pallotta J, Silen W. Effect of hyperparathyroidism and hypercalcemia on lower esophageal sphincter pressure. Am J Surg 1982; 143:36. Duh QY, Hybarger CP, Geist R, et al. Carcinoids associated with multiple endocrine neoplasia syndromes. Am J Surg 1987; 154:142. Rode J, Dhillon AP, Cotton PB, et al. Carcinoid tumour of the stomach and primary hyperparathyroidism: a new association. J Clin Pathol 1987; 40:546. Breslau NA, Pak NYC. Hypoparathyroidism. Metabolism 1979; 28:1261. Blizzard RM, Chee D, Davis W. The incidence of parathyroid and other antibodies in the sera of patients with idiopathic hypoparathyroidism. Clin Exp Immunol 1966; 1:119. Peracchi M, Bardella MT, Conte D. Late-onset idiopathic hypoparathyroidism as a cause of diarrhoea. Eur J Gastroenterol Hepatol 1998; 10:163. Matsueda K, Rosenberg IH. Malabsorption with idiopathic hypoparathyroidism responding to treatment for coincident celiac sprue. Dig Dis Sci 1982; 27:269. Cushman P Jr. Glucocorticoids and the gastrointestinal tract: current status. Gut 1970; 11:534. Irvine WJ, Barnes EW. Adrenocortical insufficiency. Clin Endocrinol Metab 1972; 1:549. Sorkin SZ. Addison's disease. Medicine 1949; 28:371. Valenzuela G, Davis T, McGroaty D, et al. Primary adrenal insufficiency: a new cause of reversible gastric stasis. Am J Gastroenterol 1990; 85:1626. Phillips JD, Stelzner M, Zeng H, et al. Alterations in gastrointestinal motility during postoperative acute steroid withdrawal. Am J Surg 1991; 162:251. Tobin MV, Aldridge SA, Morris AI, et al. Gastrointestinal manifestations of Addison's disease. Am J Gastroenterol 1989; 84:1302. Lopez-Guzman A, Salvador J, Frutos R, et al. Hypergastrinaemia in Cushing's syndrome: pituitary origin or glucocorticoid-induced? Clin Endocrinol 1996; 44:335. Kyle J, Logan JS, Neill DW, Welbourn RB. Influence of the adrenal cortex on gastric secretion in man. Lancet 1956; 1:664. Seino S, Seino Y, Matsukura S, et al. Effect of glucocorticoids on gastrin secretion in man. Gut 1978; 19:10. Conn HO, Blitzer BL. Nonassociation of adrenocorticosteroid therapy and peptic ulcer. N Engl J Med 1976; 294:473. Messer J, Reitman D, Sacks HS, et al. Association of adrenocorticosteroid therapy and peptic ulcer disease. N Engl J Med 1983; 309:21. Hume DM. Pheochromocytoma in the adult and in the child. Am J Surg 1960; 99:458. Turner CE. Gastrointestinal pseudo-obstruction due to pheochromocytoma. Am J Gastroenterol 1983; 78:214. Salehi A, Legome EL, Eichorn K, Jacobs RS. Pheochromocytoma and bowel ischemia. J Emerg Med 1997; 15:35. Interlandi JW, Hundley RF, Kasselberg AG, et al. Hypercortisolism, diarrhea with steatorrhea, and massive proteinuria due to pheochromocytoma. South Med J 1985; 78:879.

60a. Van Eeckhart P, Shungu H, Descamps FX, et al. Acute watery diarrhea as the initial presenting feature of a pheochromocytoma in an 84-year-old female patient. Horm Res 1999; 52:101. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.

Perrier NA, van Heerden JA, Wilson DJ, Warner MA. Malignant pheochromocytoma masquerading as acute pancreatitis—a rare but potentially lethal occurrence. Mayo Clin Proc 1994; 69:366. Van Thiel DH, Gavaler JS, Joshi SN, et al. Heartburn of pregnancy. Gastroenterology 1977; 72:666. Singer AJ, Brandt LJ. Pathophysiology of the gastrointestinal tract during pregnancy. Am J Gastroenterol 1991; 86:1695. Van Thiel DH, Gavaler JS, Stremple J. Lower esophageal sphincter pressure in women using sequential oral contraceptives. Gastroenterology 1976; 71:232. Hanan IM, Kirsner JB. Inflammatory bowel disease in the pregnant woman. Clin Perinatol 1985; 12:669. Deana DG, Dean PJ. Reversible ischemic colitis in young women. Association with oral contraceptive use. Am J Surg Pathol 1995; 19:454. Mann DE, Kessel ER, Mullins DL, Lottenberg R. Ischemic colitis and acquired resistance to activated protein C in a woman using oral contraceptives. Am J Gastroenterol 1998; 93:1960. Glueck CJ, Lang J, Hamer T, Tracey T. Severe hypertriglyceridemia and pancreatitis when estrogen replacement therapy is given to hypertriglyceridemic women. J Lab Clin Med 1994; 123:59. McKay AJ, O'Neill J, Imrie CW. Pancreatitis, pregnancy and gallstones. Br J Obstet Gynaecol 1980; 87:47. Roberts IM. Hyperlipidemic gestational pancreatitis. Gastroenterology 1993; 104:1560. Lawson M, Kern F, Everson GT. Gastrointestinal transit time in human pregnancy: prolongation in the second and third trimesters followed by postpartum normalization. Gastroenterology 1985; 89:996. 72. Broussard CN, Richter JE. Nausea and vomiting of pregnancy. Gastroenterol Clin North Am 1998; 27:123. 73. Walsh JW, Hasler WL, Nugent CE, Owyang C. Progesterone and estrogen are potential mediators of gastric slow-wave dysrhythmias in nausea of pregnancy. Am J Physiol 1996; 270:G506.

CHAPTER 205 THE LIVER AND ENDOCRINE FUNCTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 205 THE LIVER AND ENDOCRINE FUNCTION NICOLA DE MARIA, ALESSANDRA COLANTONI, AND DAVID H. VAN THIEL Growth Hormone Effect of Growth Hormone on the Liver Effect of Liver Disease on Growth Hormone Homeostasis Thyroid Hormones Effect of Acute Liver Disease on Thyroid Hormone Homeostasis Effect of Chronic Liver Disease on Thyroid Hormone Homeostasis Hepatic Aspects of Drug-Induced Alterations of Thyroid Hormones The Liver in Thyroid Gland Disease Hypothalamic–Pituitary–Gonadal Axis Sexual Dimorphism of Hepatic Function Estrogens and the Liver Androgens and the Liver Sex Hormone Alterations in Cirrhotic Subjects Glucose Homeostasis and the Liver Glucose Homeostasis in Acute Liver Failure Congenital Defects of Hepatic Gluconeogenesis Hypoglycemia due to Liver Disease Glucose and Lipid Metabolism in Cirrhosis Diabetes Mellitus and the Liver Lipid Metabolism in Diabetes Mellitus Protein Metabolism in Diabetes Mellitus Biliary Lipid Secretion in Diabetes Mellitus Liver Diseases Associated with Diabetes Bone and Calcium Metabolism Alcoholic Liver Disease Effect of Alcohol on the Hypothalamic–Pituitary–Gonadal Axis Effect of Alcohol on Thyroid Function Effect of Alcohol on Growth Hormone and Prolactin Effect of Alcohol on Bone and Calcium Metabolism Liver Transplantation and Endocrine Function Chapter References

The liver is a hormone-responsive organ. Not only are numerous hepatic metabolic functions regulated by hormones, but the liver also is a principal site for hormone metabolism and clearance. In general, whereas the receptors for peptide hormones (e.g., insulin, glucagon, thyroid-stimulating hormone [TSH], adrenocorticotropic hormone [ACTH], and many small peptides [e.g., hypothalamic releasing factors]) are located on the surface of hepatocytes, the receptors for lipophilic hormones (e.g., steroid hormones) are located in either the cytosol or nucleus. Receptors for small neurotransmitters or signal transmitters present a mixed pattern: acetylcholine, a-adrenergic neurotransmitter, and b-adrenergic agents are located on the cell surface. In contrast, the receptors for thyroid hormone are located within the cell, on mitochondria, and in the nucleus. The response of liver cells to a hormone is dependent on the concentration of the hormone in extracellular fluid, the affinity of the hormone for its receptor, and the number of receptors present on or in the cell. At low hormonal concentrations, only those receptors with a high affinity are activated. With supraphysiologic amounts of hormone, nonclassic receptors can become activated and produce unusual biologic responses.

GROWTH HORMONE The growth-promoting actions of growth hormone (GH) are mediated by somatomedins (insulin-like growth factors) that are produced in the liver, as well as elsewhere, in response to GH stimulation (see Chap. 12). GH interacts with both “somatogenic” and “lactogenic” receptors located on the surface of hepatocytes1 The number of somatogenic receptors on liver cells is equal in men and women. In contrast, lactogenic receptors are present only in liver cells of women.2,3 Prolactin (PRL) (see Chap. 13), unlike GH, binds only to the lactogenic receptors on liver cells. The number of these receptors is controlled by the levels of GH and PRL, and each of the two receptor classes is regulated independently. EFFECT OF GROWTH HORMONE ON THE LIVER GH regulation of hepatic production of insulin-like growth factor-I (IGF-I) is mediated largely by somatogenic receptors (see Chap. 12)4,5 and 6 GH stimulates hepatic protein synthesis and induces hepatic enzymes that regulate the flow of substrates into a number of metabolic pathways7,8,9,10,11 and 12 Moreover, GH is thought to be the factor that determines the sexually dimorphic pattern of hepatic drug metabolism13 The complex influences of GH on hepatic carbohydrate and lipid metabolism include both insulin-like and insulin antagonistic effects. Hepatic somatomedin production is regulated predominantly by GH14,15; however, whether somatomedins are synthesized de novo or only released from storage sites remains unresolved. Subsequent to in vitro GH stimulation, protein synthesis inhibitors block IGF-I production, suggesting de novo synthesis rather than a simple release of stored IGF-I16 Malnutrition, a potent depressor of somatomedin, can block the effect of GH. Thus, somatomedin concentrations are reduced in response to protein/calorie malnutrition or fasting and are not increased in response to administration of exogenous GH in these conditions17 The precise mechanisms by which nutritional factors modulate somatomedin production and release are unclear. EFFECT OF LIVER DISEASE ON GROWTH HORMONE HOMEOSTASIS Because the liver is the major site of GH degradation, hepatocyte injury is associated with a reduction in GH clearance, and serum GH levels are increased in patients with chronic liver diseases. Thus, serum GH levels also correlate with the severity of the liver disease18,19 In cirrhotic patients, a wide variety of unusual physiologic stimuli (e.g., thyrotropin-releasing hormone [TRH] and oral glucose) enhance pituitary secretion of GH. In contrast, somatomedin concentrations are reduced in chronic liver diseases, correlating with the severity of the liver disease as manifested by a variety of biochemical indices (e.g., levels of serum albumin, alkaline phosphatase, and bilirubin)20

THYROID HORMONES The liver is the principal site for extrathyroidal thyroid hormone metabolism21 The liver is also responsible for sulfation, glucuronidation, and other phase II reactions of thyroid hormones. Thyroid hormone circulates bound to three different proteins that are synthesized in the liver: thyroxine (T4)–binding globulin (TBG), T4-binding prealbumin (TBPA), and albumin. These proteins determine the free T4 (FT4) level in plasma. EFFECT OF ACUTE LIVER DISEASE ON THYROID HORMONE HOMEOSTASIS Patients with acute hepatitis have elevated serum levels of T4 without clinical signs of hyperthyroidism22,23,24 and 25 The increase is caused by reduced clearance of thyroid hormone secreted by the thyroid gland and increased TBG concentrations occurring as part of the acute-phase response24 Both conditions resolve with clinical recovery. The increase in TBG concentration observed in acute hepatic injury may in part be due to an increased release from injured hepatocytes. A correlation is found between the changes observed in serum TBG levels and the aspartate aminotransferase levels that reflect the degree of hepatic injury23 The rise in serum T4 levels coincides with the rise in serum TBG levels, and the continued euthyroid state of the patient is reflected by a lowered triiodothyronine (T3) resin uptake and a normal FT4 index. If a mistaken diagnosis of hyperthyroidism is entertained, a normal TSH response to TRH stimulation should exclude this possibility. In acute

hepatitis, serum T3 levels are highly variable. The increased reverse T3 (rT3) levels normalize with recovery,23 and serum TSH concentrations generally are normal23 The hepatic conversion of T4 to T3 is reduced in cases of acute liver injury26 The rT3 levels are generally increased, however, whereas both total and free T3 (FT3) levels are reduced in acute hepatic failure27,28 and 29 Reduced levels of circulating T4, T3, and TBG are found in some cases of massive hepatic necrosis and in cases of fulminant hepatic failure27,28 and 29 The T3/T4 ratio has been proposed as an index of the severity of liver disease and as a prognostic factor in patients with fulminant hepatic failure and those with advanced liver disease awaiting liver transplantation29,30 Overall, the serum TSH level is usually normal in patients with acute hepatitis (TSH is considered to be the single best measure of thyroid hormone status in individuals with liver disease)31 EFFECT OF CHRONIC LIVER DISEASE ON THYROID HORMONE HOMEOSTASIS Alterations in thyroid hormone levels may occur with a variety of nonthyroidal illnesses, including liver diseases. Generally, these conditions are characterized by a decreased formation of T3 from T4 and an increase in serum levels of rT332,33,34,35 and 36 These changes occur as a consequence of a reduction in 5-monodeiodinase activity, which has the dual effect of decreasing T3 production and consequently the clearance of rT3. Cirrhosis and other chronic liver diseases are associated with several important changes in the indices of thyroid hormone metabolism. Several patterns of thyroid hormone levels are recognized in individuals with various chronic liver diseases. An increase in serum total T4 (TT4) level, due to an increase in the serum level of TBG, occurs commonly in individuals with primary biliary cirrhosis and autoimmune chronic active hepatitis24 The T3 resin uptake is reduced and the FT4 index is normal. T3 resin-binding ratios are ~50% higher in patients with chronic liver disease and primary biliary cirrhosis than in control subjects37 Many chronic parenchymal liver diseases, and also primary biliary cirrhosis, involve autoimmune phenomena, and coexistent autoimmune Hashimoto thyroiditis is found in 18% to 22% of such patients. The HLA antigens A1, B8, and DR3 are often found in subjects with HT and primary biliary cirrhosis, primary sclerosing cholangitis, and autoimmune chronic active hepatitis. One of the consistent findings in chronic liver disease is a reduction in the serum T3 level; lowest levels occur in end-stage cirrhosis. This finding is attributed to a reduced extrathyroidal and particularly hepatic conversion of T4 to T3 (resulting from reduced hepatic 5-monodeiodinase)38,39,40,41 and 42 The percentage of T4 that is coverted to T3 is reduced from 35.7% to 15.6% in cirrhotic patients 43,44 Besides the low serum T3, rT3 levels are typically increased in chronic liver disease, possibly due to a reduced clearance because of the reduced 5-monodeiodinase activity. Serum levels of rT3 have been used as a prognostic indicator for survival among patients waiting for orthotopic liver transplantation and those recovering from an episode of alcoholic hepatitis45,46,47 and 48 The circulating inhibitors of extrathyroidal conversion of T4 to T3, which are found in patients with liver cirrhosis,49,50 also act as thyroid hormone–binding inhibitors; they are incompletely characterized. (Free fatty acids [FFAs] have been proposed as likely candidates for this activity.) Patients with severe chronic liver disease, as well as those with other critical illnesses, often have reduced serum TT4 and T3levels. The circulating TT4 mconcentration is more severely reduced than is the FT4, which usually is normal51 Mixed patterns have been described in patients with very advanced liver cirrhosis, in which TT3 and TT4 levels, as well as FT3and FT4 levels, can be either increased or reduced. The absence of a consistent pattern in these individuals reflects the complexity of the metabolic alterations that occur in patients with end-stage liver disease. Despite all the changes occurring in thyroid hormone levels in cirrhotic individuals, little or no change in the serum level of TSH occurs in the absence of overt thyroid disease. Basal levels are either normal or slightly increased52,53,54 and 55 Moreover, the TSH response to TRH is either normal or slightly exaggerated. These findings indicate an intact hypothalamic-pituitary axis regulation of thyroid function. Nevertheless, both the elevated serum level of cortisol seen in severely ill cirrhotic patients and their overall poor nutritional state may inhibit the thyrotrope response to low T3 levels56,57 Thus, how the euthyroid state is maintained in patients with advanced liver disease and other critical illness in the face of markedly reduced T3 concentrations is not entirely clear. The major tissue effects of T3 are mediated via T3 nuclear receptor proteins encoded by c-erbA genes (see Chap. 33). Two classes of T3 receptors are found in humans (i.e., c-erbAa and c-erbA-b).58 The levels of c-erbA-a and c-erbA-b messenger RNA in monocytes and liver tissue are increased in cirrhotic patients.59 This finding suggests an increased rate of T3 receptor synthesis despite the presence of a reduced serum T3 level. As a result, an euthyroid state is achieved in target tissues where T3 receptors are expressed. HEPATIC ASPECTS OF DRUG-INDUCED ALTERATIONS OF THYROID HORMONES Several drugs inhibit the enzymatic monodeiodination of T4 to T3, potentially causing hyperthyroxinemia. Most important for hepatologists are glucocorticoids, propranolol, and certain iodinated contrast reagents used in hepatobiliary scanning and other imaging procedures. Two distinct monodeiodinase enzymes are found within microsomes: the type I enzyme is present both in the liver and in the kidney; the type II is found solely in the pituitary gland. The drugs that have been reported to cause hyperthyroxinemia by inhibiting these enzymes can be divided into two categories—drugs that inhibit only the type I enzyme alone and drugs that inhibit both enzymes. Drugs inhibiting only type I 5-deiodinase activity include high-dose propylthiouracil, dexamethasone, and propranolol60,61 and 62 Interestingly, propranolol is the only b-adrenergic blocker that inhibits the conversion of T4 to T3. The second group of drugs that inhibit both type I and type II enzymes includes amiodarone and the radiographic contrast agents iopanoic acid and ipodate, which are used for cholangiography63 All of these drugs contain iodine and have a structure that resembles that of thyroxine. The observed changes in thyroid function studies that occur after the use of these drugs include an increased serum T4 and a decreased serum T3 secondary to impaired peripheral conversion and clearance, a rise in serum rT3, and a rise in serum TSH secondary to impaired T4 conversion to T3. Cholecystographic agents also elevate T4 levels by inhibiting the hepatic uptake and binding of T4.63 Failure of a patient treated with amiodarone to develop hyperthyroxinemia should suggest the presence of hypothyroidism64,65 Finally, the clinical onset of Hashimoto thyroiditis after interferon-a therapy in patients with chronic viral hepatitis has been observed. Most such patients have low titers of antithyroid antibodies before starting interferon therapy.66 Conversely, the safety of interferon treatment in patients with evidence of immune dysfunction, including the presence of antithyroid autoantibodies, has been confirmed67,68 THE LIVER IN THYROID GLAND DISEASE HYPERTHYROIDISM Abnormal liver function tests may occur in 15% to 76% of individuals with hyperthyroidism65,66,67 and 68 The abnormalities observed include mild increases in serum glutamic-oxaloacetic transaminase (SGOT, AST), bilirubin, and, most commonly, alkaline phosphatase69 Alkaline phosphatase elevations can originate from liver or bone or both. The hepatic abnormalities seen in hyperthyroidism steadily improve slowly over many months as the patient returns to a euthyroid state. Although elevated serum bilirubin occurs in hyperthyroidism, clinical jaundice develops infrequently. The pathogenesis of the hepatic dysfunction leading to jaundice is unclear, but it may be due to tissue hypoxia. Importantly, excess thyroid hormone increases the oxygen requirements for mitochondrial metabolism70 Moreover, hyperthyroidism affects both overall bile production and the amount of bile acids present in bile. These thyroid hormone–induced changes in the bile acid production and pool size may contribute to the pruritus occasionally associated with hyperthyroidism71 HYPOTHYROIDISM No consistent or specific abnormalities of hepatic function are found in patients with hypothyroidism. Serum levels of SGOT may be abnormal in as many as 50% of these patients, however, suggesting coexistent hepatocellular disease. When the source of the transaminase elevation is investigated, however, concomitant increases in serum creatine kinase (MM isoenzyme) and aldolase levels usually indicate a skeletal muscle site of origin. In myxedema, release of these enzymes from muscle is increased and plasma clearance is decreased.

HYPOTHALAMIC–PITUITARY–GONADAL AXIS The normal function of the hypothalamic–pituitary–gonadal (HPG) axis is frequently disturbed by liver disease (see Chap. 16). Advanced cirrhosis in men is associated with phenotypic feminization, suggesting an alteration in plasma sex hormone levels. The findings of feminization are more pronounced in alcoholic individuals, because of their associated gonadal damage. The serum estrogen/androgen ratio of cirrhotic patients is usually increased; plasma testosterone and

dehydroepiandrosterone sulfate (DHEAS) levels are reduced; and estradiol levels range from normal to moderately elevated72,73 Because estrogens can act as tumor promoters or at least as hepatocyte proliferation stimuli, estrogens may play a cocarcinogenic role in the development of hepatocellular carcinoma (HCC) in men with cirrhosis74,75 In hemochromatosis, the abnormal accumulation of iron in the hypothalamus, pituitary, and gonads can impair normal hormonal function (see Chap. 116 and Chap. 131). In some patients, gonadotropin levels may be low despite advanced gonadal failure, indicating hypothalamic-pituitary involvement76,77,78,79,80 and 81 Thus, sexual dysfunction secondary to hypogonadotropic hypogonadism is a common complication of end-stage idiopathic hemochromatosis that can occur in the absence of clinically evident liver disease. Iron-mediated hepatic damage can play a contributory role late in the natural history of the disease. Hemochromatosis patients with cirrhosis have lower serum free testosterone and estradiol concentrations than do patients without cirrhosis.82 Historically, the description and characterization of sex hormone receptors have been limited to target tissues (e.g., ovary, testes, uterus, and prostate), sites where the levels of these somewhat labile proteins are maximal. Classic estrogen receptors (ERs), however, are present in both male and female liver83,84,85,86,87 and 88 These high-affinity ERs bind estradiol and various other estrogenic compounds (e.g., phytoestrogens, antiestrogens, and xenobiotics like chlorophenothane [DDT])83,87,88,89,90 and 91 Male liver contains ~25% of the ERs in female liver.83 Androgens repress and castration increases the male ERs. Similarly, estrogen treatment of men increases hepatic ER activity. In women, the liver contains ~25% as many cytosolic ERs as does the uterus; hepatic ER activity differs from that of classic steroid-receptive tissues. Much higher doses of estrogen are required to translocate the ERs of the liver to the nucleus of the cell. Presumably, this occurs because the cytosolic metabolism of estrogen by the liver limits the availability of the steroid for its receptor.92 Moreover, because of this metabolism within the liver cytosol, hepatic ERs are exposed to many different estrogenic metabolites and xenobiotics not found in classic ER-containing tissues. The hepatic ER in humans has high affinity and low capacity; it is specific for both steroidal and nonsteroidal estrogens93 Besides the cytosolic ER, human liver also contains ERs in the nuclear fraction,94 suggesting that they are functional.

SEXUAL DIMORPHISM OF HEPATIC FUNCTION Men and women differ markedly in their ability to metabolize different classes of xenobiotics95,96,97,98 and 99 Generally, men have a higher hepatic content of microsomal oxidative enzymes, and women have a greater capacity for reductive activity (Table 205-1). Androgens appear to be the major steroids that determine the “masculine” pattern of hepatic metabolic function. In fact, most of the metabolic functions that are elevated in men require androgen imprinting (a brief surge of testicular androgen synthesis early in life) to achieve and maintain adult levels of masculine activities. The imprinting process occurs at the level of the hypothalamus or pituitary. Whether a direct effect of imprinting on the liver also occurs is unclear. Some imprinted functions require the constant presence of androgen after puberty for expression, whereas others retain partial activity even after castration. Certain enzymes (e.g., 5a-reductase; see Chap. 114) are low in livers of adult males but high in livers of women, presumably because of a “feminizing factor.”

TABLE 205-1. Sexual Dimorphism in Hepatic Processes

The pituitary plays an important role in the development of the sexually dimorphic male and female patterns of hepatic steroid- and drug-metabolizing enzyme systems. Developmentally, male and female rats have similar male-like hepatic enzyme patterns until ~30 days of age. Subsequently, in females, the hypothalamus secretes a “feminizing factor” (apparently GH) that induces the female pattern of hepatic steroid metabolism.96,97 A sexual difference is seen in the pattern of GH release100 The male pattern of GH secretion is programmed by the action of neonatal androgens at the level of the hypothalamus, as are those for luteinizing hormone (LH) and follicle-stimulating hormone (FSH). Moreover, feminization of hepatic steroid metabolism occurs in hypothalamectomized (HPX) male rats given GH (twice daily), ACTH, and T4. Continuous administration of GH, mimicking the female pattern of GH secretion, feminizes hepatic steroid metabolism in normal and HPX males and “refeminizes” HPX and ovariectomized females101,102 Similarly, if male rats are castrated or treated with estrogen, a change to the female pattern of GH secretion is observed, concomitant with a change to the feminine pattern of hepatic steroid metabolism102,103 Collectively, these data suggest that female hepatic function, evidenced by higher hepatic PRL receptor levels, arises from continuous GH receptor occupancy, and that the male pattern originates as a result of a pulsatile pattern of GH receptor occupancy. ESTROGENS AND THE LIVER Estrogen administration increases the synthesis of sex steroid–binding globulin, TBG, transcortin, ceruloplasmin, and other secretory proteins104,105,106,107 and 108 Estrogen also increases the number of hepatocyte low-density lipoprotein (LDL) receptors, thereby increasing hepatic uptake of LDL cholesterol from the blood109 Moreover, ovariectomized female rats given estrogen display an increase in very-low-density lipoprotein (VLDL) triglycerides, but not in cholesterol, compared with controls. Estrogens promote the development of hepatic neoplasms associated with increased hepatocyte regenerative activity110 For example, estrogens promote diethylnitrosamine-induced liver tumors, and changes are found in the content and the distribution of ERs in the livers of patients with oral contraceptive–associated focal nodular hyperplasia and hepatic adenomas. This indicates a possible association between estrogen and its receptor, and the processes of hepatocyte proliferation and neoplasia. The hepatocyte ER content increases sharply after partial hepatectomy, both in the cytosol and in the nuclear fraction. Simultaneously, the hepatic AR content declines in both the cytosol and the nucleus. The plasma estradiol levels increase, whereas testosterone levels decrease111,112 These changes do not occur after abdominal surgery without hepatic resection. Moreover, both in vitro and in vivo, tamoxifen, an antiestrogen, inhibits the proliferative response of liver cells113 These data strongly suggest that estradiol contributes to the hepatic regenerative process, and a feminization of the liver appears to be essential during the proliferative response of the liver to an injury114,115 No alterations in the affinity of the ERs for estrogen occur with regeneration. The redistribution of the ERs from the cytosol to the nucleus occurs at the same time as the stimulation of DNA synthesis and reaches a maximum at 48 hours (the time at which the highest mitotic index occurs). Thus, the active translocation of the ERs into the nucleus correlates with several other markers of hepatocyte regeneration, including important biochemical functions (e.g., DNA polymerase activity, protein synthesis, and deoxythymidine kinase activity). The effect of ER translocation after partial hepatectomy may be related to an increase in serum estradiol or to an increase in hepatic intracellular estradiol, resulting from a decrease in the estrogen-metabolizing capacity of the liver remnant. The fact that estrogen administration to nonhepatectomized rats causes a translocation of ERs to the nucleus, as well as increased liver weight and DNA synthesis, supports this view. ESTROGEN-INDUCED LIVER DISEASE Numerous examples of liver-estrogen interactions have been associated with human liver disease. Women exposed to oral contraceptive steroids are at increased risk for hepatic neoplasms, jaundice, cholestatic hepatitis, gallstones, and hepatic vein thrombosis (Budd-Chiari syndrome). Pregnant women, who are exposed to gestational levels of both estrogenic and progestational hormones, develop liver abnormalities (e.g., fatty liver of pregnancy, intrahepatic cholestasis of pregnancy, and gallstones).116 The liver responds to estrogen exposure and also plays a major role in the metabolism of sex steroid hormones, converting them to less potent compounds. Many of these metabolites can alter the rate of metabolism and excretion of other hormones and drugs, further influencing the interaction between these

compounds and the hepatocyte. For example, catechol estrogens, although weak, are noncompetitive inhibitors of the demethylation of mestranol, a 17a-ethinyl estrogen found in many oral contraceptive steroids. A reduction in their catabolism could expose the hepatocyte to higher levels of a potential hepatotoxin for long periods and thereby increase the likelihood of an untoward estrogen-hepatocyte interaction. ANDROGENS AND THE LIVER Androgenic hormones exert powerful effects on the liver, whether directly or indirectly through their influence on the hypothalamus and pituitary. The presence of a hepatic androgen receptor (AR) is receiving increasing attention, especially with respect to its relationship with HCC. The AR content is higher in tumoral tissue than in nontumoral portions of the liver,117 perhaps explaining the higher worldwide incidence of HCC in men than in women. This finding also supports the use of antihormonal therapy for the treatment of HCC, particularly in patients who express high AR levels118 ANDROGEN-INDUCED LIVER DISEASE The use of androgenic anabolic steroids is associated with a significant risk of liver disease. The most common manifestation is a mild hepatic dysfunction without jaundice that resolves without sequelae on discontinuation of the drug. In contrast, androgen-induced cholestasis can be severe and may take weeks or months to resolve. The cholestasis is caused by a direct hepatotoxic effect of the drug, related to the alkyl group on the steroid C17 position. Androgens also interfere with the excretion of conjugated bilirubin into the canaliculus. Other problems that occur with androgen treatment include peliosis hepatis and hepatic neoplasia. Some of the reported cases of HCC associated with androgen use have been disputed on pathologic grounds, however; some of them actually appear to be hepatic adenomas. Because many of the patients treated with anabolic steroids also have Fanconi anemia (which is associated with malignancy), independent of androgen exposure, the HCC occurring in androgen-treated patients may be a component of the underlying Fanconi syndrome rather than a consequence of the administered androgen. SEX HORMONE ALTERATIONS IN CIRRHOTIC SUBJECTS Gynecomastia, impotence, loss of libido, and changes in body hair distribution are typical signs and symptoms in male patients with cirrhosis. The disturbances observed in the HPG axis in patients with liver disease vary, depending on the cause of the liver disease. Hypogonadism and feminization are common among individuals with chronic alcoholic liver disease. Hypogonadism alone is common in advanced hemochromatosis, whereas in patients with non–alcohol-related cirrhosis and in those without iron overload, such signs are much less prevalent and, when present, are much less severe119,120 and 121 Limited data are available concerning the alterations of the HPG axis in indi- viduals with advanced non–alcohol-related liver disease. The levels of circulating testosterone and dehydroepiandrosterone are slightly decreased in nonalcoholic cirrhotic patients, whereas estradiol concentrations are usually normal. These abnormalities depend on the severity of liver disease and are more pronounced in individuals with higher Child-Pugh scores119,122 Gynecomastia and impotence occur in cirrhotic men and are enhanced by the frequent long-term use of the diuretic drug spironolactone (which is a competitive antagonist of aldosterone at mineralocorticoid receptors and can also bind to testosterone receptors). Spironolactone reduces testosterone levels and slightly increases estradiol levels.123 Among women with non–alcohol-related liver cirrhosis, amenorrhea occurs in up to 50% of cases. Irregular bleeding, oligomenorrhea, or metrorrhagia occurs in those with chronic liver disease without cirrhosis. As a consequence, conception and pregnancy are uncommon in women with advanced chronic liver disease. Many cirrhotic women manifest a state of hypogonadotropic hypogonadism that can be related to both malnutrition and an impaired intermediate metabolism.124,125 In other cases, the plasma levels of both estradiol and testosterone are increased, in part due to a portosystemic recirculation of weak androgens with subsequent peripheral conversion to estrogens in fat as well as in other tissues. In cases of advanced liver disease complicated by the presence of hepatic encephalopathy, the central release of neuromediators (i.e., dopamine and norepinephrine) is altered, influencing the pulsatile release of gonadotropin-releasing factors from the hypothalamus. This may adversely affect neuroendocrine gonadal interrelationships. Decreased central levels of dopamine may contribute to the well-known hyperprolactinemia of cirrhotic patients, further compromising the release of gonadotropins.124

GLUCOSE HOMEOSTASIS AND THE LIVER The liver occupies a unique position in protein, carbohydrate, and lipid homeostasis. After feeding, the liver actively takes up glucose and uses it for fuel or stores it as glycogen. During fasting, the liver synthesizes glucose from amino acids. Because of these two opposing functions of storage and synthesis, the liver maintains a relatively stable plasma glucose concentration in the fasting and the fed state. With the ingestion of a meal, a large bolus of glucose is delivered to the liver via the portal vein. The fact that the peripheral glucose concentration rises only slightly (50% above baseline) after a meal implies that the liver is the principal site of glucose uptake after a meal. With fasting or hypoglycemia, hepatic glycogen and peripheral proteins are broken down and made available to the liver as glucose precursor, so that the crucial energy needs of the body can be maintained. The tight homeostatic control of peripheral glucose levels is maintained by hepatic glucose uptake in response to increases in portal venous glucose. This process does not appear to be insulin dependent, but results from the enzyme kinetics of the initial step in glucose handling by the hepatocyte. Specifically, by facilitated diffusion, glucose is taken up by the liver, and within the range of glucose concentrations experienced by the liver, this mechanism is saturated. Once glucose enters the hepatocyte, it is rapidly phosphorylated to glucose-6-phosphate (G6P) by glucokinase. Although hepatic glucokinase is in relative excess compared with its substrate concentration, the amount of this enzyme within the liver is under the control of glucose and insulin. In the fasting state, or with acute insulin deficiency, the level of hepatic glucokinase declines. Conversely, during the fed state, the intrahepatic level of the enzyme increases, as does the glucose level.126 Once glucose has been phosphorylated, it can be stored as glycogen or broken down to acetyl coenzyme A and further converted to an amino acid, fatty acid, or energy. In the fed state, glycogen synthesis predominates. Insulin also appears to have an important effect on hepatic glucose production. Low doses of insulin inhibit glycogen breakdown, and high doses inhibit glucose synthesis from amino-acid precursors. Thus, low doses of insulin lead to hepatic glycogen accumulation by inhibiting its breakdown. At higher levels of insulin, both glycogenolysis and gluconeogenesis are inhibited. The net result is a smaller increase in the hepatic venous glucose concentration than would be expected, with storage of most of the ingested glucose as glycogen. In the face of high splanchnic glucose concentrations, insulin stimulates a net hepatic glucose uptake.127,128 Neither the ratio of insulin to glucagon nor the absolute amount of glucagon appears to be important in this process. This difference between portal venous and systemic insulin levels suggests that the liver is the major site of glucose regulation and is the main site of insulin degradation. This occurs because of binding of insulin to specific insulin receptors on the hepatocyte surface, followed by internalization of the receptor and its bound hormone, with the subsequent breakdown of the bound insulin by cytosolic proteases. The receptor is then recycled to the cell surface. Thus, the amount of insulin degraded by the liver is determined by the amount bound to hepatocyte receptors, which, in turn, depends on a variety of stimuli that control insulin receptor number, affinity, or both. GLUCOSE HOMEOSTASIS IN ACUTE LIVER FAILURE Whereas glucose intolerance and insulin resistance are common features in acute hepatitis, hypoglycemia is a frequent complication of massive hepatic necrosis129,130,131 and 132 In acute hepatic failure, hypoglycemia can be symptomatic, persistent, and severe. Its prevention is based on a close monitoring of blood glucose levels as well as on the intravenous administration of high glucose loads. Also, because insulin and glucagon promote hepatic regeneration, their administration may have a beneficial effect in the treatment of acute liver failure.133 The ability of the liver to produce glucose (as a result of glycogenolysis as well as gluconeogenesis) is reduced to a critical level in subjects with acute liver failure134 In addition, a state of hyperinsulinemia occurs, as a result of the loss of the normal metabolism of insulin by the liver. This situation is aggravated further by the portosystemic shunting of blood because of liver necrosis135 CONGENITAL DEFECTS OF HEPATIC GLUCONEOGENESIS A deficiency of the enzyme fructose-1,6-diphosphatase occurs in humans.136 Individuals with this deficiency are chronically ill, with metabolic acidosis and hepatomegaly, and manifest potentially lethal episodes of hypoglycemia that often are precipitated by infection or other metabolic stresses. The glycogenolytic system in these patients is intact; glucagon administration causes a hyperglycemic response if performed in the postprandial period. After a prolonged fast, however, no response occurs. This enzyme deficiency can be established by liver biopsy with subsequent assay of the activity of the enzyme fructose-1,6-diphosphatase. Children with an absolute deficiency of the enzyme phosphoenolpyruvate carboxykinase manifest lactic acidosis and hypoglycemia associated with disturbances in pyruvate oxidation and organic acidemia (see Chap. 161).137

HYPOGLYCEMIA DUE TO LIVER DISEASE Hypoglycemia commonly occurs in Reye syndrome because of mitochondrial injury and the development of transient deficiencies of the mitochondrial enzymes pyruvate carboxylase and pyruvate dehydrogenase. The acquired deficiencies markedly limit the ability of the liver to convert gluconeogenic precursors, such as alanine and pyruvate, into glucose. A similar form of hypoglycemia can occur in women with acute fatty liver of pregnancy. Hepatotoxin-induced liver disease, particularly that produced by mitochondrial toxins (e.g., phosphorus, chloroform, carbon tetrachloride, propylene glycol, and halothane) is frequently associated with hypoglycemia for similar reasons. These agents reduce gluconeogenesis because of transient hepatic injury or lethal hepatic necrosis, which is associated with a loss of pyruvate decarboxylase and pyruvate dehydrogenase activities within the liver. Not unexpectedly, hypoglycemia can occur with massive hepatic necrosis due to any agent, virus, or drug, if the injury is severe enough or if hepatic infarction occurs.138,139 Fortunately, the hepatic reserve for gluconeogenesis is substantial, and 85% to 95% of the liver must be dysfunctional before clinically important hypoglycemia occurs. When this does occur, the resulting hypoglycemia can be severe and usually exceeds that which can be explained by a reduction in gluconeogenesis alone. It seems to be compounded further by a loss of hepatic mechanisms for insulin degradation and glycogenolysis. Thus, the combination of insulin excess, inadequate gluconeogenesis, and inadequate glycogenolysis, all occurring together, produce the hypoglycemia. Because of the considerable hepatic reserve for gluconeogenesis, no consistent, direct association is found between hypoglycemia and any common hepatic injury or any laboratory parameter of hepatic function. HCC is occasionally associated with the development of hypoglycemia, which occurs because of tumor production of massive amounts of insulin-like growth factors (see Chap. 219). GLUCOSE AND LIPID METABOLISM IN CIRRHOSIS The liver disposes of the glucose load after meals and clears the portal venous blood of most of its insulin and glucagon. Much of the glucose cleared by the liver after a meal is stored within that organ as hepatic glycogen or is converted to triglycerides. High portal venous levels of insulin, low portal venous levels of glucagon, and an intact liver are required for these events to occur at normal rates.52,140,141 Several abnormalities of glucose metabolism occur in individuals with cirrhosis.52,142,143,144 and 145 In end-stage liver disease, fasting hypoglycemia occurs as a result of impaired glycogen synthesis, impaired gluconeogenesis, and poor nutrition. Moreover, hypoglycemia indicates a very poor short-term prognosis. Normally, after an overnight fast, 70% to 80% of the glucose available in plasma is derived from hepatic glycogen. This decreases to 30% to 40% in cirrhotic patients, who have both a reduced hepatic glycogen content and liver cell dysfunction.52,142,143,144,145 and 146 The prevalence of overt glucose intolerance in cirrhotic patients varies greatly (54–92%).147,148,149 and 150 Up to 40% of cirrhotic patients are diabetic147 The pathogenic mechanisms responsible for the glucose intolerance in cirrhotic individuals are incompletely understood, although a combination of impaired insulin secretion rate and reduced insulin sensitivity may be responsible. An elevated postprandial circulating insulin level and hyperglycemia are consistent features of advanced cirrhosis and strongly suggest peripheral insulin resistance. An increased rate of insulin secretion in response to intravenous hyperglycemic stimuli occurs in cirrhotic patients with impaired glucose tolerance and mild diabetes. The insulin response to an oral glucose load, however, appears to be blunted and is not significantly greater than that of control subjects until 179;90 minutes after the ingestion of a glucose load. An impaired glucose tolerance in response to an oral glucose load has been reported in up to 80% of cirrhotic patients.149 Moreover, an abnormal intravenous glucose intolerance (defined as a 2-hour serum glucose level of >7.8 mmol/L after 25 g intravenous glucose) was found in 19% of cirrhotic patients. The augmented insulin response to an intravenous glucose load supports the concept of an impaired pancreatic B-cell response to hyperglycemia. The late insulin response to an oral glucose load suggests that some intestinal factor(s) might be responsible for the changes in insulin secretion.151 Unlike insulin, C peptide is not degraded by the liver. In cirrhotic patients, increases in both C peptide and insulin levels have been reported during fasting and often in response to an oral glucose load. However, the C peptide/insulin ratio, an index of hepatic insulin uptake, is lower in cirrhotic patients than in controls. This finding suggests that cirrhotic individuals have a reduced hepatic uptake and degradation of insulin.152,153 and 154 Clearly, the presence of hepatocyte dysfunction and both intrahepatic and extrahepatic portal systemic shunts in cirrhotic patients would be expected to contribute to the reduced hepatic insulin uptake and clearance of insulin. In healthy subjects, the insulin uptake by liver is ~50% of the insulin load. In patients with alcoholic liver cirrhosis, uptake has been estimated to be only 13%.155 However, another study evaluating insulin and glucagon catabolism by hepatic tissue obtained from cirrhotic patients and controls failed to show a significantly reduced degradation of these two hormones.156 Glucagon-like peptide-1 reverses the insulin secretion defects in liver cirrhosis.156a Impaired negative feedback inhibition by circulating insulin levels in cirrhotic patients is yet another possible explanation for the hyperinsulinemia and insulin hypersecretion seen in cirrhosis. In normal subjects, C-peptide secretion is suppressed by as much as 50% after 30 minutes of a euglycemic hyperinsulinemic infusion.157 In contrast, in cirrhotic patients insulin suppression is not achieved even after 120 minutes of a euglycemic hyperinsulinemic infusion. This finding suggests an insensitivity of B cells to suppression by insulin in cirrhotic patients.158 Non–insulin-dependent diabetes mellitus (type 2) is more common than insulin-dependent diabetes (type 1) in cirrhotic patients. Type 2 cirrhotic patients appear to have one or more insulin-receptor defects that contribute to their insulin resistance. This reduced sensitivity of cells to insulin may occur at either the insulin-receptor or postreceptor level. Moreover, an additional defect in glucose transport mechanisms may be present. Insulin binding to monocytes and adipocytes is reduced in cirrhotic subjects according to the severity of their glucose intolerance.159,160 Presently, the evidence that altered insulin-receptor binding contributes to the insulin resistance seen in cirrhosis is inconclusive. The data supporting the possibility of a postbinding defect are much more convincing.161,162 Examination of the transmembrane transport of labeled glucose in isolated adipocytes or skeletal muscle obtained from cirrhotic patients demonstrates a significant reduction in insulin-induced glucose uptake. Reduced levels of G6P are observed, whereas the activity of the phosphorylative chain is normal. Thus, insulin resistance in cirrhotic patients is characterized by reduced intracellular availability of G6P caused by a defect in glucose transport. Cirrhotic patients also may manifest a “glucose resistance,” that is, a reduced tissue uptake of glucose that occurs independent of insulin. The insulin resistance in cirrhotic patients can be explained in part by the presence of circulating factors that antagonize insulin effects peripherally. Among the putative insulin antagonists present in the plasma of cirrhotic patients are the counterregulatory hormones (i.e., GH, glucagon, cortisol, and catecholamines). The malnutrition common to advanced cirrhosis also alters glucose homeostasis by reducing the intracellular availability of various cofactors for enzymes involved in overall glucose metabolism. The hyperglucagonemia that occurs in cirrhotic subjects163,164,165 and 166 leads to a glucose overproduction, which contributes to the observed glucose intolerance. The data available for cirrhotic patients with impaired glucose tolerance fail to demonstrate a relationship between the degree of glucose intolerance and the level of hyperglucagonemia. Despite the hyperglucagonemia, hepatic glucose production is lower in cirrhotic patients than in normal individuals. This reflects both reduced hepatocyte function and reduced access of individual liver cells to glucose in portal venous blood. The possibility exists that, because of the portal systemic shunts within and around the liver in cirrhotic individuals, the diminished hepatocyte mass of cirrhotic patients actually experiences a reduced glucagon stimulus. No relationship between the elevated cortisol levels and glucose intolerance has been reported in cirrhotic patients.166 Thus, increased levels of adrenal counterregulatory hormones are unlikely to be related to their insulin resistance. High plasma FFA levels have also been proposed as possible insulin antagonists in cirrhotic individuals. More important, basal FFA levels correlate with the degree of glucose intolerance observed.166 In type 2 diabetes, elevated circulating FFA levels impair glucose utilization, especially in muscle tissue, thereby reducing the effect of insulin. A reduction in plasma FFA through the use of nicotinic acid is effective in ameliorating the glucose intolerance in individuals with type 2 diabetes without cirrhosis; however, it has little effect on the whole body insulin sensitivity seen in cirrhotic patients. A malabsorption of lipids frequently occurs in patients with liver disease, not only in those with cholestasis but also in those with parenchymal liver disease. Moreover, lipid synthesis and transfer rates are impaired in cirrhotic patients. They have both impaired FFA synthesis and impaired VLDL production. In cirrhosis, plasma triglycerides are elevated and are carried by LDL rather than VLDL. In advanced liver disease, however, both plasma triglycerides and VLDL levels are reduced. Of interest, diabetes-induced chronic vascular complications, such as diabetic retinopathy, are less frequent when diabetes is associated with chronic liver disease. Cirrhosis is a hypotensive condition, and the lower prevalence of hypertension in such patients may explain this observed reduction in vascular complications.167

DIABETES MELLITUS AND THE LIVER Diabetes mellitus comprises a group of heterogeneous disorders that are characterized by hyperglycemia and, in the more severe cases, ketosis and protein wasting.

Diabetes is associated with an increased risk of certain associated diseases, including those involving the liver: glycogenosis (50% of those with type 1 diabetes), fatty liver (50% of those with type 2 diabetes), steatonecrosis (10–20% of those with type 2 diabetes), and cirrhosis of the micronodular variety (16% of those with type 2 diabetes). LIPID METABOLISM IN DIABETES MELLITUS Hypercholesterolemia occurs in those with poorly controlled diabetes and improves with insulin therapy. Early studies demonstrated an increase in total body cholesterol synthesis in those with poorly controlled diabetes, which decreases with the initiation of better insulin therapy. In one study,168 however, total body cholesterol synthesis increased in some diabetic patients after initiation of insulin therapy. Animal data169,170 demonstrate a decreased hepatic cholesterol synthesis in uncontrolled diabetes. This is difficult to explain in the face of an increased total body cholesterol synthesis, because normally the liver contributes more than half of the total cholesterol synthesis.171 This may occur because intestinal cholesterol synthesis markedly increases in experimentally induced diabetes. Cholesterol formed in the intestine is carried by chylomicrons to the liver, where it contributes to the regulation of hepatic cholesterol synthesis by negative feedback.172 A specific sequence of events has been suggested in experimentally induced diabetes. The observed increase in total body cholesterol synthesis is predominantly due to an increased intestinal cholesterol synthesis, which is partially offset by a diminished hepatic cholesterol synthesis. The institution of insulin therapy reestablishes a more normal rate of intestinal cholesterol synthesis and decreases the amount of cholesterol delivered to the liver from the intestine. This removes the suppression of hepatic cholesterol formation, accounting for the observed increase in hepatic cholesterol synthesis that accompanies the initiation of insulin therapy. The liver is an important site of cholesterol synthesis and lipoprotein uptake and metabolism. Altered lipoprotein metabolism may contribute to the hypercholesterolemia observed in diabetic patients. Two defects in lipoprotein metabolism have been identified in diabetic individuals: an increased hepatic VLDL output and a decreased hepatic VLDL uptake. Because 20% of the VLDL particle is made up of cholesterol, the cholesterol concentration in plasma must increase as VLDL levels increase. The cholesterol carried in LDL, a particle that has a significant positive correlation with the development of atherosclerosis, is elevated in patients with type 1 diabetes mellitus. In humans, the LDL particle is formed almost exclusively from the hydrolysis of VLDL. LDL particles are cleared from the plasma by an LDL receptor–mediated mechanism. Quantitatively, however, the liver is thought to be the most important site of LDL cholesterol uptake (see Chap. 162). PROTEIN METABOLISM IN DIABETES MELLITUS The liver uses endogenous and peripherally supplied amino acids and is exposed to high levels of portal venous amino acids absorbed from the intestinal tract after feeding. Fifty percent of the amino acids metabolized by the liver is derived from hepatic protein degradation.173 The other half is derived from extrahepatic sources, either peripheral tissue sites or dietary sources. In the fed state, most amino acids delivered to the liver are derived from dietary proteins (90 g). Moreover, 50 g of amino acids are derived from exfoliated cell proteins that have been digested and absorbed by the intestine, 16 g are derived from secreted gastrointestinal enzymes, and 1 to 2 g are generated by exuded plasma proteins. In the fasting state, the total amount of amino acids that influx through the portal vein to the liver is reduced to approximately one-sixth of that found in the fed state and is derived almost exclusively from exfoliated intestinal cell protein. The liver adapts to this reduction in amino acids from the portal vein by metabolizing endogenous amino acids to supply its needs. Hepatic protein synthesis is under hormonal control. This is true for proteins retained within the hepatocyte and for those manufactured in the liver and secreted into the systemic circulation. Adequate levels of insulin and glucagon are also necessary for the intrinsic hepatic protein synthesis that is observed during hepatic regeneration after injury. Both hormones appear to stimulate hepatic DNA synthesis, protein synthesis, and subsequent cell division. Secretory protein synthesis by the liver is modulated by fasting and by protein ingestion. Albumin synthesis promptly decreases with fasting. This reduction in albumin synthesis is associated with a decreased hepatic RNA concentration, a decreased hepatic protein concentration, and a reduced state of aggregation of the endoplasmic reticulum–bound polysomes. The free polysomes, however, do not appear to be affected significantly by fasting. These changes are quickly reversed with feeding or after amino-acid delivery to the liver. Fasting affects the hepatic level of endoplasmic reticulum–bound polysomes, which are considered important in secretory protein synthesis, without affecting the hepatic levels of free polysomes, which are thought to be responsible for endogenous protein synthesis. A basal insulin level is necessary to maintain the state of aggregation of the endoplasmic reticulum–bound polysomes for secretory protein synthesis. In insulin-deficient animals, a loss of rough endoplasmic reticulum occurs, together with a proliferation of smooth endoplasmic reticulum, a reduced amino-acid incorporation into protein, and a decreased amount of rough endoplasmic reticulum–bound ribosomes existing as polyribosomes. The relative sparing of the smooth endoplasmic reticulum suggests that patients with diabetes mellitus predominantly have a defect in secretory protein synthesis, although they maintain their synthesis of essential intracellular proteins. BILIARY LIPID SECRETION IN DIABETES MELLITUS The influence of diabetes mellitus on biliary lipid secretion is seen most often in those with type 2 diabetes, who tend to be overweight and to secrete a lithogenic bile. When the duodenal bile obtained from patients with maturity-onset diabetes is compared with that of obesity-matched controls, however, the bile from both groups is similarly supersaturated.174,175 This suggests that obesity, rather than diabetes, is the etiologic factor responsible for gallstones, which are common in patients with obesity and type 2 diabetes. Interestingly, nonobese Pima Indians with type 2 diabetes also secrete a supersaturated bile. (This phenomenon tends to occur in this group even when diabetes is not present. Nonobese, non-Indian patients with significant fasting hyperglycemia, who do not develop ketoacidosis, have an increase in bile acid and hepatic cholesterol synthesis without an increase in bile acid pool size or cholesterol saturation.) Therefore, in patients with type 2 diabetes, the secretion of a lithogenic bile appears to be related more to obesity than to a genetic predisposition associated with the diabetes. The bile acid production and secretion rates in type 1 diabetes have not been investigated as carefully as those in type 2 diabetes. In a study176 of 12 patients with type 1 diabetes who were not overweight, no increase in biliary cholesterol saturation was found. Moreover, the total biliary lipid concentration was found to be lower in the diabetic group than in the control group. EFFECT OF INSULIN THERAPY Interestingly, insulin therapy changes the biliary saturation index. A lack of diabetic control is associated with increased bile salt synthesis and pool size, and insulin therapy decreases the bile salt secretion rate and reduces the bile acid pool size. This yields a net increase in the cholesterol saturation of bile. With experimentally induced diabetes, a decrease in hepatic cholesterol synthesis can be reversed with the institution of insulin therapy. In humans, cholesterol synthesis appears to be an important determinant of biliary cholesterol secretion. Biliary cholesterol secretion is higher in those with uncontrolled and insulin-treated diabetes than in controls. A decrease in the bile acid pool size, associated with an increased secretion of biliary cholesterol, causes the formation of a supersaturated bile, a necessary prerequisite for gallstone formation. Therefore, by decreasing bile salt secretion and increasing biliary cholesterol secretion, insulin therapy can actually increase the risk for gallstone formation. CHOLELITHIASIS AND DIABETES Many normal individuals secrete a supersaturated bile. Thus, the mere presence of a lithogenic bile does not completely explain the increased frequency of gallstones in diabetic individuals. The diabetic patient has another risk factor for gallstone formation not present in the normal population: decreased gallbladder contractility.177,178 This decreased contractility results in incomplete emptying of the gallbladder, which promotes cholesterol nucleation; the result is stone formation and growth, particularly if the bile is supersaturated. Data indicate that surgical intervention for gallstone disease in patients with diabetes and cholelithiasis is associated with a higher mortality rate than that seen in the normal population. In diabetic patients, emergency gallbladder surgery has been associated with a 10% to 20% mortality rate, compared with 1% to 4% in nondiabetic persons. These data prompted the suggestion that screening of patients with diabetes to identify silent gallstones should be considered, so that elective surgical intervention could be instituted before acute cholecystitis occurs. The available data,179 however, suggest that the increased surgical risk observed in diabetic patients is primarily related to the coexistence of vascular and renal disease. In the diabetic patient without confounding vascular or renal disease, surgical intervention for acute cholecystitis is not associated with a mortality greater than that observed for the general population. Currently, whether elective surgery for silent gallstones in the subpopulation of diabetic patients with vascular or renal disease actually decreases the mortality rate from cholelithiasis is unclear. LIVER DISEASES ASSOCIATED WITH DIABETES Various hepatic disorders are associated with diabetes mellitus. Moreover, several hepatic histopathologic lesions are observed with increased frequency in diabetic

populations. GLYCOGEN DEPOSITION The most common lesion seen in diabetes mellitus is an increase in liver glycogen, found both at autopsy and in biopsy material. The incidence rate of increased hepatic glycogen is as high as 80%. Increased glycogen deposits are observed especially in patients with brittle diabetes who are prone to hypoglycemia. Also, insulin therapy in these patients increases hepatic glycogen deposition further. Together, these findings suggest that an intermittent excess in insulin levels associated with exogenous insulin therapy causes the glycogen deposition and the associated hypoglycemia. The increased liver glycogen and the accompanying hepatomegaly can be quickly reversed if appropriate insulin therapy is instituted. FATTY LIVER Fatty liver is defined as a hepatic accumulation of lipid, usually in the form of triglyceride, that exceeds 5% of the liver weight. An increase in lipolysosomes occurs in diabetic patients with fatty liver and apparently correlates with the level of increased serum cholesterol. Hypercholesterolemia may be an important contributing factor to the pathogenesis of the fatty liver associated with diabetes. The prevalence of fatty liver in diabetic patients varies considerably, but biopsy studies reveal that ~50% of diabetic patients have excess fat in their livers. In several series of patients with fatty liver, diabetes was the cause in 4% to 46%. This wide range can be explained by the rate at which concurrent obesity is seen in those with type 2 diabetes. Because obesity appears to be a major cause of fatty liver, and because the incidence of fatty liver in patients with diabetes mellitus without concurrent obesity is unknown, isolating diabetes as the specific cause of the fatty liver in diabetic patients is extremely difficult. Nevertheless, the presence of a fatty liver in diabetic patients correlates directly with age, inversely with the severity of the carbohydrate abnormality, and directly with the duration of the diabetes. However, it does not correlate with the degree of diabetic control.180,181 Rarely do patients with type 1 diabetes develop fatty liver. In contrast, in the type 2 diabetic population, fatty liver is found in >50% of patients. In this population, insulin insensitivity, rather than the degree of glucose intolerance, predicts the occurrence of fatty liver. Although patients with fatty livers often present with hepatomegaly, this finding is not invariably present. In diabetes, no abnormality of liver enzymes reliably predicts excess hepatic fat. Histologically, the fat appears as both large and small droplets within the hepatic cytosol (Fig. 205-1). With intensive fatty infiltration, the cytoplasmic fat coalesces into large droplets that displace the nucleus eccentrically. Presumably, in severe cases this can cause hepatocyte injury and death. In areas of severe fatty infiltration, an increase in connective tissue is occasionally seen. Inflammation, as manifested by the presence of polymorphonuclear cells, is typically absent.182 Because of the poor correlation between the various laboratory tests for hepatic injury and function and the presence or absence of fat in liver cells, liver biopsy appears to be the only reliable way to make a reliable diagnosis of a fatty liver in diabetic and nondiabetic patients. Computed tomographic scanning has been somewhat helpful, whereas ultrasonographic techniques cannot distinguish fat from fibrosis.

FIGURE 205-1. Photomicrograph of a fatty liver in a diabetic patient showing ballooned fat-filled hepatocytes and a Mallory lesion (arrow).

The pathophysiologic mechanisms responsible for the development of a diabetic fatty liver are not understood completely. In diabetes mellitus, FFAs released from adipocytes are taken up by the liver in a concentration-dependent manner. The fate of these FFAs is either oxidation to ketone bodies, or esterification to phospholipids and triglycerides with subsequent excretion as part of VLDL particles. Fatty liver occurs when the rate of hepatic triglyceride synthesis exceeds the rate of hepatic secretion of VLDL. This can result from different mechanisms: (a) increased hepatic FFA concentrations derived from hepatic synthesis; (b) excess dietary intake or peripheral lipolysis; (c) decreased oxidation of fatty acids to ketone bod- ies; and (d) decreased output of triglycerides in VLDL particles. Which mechanism accounts for the development of fatty liver in any patient depends on the type of diabetes and the degree of diabetic control achieved. Fatty liver seen in type 1 diabetes occurs only when diabetic control is inadequate. In the presence of reduced serum insulin levels, a marked increase in hepatic FFA concentrations occurs that stimulate ketone body production and hepatic triglyceride synthesis. Glucagon also inhibits triglyceride secretion as VLDL, but does not inhibit triglyceride synthesis. These data suggest that the fatty liver seen in those with poorly controlled type 1 diabetes is due to an increased influx of FFAs and impaired VLDL secretion of triglycerides. In type 2 diabetes, the imbalance between hepatic triglyceride synthesis and VLDL secretion occurs for different reasons. Hepatic FFA concentrations are increased because of a greater intake of dietary fat and carbohydrate that contributes to elevated plasma FFA concentrations. Triglyceride synthesis is stimulated because of the increased hepatic FFA content from both endogenous and exogenous sources. Because no evidence exists to suggest that triglyceride secretion is impaired in type 2 diabetes, the intracellular lipid accumulation presumably occurs because the rate of triglyceride synthesis exceeds the liver's capacity to secrete newly formed triglyceride as VLDL. The clinical significance of fatty liver in patients with diabetes mellitus is debatable. Generally, fatty liver is not believed to progress to more severe disease. In pancreatectomized dogs, however, the progression of fatty liver to fibrosis and cirrhosis has been documented. Similarly, serial liver biopsies in diabetic patients with fatty liver have demonstrated progression to cirrhosis. Also, fatty steatosis, pericentral fibrosis, and intracellular hyaline occurred in the livers of a small series of women with poorly controlled diabetes.170 Steatonecrosis is seen more commonly in alcoholic individuals, in whom the presence of this histologic lesion is associated with a polymorphonuclear leukocytic inflammation not seen in diabetic patients. Despite the prevailing opinion that the fatty liver of diabetes does not lead to cirrhosis, many researchers report an increased prevalence of cirrhosis in diabetic patients. Approximately a four-fold increased incidence of cirrhosis occurs in diabetic patients. HEPATIC TOXICITY OF ORAL HYPOGLYCEMIC AGENTS The treatment of diabetes with oral hypoglycemic agents (see Chap. 142) also appears to increase the risk for liver disease. The most commonly used group of oral hypoglycemics are the sulfonylurea agents (e.g., chlorpropamide, tolazamide, tolbutamide, acetohexamide, glyburide, and glipizide). These agents produce jaundice, although the frequency of reported hepatotoxic reactions varies with the particular agent used. The reported incidence of toxicity is greatest with chlorpropamide and has been reported at 0.5% to 1%.183 Hepatic sulfonylurea toxicity usually consists of a cholestatic reaction, although hepatocellular injury or a mixed picture can be seen. Rarely, oral hypoglycemic agents are implicated as a cause of granulomatous liver disease.184 Hepatocellular necrosis appears to be most common with acetohexamide.

BONE AND CALCIUM METABOLISM Both osteoporosis and osteomalacia occur in patients with longstanding liver disease. Defects in GH-binding protein and IGF concentrations contribute to the pathogenesis of these disorders in cirrhotic individuals. Because of the hypoalbuminemia of many cirrhotic patients, the total calcium concentration often is low. Ionized calcium lev- els, however, are normal. Poor nutrition and alcohol-induced defects in the intestinal absorption of calcium are factors other than hypoalbuminemia that may contribute to the hypocalcemia. In cirrhotic patients, parathyroid hormone (PTH) levels usually are normal. Nonetheless, PTH metabolism is altered. In patients with primary biliary cirrhosis, increased levels of the C-terminal PTH peptide, a proteolytic product of intact PTH, are found.185 The biologic significance of increased levels of this 70-84 PTH product is

unclear. The liver also plays an important role in vitamin D metabolism. Most patients with liver disease have normal 1,25-hydroxyvitamin D3 (1,25-OHD3) and 25-hydroxyvitamin D3 (25-OHD3) levels, and only the few with very severe liver failure have reduced 25-OHD3 levels. The exception to this rule is primary biliary cirrhosis, in which the incidence of 25-OHD3 deficiency is high. Among the mechanisms that explain the finding of reduced serum 25-OHD3 concentrations in patients with cirrhosis is the reduced activity of 25-hydroxylase in the liver with an impaired conversion of vitamin D3 to 25-OHD3. Synthesis of vitamin D–binding protein by the liver is reduced.186,187 Other factors that also may contribute to vitamin D deficiency in end-stage cirrhosis include decreased sun exposure, poor nutrition, intestinal malabsorption of vitamin D, and a disturbed enterohepatic circulation of the vitamin. Osteoporosis is the most common bone disease seen in individuals with advanced liver disease (occurring in 21–39%).188,189 and 190 Long-term immobility, poor nutrition, and a reduced muscle mass are features commonly noted in individuals with cirrhosis. Each of these factors can augment the development of osteoporosis. In addition, in primary biliary cirrhosis the proliferation of osteoblasts may be inhibited by increased levels of conjugated bilirubin.191 Thus, prolonged hyperbilirubinemia may be an important factor in the pathogenesis of accelerated bone loss in patients with cholestatic liver disease. In cirrhotic patients, histomorphometric evidence suggests that bone formation is reduced (low turnover), whereas bone resorption rates are normal.186,188,192 A low osteocalcin level is a good marker of bone formation, and urinary pyridinium cross-link compounds reflect bone resorption rates. Both help to confirm the presence of a low bone turnover in the osteoporosis of cirrhotic patients.

ALCOHOLIC LIVER DISEASE EFFECT OF ALCOHOL ON THE HYPOTHALAMIC–PITUITARY–GONADAL AXIS ALCOHOLIC MEN Hypogonadism. Most alcohol-abusing male patients, particularly those with cirrhosis, present with signs of hypogonadism (e.g., loss of secondary male sex characteristics, testicular atrophy, and infertility)193,194,195,196,197,198 and 199 (Fig. 205-2).

FIGURE 205-2. Photomicrographs of the testes obtained from (A) an alcoholic man with testicular atrophy and (B) a healthy control. ×400

A reduction in circulating testosterone levels is the most common finding in alcoholic men with clinical manifestations of hypogonadism.193,199 The toxic effect of alcohol rather than the presence of alcohol-induced cirrhosis may be responsible for the androgen deficiency of male alcoholics.200,201,202,203,204 and 205 In fact, in experimental models, the reduction in circulating testosterone follows chronic alcohol feeding,204a even in the absence of liver injury. Several factors influence the impairment of testosterone biosynthesis caused by alcohol. Not only are ethanol and its metabolite, acetaldehyde, directly toxic to Leydig cells,205,206 but a disruption of the hypothalamic–pituitary–gonadal axis also occurs as a consequence of alcohol abuse. A central defect at the level of the pituitary and/or hypothalamus that is associated with chronic alcohol intake contributes to the hypogonadism of alcoholic men.193,207,208 Chronic alcohol exposure decreases circulating LH levels, and the response of LH to LH-releasing hormone (LHRH) is reduced in alcoholics.209,210 Hyperprolactinemia is often found in male alcoholics, particularly those with cirrhosis. Elevated PRL levels may participate in the pathogenesis of hypogonadism by inhibiting gonadotropin secretion, thereby further decreasing sex hormone synthesis as well as germ cell production and maturation.195,211 Although, in the absence of liver disease, alcohol enhances the hepatic metabolism of testosterone, in the presence of cirrhosis, ethanol increases testosterone binding to the plasma androgen carrier sex hormone–binding globulin; this reduces the metabolic clearance rate of the hormone.194 Feminization. Alcoholic men often manifest clinical gynecomastia, alterations in body fat distribution, and a female escutcheon. In addition, hyperestrogenization occurs in chronic alcoholic men.212 Although alcohol abuse can cause hypogonadism even in the absence of liver disease, cirrhosis, acting in concert with alcohol, produces feminization. Alcoholic men with chronic liver disease have elevated circulating estradiol and estrone levels.213 These are produced from weak androgens of adrenal gland origin.209 In alcoholic men, although normal or slightly increased concentrations of estradiol are commonly found, levels of estrone are increased by the aromatization of androstenedione.202 Both alcohol and acetaldehyde stimulate the adrenal cortex to increase the secretion of weak adrenal androgens that serve as estrogen precursors. Moreover, ethanol also increases the activity of aromatase, which converts androgens to estrogens. Normally, weak androgens are biotrans-formed by the liver and excreted in the urine; however, in liver disease, these hormones escape hepatic clearance and are shunted via venous collaterals to sites in the body where they undergo aromatization.203 The metabolic clearance rate for estradiol is normal in alcoholic subjects, even in the presence of advanced cirrhosis. On the other hand, in subjects with alcoholic cirrhosis, the reduced albumin may result in a relatively high level of free estrogen in the blood. Compounding this problem is the reduction in hepatic estrogen-binding proteins, which enables more estrogen receptors to produce an estrogenic effect. ALCOHOLIC WOMEN Hypogonadism. Alcohol abuse results in profound changes in the hormonal status, physical appearance, and reproductive performance of women. In women, chronic alcohol abuse may be associated with severe hypogonadism, as manifested by the loss of secondary sexual characteristics, amenorrhea, and early menopause, because the secretion of estrogens and gonadotropin is reduced. A reduction in estradiol and progesterone levels as well as ovulatory failure characterize the ovarian failure seen in alcoholic women. The ovulatory failure leads directly to infertility and to the absence of corpora lutea and the midcycle LH surge.214,215,216,217 and 218 Both increased PRL and decreased LH levels correlate with the severity of alcoholic liver disease as expressed by the Child-Pugh score. Such findings suggest that hormone levels may have a prognostic value in assessing the severity of liver disease. Endocrine Effect of Alcohol in Postmenopausal Women. Normal postmenopausal women who drink moderately (one drink per day or less) have higher estradiol levels than do women who do not drink. Whether alcohol or estrogenic phytoestrogens (principal nutrients of plant origin in alcoholic beverages) are responsible for this effect is unclear. (Phytoestrogens interact with estrogen receptor–binding proteins and experimentally produce an estrogenic response that is dose dependent.219) In cirrhotic alcoholic postmenopausal women, estradiol levels are higher and testosterone levels are lower than in controls. The conversion of androgens to estrogens is enhanced by alcohol abuse in alcoholic cirrhotic postmenopausal women as it is in men. As a consequence of the increased levels of estrogen, the production and secretion of pituitary gonadotropins is reduced. Despite the increase in estrogen levels in alcoholic cirrhotic women, they appear to be defeminized, with a marked loss

of secondary sex characteristics.220,221 EFFECT OF ALCOHOL ON THYROID FUNCTION Alcohol abuse has powerful effects on the hypothalamic–pituitary–thyroidal (HPT) axis and the thyroid gland. The severity of concomitant alcohol-induced liver disease is an important determinant of the severity of HPT axis disruption. Multiple thyroid function abnormalities may occur. Chronic alcoholism is associated with a marked decrease in thyroid volume222 and thyroid stromal fibrosis.223 However, no relationship between thyroid volume and fibrosis and any index of thyroid function has been reported. Acute alcohol administration reduces T3 levels both in alcoholic persons and in normal subjects by reducing hepatic deiodination of T4 to T3, due to an alcohol-induced hepatocellular injury.224,225 and 226 A reduction in the circulating levels of T4 and, to a greater extent, T3 occurs in alcoholic subjects with liver disease.227,228,229,230 and 231 Despite these alterations in the profiles of thyroid hormones, alcoholic patients remain clinically euthyroid.228,232 TSH levels are usually within the normal range.228,233,234,235 and 236 The most consistent finding in alcoholic individuals is blunting of the TSH response to TRH, suggesting that a defect in the hypothalamic-pituitary axis occurs as a consequence of alcohol abuse.237,238,239 and 240 In vitro, at physiologic doses, alcohol exposure modifies the number of TSH-binding sites on thyroid gland cell membranes as well as the thyroid response to TSH.241 This may represent a self-correcting effort involving the HPT axis. EFFECT OF ALCOHOL ON GROWTH HORMONE AND PROLACTIN In alcoholic persons, the increase in estrogen levels and the impaired synthetic ability of the liver due to the toxic effect of alcohol each contribute to a dysregulation of GH production and release (see Chap. 12). Chronic alcoholic subjects with cirrhosis have increased circulating levels of GH. A hypothalamic defect in neuroregulation of GH may exist in alcoholic patients with cirrhosis, as TRH abnormally stimulates the secretion of GH in alcoholic individuals with or without alcohol-induced liver disease, but not in normal volunteers.242,243 A single dose of ethanol does not affect PRL levels244; however, circulating PRL levels are increased in alcoholic individuals, particularly those with cirrhosis.245 An increased abundance of PRL-secreting cells are found in the pituitaries of alcoholic cirrhotic patients.246 PRL secretion is regulated primarily through inhibitory mechanisms (see Chap. 13). Dopamine is the most important PRL-inhibiting factor.247 In cirrhotic patients, dopaminergic drugs fail to suppress PRL and GH secretion, suggesting a dysregulation in dopaminergic systems, particularly in those patients with hepatic encephalopathy.248 This abnormal GH and PRL secretion most likely occurs because of portal hypertension and portosystemic shunting of adrenergic monoamines and other substances that affect dopaminergic neuroreactivity at the level of the hypothalamus.249,250 Hypersecretion of PRL has been implicated as one of the factors that mediate the ethanol-induced hypogonadism in male alcoholics. In experimental models, acute alcohol administration stimulates PRL secretion from the pituitary gland in a dose-dependent manner. Moreover, ethanol has a direct stimulatory effect on PRL secretion by the adenohypophysis.251 TRH stimulates the secretion of GH and PRL in individuals with alcohol-induced liver disease,252,253 and 254 an effect not seen in normal subjects. The hypothesis has been raised that a dysregulation of TRH secretion in cirrhotic patients also contributes to the abnormalities of GH and PRL secretion, which are noted particularly in alcoholic cirrhotic patients.236 EFFECT OF ALCOHOL ON BONE AND CALCIUM METABOLISM The effect of alcohol on bone metabolism is multifactorial. Normal osteoblast function depends on an adequate state of nutrition and an intact endocrine system. The effect of alcohol on bone metabolism is a consequence of both direct toxicity and an indirect action through an impairment of the individual's hormonal and nutritional status. Hypophosphatemia is yet another factor that contributes to the osteodystrophy seen in chronic alcoholics.255,256 Alcoholic patients often present with a notable reduction in bone mass,256a of trabecular bone in particular, associated with increased bone fragility compared with age-matched controls.257,258,259,260 and 261 Alcohol is thought to be responsible for the imbalance between bone formation and resorption. It decreases bone formation by reducing the number of osteoblasts.262,263 Circulating levels of osteocalcin, a marker of osteoblast function, are reduced in chronic alcoholism.264 Liver disease, a condition that often accompanies chronic alcohol abuse, may itself be associated with disrupted bone metabolism (hepatic osteodystrophy). Osteoporosis occurs predominantly in hypogonadic individuals; reduced sex hormone concentrations may lead to a decreased activity of osteoblasts.264 GH stimulates the proliferation of the osteoblasts and controls bone remodeling by modulating the production of IGF-I and IGF-II. Chronic alcohol abuse causes a reduction in both GH and IGF-I levels.265 Histomorphometric evidence of bone resorption is observed after moderate to heavy chronic alcohol intake. Osteoclasts are more abundant in bone biopsies obtained from alcoholic persons than in those from nonalcoholic individuals.266 Moreover, alcohol seems to directly stimulate osteoclast activity in vitro.267 In alcoholic individuals with cirrhosis, both malabsorption of calcium and vitamin D deficiency contribute to a chronic hypocalcemia. Ethanol ingestion interferes with intestinal calcium absorption by reducing the duodenal transport of the calcium ion.268,269 Moreover, alcohol ingestion is followed by an increased urinary excretion of both calcium and magnesium.270,271 The alterations in calcium homeostasis observed in alcoholic patients may be partly due to an alcohol-induced primary alteration of magnesium homeostasis, as magnesium is the principal regulator of PTH secretion and PTH action at peripheral tissue sites.272,273 and 274 Chronic alcoholic individuals have PTH levels that are either elevated or at the upper level of the normal range.

LIVER TRANSPLANTATION AND ENDOCRINE FUNCTION Orthotopic liver transplantation (OLTx) is the treatment of choice for end-stage liver disease. Although OLTx is effective in correcting liver function in a short period of time, a complete correction of the various metabolic and hormonal imbalances associated with advanced liver disease is usually delayed. The immunosuppressive therapy required after transplantation to prevent graft rejection contributes largely to this phenomenon. In the first few months after OLTx, the high doses of corticosteroids and of tacrolimus or cyclosporine that are required are responsible for maintaining insulin resistance, impaired insulin secretion, and a hyperlipidemia consisting of hypertriglyceridemia and hypercholesterolemia. Both plasma glucagon and plasma insulin levels are higher than normal for several months after successful OLTx, presumably as a consequence of an overproduction of both hormones by the pancreas. Within 2 years posttransplantation, insulin and glucose metabolism normalizes. Protein metabolism remains altered, however, as several amino-acid transport systems in the hepatocyte are normally under neural control, and the transplanted liver is denervated. This, combined with the use of corticosteroids and cyclosporine or tacrolimus, probably accounts for the abnormality in substrate handling noted in individuals with liver transplants. The cirrhosis-induced alteration of sex steroid homeostasis is significantly corrected after OLTx. Testosterone and gonadotropin levels usually return to normal within 6 months. A residual gonadal failure that may persist in patients with chronic alcoholism is associated with increased levels of FSH and LH. This residual hypergonadotropic hypogonadism is a consequence of an irreversible alcohol-induced gonadal injury. Finally, both cyclosporine and prednisone tend to reduce serum testosterone levels in men. Usually, women achieve normal menstrual function and fertility after successful OLTx. A return of menses can occur as soon as 2 months post-OLTx. Pregnancy has been reported to occur as early as 3 weeks after OLTx. Pregnant transplant recipients require the same doses of immunosuppressive agents as their nonpregnant peers. Successful pregnancy has been reported, with no adverse fetal consequences of the administration of standard doses of immunosuppressive agents throughout the pregnancy. Maternal and perinatal outcomes are generally favorable. Pregnancies in transplant recipients require careful monitoring, however, because of an increased risk of preterm delivery, of preeclampsia, and of perineal infection in women who have undergone OLTx before their pregnancies. CHAPTER REFERENCES 1. Ranke MB, Stanley CA, Tenore A, et al. Characterization of somatogenic and lactogenic binding sites in isolated hepatocytes. Endocrinology 1976; 99:1033. 2. Postel-Vinay M-C, Cohen-Tanugi E, Charrier J. Growth hormone receptors in rat liver membranes: effects of fasting and refeeding, and correlation with plasma somatomedin activity. Mol Cell

3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97.

Endocrinol 1982; 28:657. Delahaye-Zerras MC, Mertami H, Martini JF, et al. Expression of the growth hormone receptor gene in human digestive tissue. J Clin Endocrinol Metab 1994; 78:1473. Bala RM, Bohnet HG, Carter JN, Friesen HG. Effect of ovine prolactin on serum somatomedin bioactivity in hypophysectomized female rats. Can J Physiol Pharmacol 1978; 56:984. Francis MJO, Hill DJ. Prolactin-stimulated production of somatomedin by rat liver. Nature 1975; 255:167. Holder AT, Wallis M. Actions of growth hormone, prolactin and thyroxine on serum somatomedin-like activity and growth in hypopituitary dwarf mice. J Endocrinol 1977; 74:223. Cheek DB, Graystone JE. The action of insulin, growth hormone and epinephrine on cell growth in the liver, muscle, and brain of the hypophysectomized rat. Pediatr Res 1969; 3:77. Wong BS, Chenoweth ME, Dunn A. Possible growth hormone control of liver glutamine synthetase activity in rats. Endocrinology 1980; 106:268. Gebhardt R, Mecke D. The role of growth hormone, dexamethasone and triiodothyronine in the regulation of glutamine synthetase in primary cultures of rat hepatocytes. Eur J Biochem 1979; 100:519. Korner A, Hogan BLM. The effect of growth hormone on inducible liver enzymes. In: Pecile A, Muller EE, eds. Growth and growth hormone. Amsterdam: Excerpta Medica, 1971:98. Paleckar AG, Collipp PJ, Maddaiah VI. Growth hormone and rat liver mitochondria: effects on urea cycle enzymes. Biochem Biophys Res Commun 1981; 100:1604. Raina A, Holtta E. The effect of growth hormone on the synthesis and accumulation of polyamines in mammalian tissues. In: Pecile A, Muller EE, eds. Growth and growth hormone. Amsterdam: Excerpta Medica 1972:143. Gustafsson JA, Mode A, Norstedt G, et al. The hypothalamo-pituitary-liver axis: a new hormonal system in control of hepatic steroid and drug metabolism. In: Litwack G, ed. Biochemical actions of hormones, vol 7. New York: Academic Press, 1980:47. D'Ercole AJ, Stiles AD, Underwood LE. Tissue concentrations of somatomedin C: further evidence for multiple sites of synthesis and paracrine or autocrine mechanisms of action. Proc Natl Acad Sci U S A 1984; 81:935. Vassilopoulou-Sellin R, Phillips LE. Extraction of somatomedin activity from rat liver. Endocrinology 1982; 110:582. Mayer PW, Schalch DS. Somatomedin synthesis by a subclone of Buffalo rat liver cells: characterization and evidence for immediate secretion of de novo synthesized hormone. Endocrinology 1983; 113:588. Merimee TJ, Zapf J, Froesch ER. Insulin-like growth factors in the fed and fasted states. J Clin Endocrinol Metab 1982; 55:999. Ilan Y, Oren R, Tur-Kaspa R. Elevated growth hormone levels in patients with non-alcoholic chronic liver disease. J Gastroenterol Hepatol 1993; 8:448. Moller S, Gronbaek M, Main K, et al. Urinary growth hormone (U-GH) excretion and serum insulin-like growth factor 1 (IGF-1) in patients with alcoholic cirrhosis. J Hepatol 1993; 17:315. Wu A, Grant DB, Hambley J, Levi AJ. Reduced serum somatomedin activity in patients with chronic liver disease. Clin Sci Mol Med 1974; 47:359. Hepner GW, Chopra IJ. Serum thyroid hormone levels in patients with liver disease. Arch Intern Med 1979; 139:1117. Vannotti A, Beraud T. Functional relationships between the liver, the thyroxine binding protein of serum, and the thyroid. J Clin Endocrinol Metab 1959; 19:466. Gardner DF, Carithers RL, Utiger RD. Thyroid function tests in patients with acute and resolved hepatitis B virus infection. Ann Intern Med 1982; 96:450. Schlusser GC, Schaffner F, Korn F. Increased serum thyroid hormone binding and decreased free hormone in chronic active liver disease. N Engl J Med 1978; 299:510. Ross DS, Daniels GH, Dienstag JL, Ridgway EC. Elevated thyroxine levels due to increased thyroxine binding globulin in acute hepatitis. Am J Med 1983; 74:564. Hepner GW, Chopra IJ. Serum thyroid hormone levels in patients with liver disease. Arch Intern Med 1979; 139:1117. Pagliacci MC, Pelicci G, Francisci D, et al. Thyroid function tests in acute viral hepatitis: relative reduction in serum thyroxine levels due to T 4-TBG binding inhibitors in patients with severe liver cell necrosis. J Endocrinol Invest 1989; 12:149. Van Thiel DH, Stone BG, Schade RR. The liver and its effect on endocrine function in health and diseases. In: Schiff L, Schiff ER, eds. Diseases of the liver. Philadelphia: JB Lippincott, 1987:129. Kano T, Kojima T, Takahashi T, Muto Y. Serum thyroid hormone levels in patients with fulminant hepatitis: usefulness of rT3 and the rT3/T3 ratio as prognostic indices. Gastroenterol Jpn 1987; 22(3):344. Itoh S, Yamaha Y, Oda T, Kawagoe K. Serum thyroid hormone, triiodothyronine, thyroxine, and triiodothyronine/thyroxine ratio in patients with fulminant, acute, and chronic hepatitis. Am J Gastroenterol 1986; 81:444. Faber J, Kirkegaard C, Rasmussen B, et al. Pituitary thyroid axis in critical illness. J Clin Endocrinol Metab 1987; 65:315. Schimmel M, Utiger RD. Thyroidal and peripheral production of thyroid hormones: review of recent findings and their clinical implications. Ann Intern Med 1977; 878:760. Chopra IJ, Hershman JM, Pardridge WM, Nicoloff JT. Thyroid function in nonthyroidal illness. Ann Intern Med 1983; 98:946. Wartofsky L, Burman KD. Alterations in thyroid function in patients with systemic illness: the “euthyroid sick syndrome.” Endocr Rev 1982; 3:164. Kaptein EM, Grieb DA, Spencer CA, et al. Thyroxine metabolism in the low thyroxine state of critical nonthyroidal illnesses. J Clin Endocrinol Metab 1981; 53:764. Baumgarten A, Rommelspacher H, Otto M, et al. Hypothalamic-pituitary-thyroid (HPT) axis in chronic alcoholism. I. HPT axis in chronic alcoholics during withdrawal and after three weeks of abstinence. Alcoholism 1994; 18:284. Schussler GC, Schaffner F, Hurley J, Shapiro J. Thyroid function in primary biliary cirrhosis. Clin Res 1979; 27:259A. Mendel CM, Cavalieri RR, Weisiger RA. Uptake of thyroxine by the perfused rat liver: implications for the free hormone hypothesis. Am J Physiol 1988; 255:E110. Elta GH, Sepersky RA, Goldberg MJ, et al. Increased incidence of hypothyroidism in primary biliary cirrhosis. Dig Dis Sci 1982; 28:971. Chopra IJ, Solomon DH, Chopra U, et al. Alterations in circulating thyroid hormones and thyrotropin in hepatic cirrhosis: evidence for euthyroidism despite subnormal serum triiodothyronine. J Clin Endocrinol Metab 1974; 39:501. Itoh S, Matsuo S, Oda T, et al. Triiodothyronine level and triiodothyronine/thyroxine ratio in HBeAg-positive chronic hepatitis patients treated with prednisolone-withdrawal. Dig Dis Sci 1990; 35:1110. Green JRB, Snitcher ES, Mowat NAG, et al. Thyroid function and thyroid regulation in euthyroid men with chronic liver disease. Clin Endocrinol (Oxf) 1977; 7:453. Nomura S, Pittman CS, Chambers JB, et al. Reduced peripheral conversion of thyroxine to triiodothyronine in patients with hepatic cirrhosis. J Clin Invest 1975; 56:643. Inada M, Sterling M. Thyroxine turnover and transport in Laënnec's cirrhosis of the liver. J Clin Invest 1967; 46:1275. Van Thiel DH, Gavaler JS, Tarter RE, et al. Pituitary and thyroid hormone levels before and after orthotopic hepatic transplantation and their responses to thyrotropin-releasing hormone. J Clin Endocrinol Metab 1985; 60:569. Israel Y, Walfish PB, Orrego H, et al. Thyroid hormones in alcoholic liver disease: effect of treatment with 6-N-propylthiouracil. Gastroenterology 1979; 76:116. Walfish PG, Orrego H, Israel Y, et al. Serum triiodothyronine and other clinical and laboratory indices of alcoholic liver disease. Ann Intern Med 1979; 91:13. Orrego H, Kalant H, Israel Y, et al. Effect of short-term therapy with propylthiouracil in patients with alcoholic liver disease. Gastroenterology 1979; 76:105. Chopra IJ, Hung TS, Beredo A, et al. Serum thyroid hormone binding inhibitor in non-thyroidal illnesses. Metabolism 1986; 35:152. Suzuki Y, Nanno M, Gemma R, Yoshimi T. Plasma free fatty acids: inhibitor of and thyroid hormone binding inhibitor Woeber K, Maddux B. Thyroid hormone binding in nonthyroid illness. Metabolism 1981; 30:412. Van Thiel DH. The liver and the endocrine system. In: Arias IM, Jakoby WB, Popper H, et al., eds. The liver: biology and pathobiology. New York: Raven Press, 1988:1007. Van Thiel DH, Tarter R, Gavaler JS, et al. Thyroid and pituitary hormone responses to TRH in advanced nonalcoholic liver disease. J Endocrinol Invest 1986; 9:479. Van Thiel DH, Gavaler JS, Tarter R, et al. Pituitary and thyroid hormone levels before and after orthotopic hepatic transplantation and their responses to thyrotropin-releasing hormone. J Clin Endocrinol Metab 1985; 60:569. Van Thiel DH, Udani M, Schade RR, et al. Prognostic value of thyroid hormone levels in patients evaluated for liver transplantation. Hepatology 1985; 5:862. Wilber JF, Utiger RD. The effect of glucocorticoids on thyrotropin secretion. J Clin Invest 1969; 48:2096. Borst GC, Osburne RC, O'Brian JT, et al. Fasting decreases thyrotropin responsiveness to thyrotropin-releasing hormone: a potential cause of misinterpretation of thyroid function tests in the critically ill. J Clin Endocrinol Metab 1983; 57:380. Sap J, Munoz A, Damm K, et al. The c-erb-A protein is a high-affinity receptor for thyroid hormone. Nature 1986; 324:635. Williams GR, Franklyn JA, Neuberger JM, Sheppard MC. Thyroid hormone receptor expression in the “sick euthyroid” syndrome. Lancet 1989; 2:1477. Silva JE, Larsen PR. Contributions of plasma triiodothyronine and local thyroxine monodeiodination to triiodothyronine to nuclear triiodothyronine receptor saturation in pituitary, liver and kidney of hypothyroid rats. J Clin Invest 1978; 61:1247. Chopra IJ, Williams DE, Orgiazzi J, Solomon DH. Opposite effects of dexamethasone on serum concentrations of 3,3,5-triiodothyronine (T 3). J Clin Endocrinol Metab 1975; 41:911. Cooper DS, Daniels GH, Ladenson PW, Ridgeway EC. Hyperthyroxinemia in patients treated with high-dose propranolol. Am J Med 1982; 73:867. Felicetta JV, Green WL, Melp WB. Inhibition of hepatic binding of thyroxine by cholecystographic agents. J Clin Invest 1980; 65:1032. Martino E, Safran M, Aghini-Lombardi F, et al. Environmental iodine intake and thyroid dysfunction during chronic amiodarone therapy. Ann Intern Med 1984; 101:28. Amico JA, Richardson V, Alpert B, Klein I. Clinical and chemical assessment of thyroid function during therapy with amiodarone. Arch Intern Med 1984; 144:487. Taniguchi Y, Murakami T, Nakanishi K, et al. Two cases of hypothyroidism associated with alpha-interferon therapy. Intern Med 1992; 31:373. Fattovich G, Betterle C, Brollo L, et al. Autoantibodies during alpha-interferon therapy for chronic hepatitis B. J Med Virol 1991; 34:132. Barreca T, Picciotto A, Franceschini R, et al. Long-term therapy with recombinant interferon alpha 2b in patients with chronic hepatitis C: effects on thyroid function and autoantibodies. J Biol Regul Homeost Agents 1993; 7:58. Thompson P, Stru O. Abnormalities of liver function tests in thyrotoxicosis. Mil Med 1978; 143:548. Goglia F, Liverini G, Lanni A, Barletta A. Mitochondrial DNA, RNA and protein synthesis in normal, hypothyroid and mildly hyperthyroid rat liver during cold exposure. Mol Cell Endocrinol 1988; 55:141. Pauletzki J, Stellare F, Paumgartner GH. Bile acid metabolism in human hyperthyroidism. Hepatology 1989; 9:852. Demelia L, Solinas A, Poma R, et al. Hypothalamo-pituitary-adrenal function in liver cirrhosis of viral etiology. Ann Ital Med Int 1991; 6:203. Maruyama Y, Adachi Y, Aoki N, et al. Mechanism of feminization in male patients with non-alcoholic liver cirrhosis: role of sex hormone-binding globulin. Gastroenterol Jpn 1991; 26:435. Farinati F, De Maria N, Fornasiero A, et al. Unresectable hepatocellular carcinoma: prospective controlled trial with the anti-estrogen drug tamoxifen in patients with unresectable hepatocellular carcinoma. Dig Dis Sci 1992; 37:659. Frezza EE, Gerunda GE, Farinati F, et al. Sex hormones and trace elements in rat CCL4-induced cirrhosis and hepatocellular carcinoma. Eur J Cancer Prev 1993; 2:357. Walton C, Kelly WF, Laing I, Bulock OE. Endocrine abnormalities in idiopathic haemochromatosis. Q J Med 1983; 52:99. McNeil LW, McKee LC Jr, Lorber D, Robin D. The endocrine manifestations of hemochromatosis. Am J Med Sci 1983; 285:7. Charbonnel B, Chupin M, LeGrand A, Guillon J. Pituitary function in idiopathic haemochromatosis. Acta Endocrinol (Copenh) 1981; 98:178. Bezwoda WR, Bothwell TH, Vanderwalt LA, et al. An investigation into gonadal dysfunction in patients with idiopathic haemochromatosis. Clin Endocrinol (Oxf) 1977; 6:377. Burrows GH, Barrea A. Copper stimulates the release of luteinizing hormone releasing hormone from isolated hypothalamic granulae. Endocrinology 1982; 115:1456. Barrea A, Cho G. Evidence that copper-amino acid complexes are potent stimulators of the release of luteinizing hormone releasing hormone from isolated hypothalamic granulae. Endocrinology 1984; 115:936. Piperno A, Rivolta MR, D'Alba R, et al. Preclinical hypogonadism in genetic hemochromatosis in the early stage of the disease: evidence of hypothalamic dysfunction. J Endocrinol Invest 1992; 15:423. Eagon PK, Fisher SE, Imhoff AF, et al. Estrogen binding proteins in male rat livers: influences of hormonal changes. Arch Biochem Biophys 1980; 201:486. Eagon PK, Zdunek JR, Van Thiel DH, et al. Alcohol-induced changes in hepatic estrogen binding proteins. Arch Biochem Biophys 1981; 211:48. Aten RF, Dickson RB, Eisenbeld AJ. Estrogen receptor in adult male rat liver. Endocrinology 1978; 103:1629. Powell-Jones W, Thompson C, Nayfeh SN, Lucier GW. Sex differences in estrogen binding by cytosolic and nuclear components of rat liver. J Steroid Biochem 1980; 13:219. Norstedt G, Wrange O, Gustafsson JA. Multihormonal regulation of the estrogen receptor in rat liver. Endocrinology 1981; 108:1190. Porter LE, Elm M, Van Thiel DH, Eagon PK. Hepatic estrogen receptor in human liver disease. Gastroenterology 1987; 92:735. Powell-Jones W, Raeford S, Lucier GW. Binding properties of zearalenone myocotoxins to hepatic estrogen receptors. Mol Pharmacol 1981; 20:35. Kneifel R, Katzenellenbogen BS. Comparative effects of estrogen and anti-estrogen on plasma renin substrate levels and hepatic estrogen receptors in rats. Endocrinology 1981; 108:545. Kupfer D, Bulger WH. Estrogenic properties of DDT and its analogs. In: McLaughlin JA, ed. Estrogens in the environment. New York: Elsevier/North Holland, 1980:239. Dickson RB, Eisenfeld AJ. 17-ethinylestradiol is more potent than estradiol in receptor interactions with isolated hepatic parenchymal cells. Endocrinology 1981; 108:1551. Porter LE, Elm MS, Van Thiel DH, et al. Characterization and quantitation of human hepatic estrogen receptor. Gastroenterology 1983; 84:704. Porter LE, Elm MS, Van Thiel DH, Eagon PK. Estrogen receptor in human liver nuclei. Hepatology 1984; 4:1085. Colby HD. Regulation of hepatic drug and steroid metabolism by androgens and estrogens. In: Thomas JA, Singal RL, eds. Advances in sex hormone research. Baltimore: Urban and Schwartzenberg, 1980:346. Bardin CW, Catteral JS. Testosterone: a major determinant of extragenital sexual dimorphism. Science 1981; 211:1285. Gustafsson JA, Mode A, Norstedt G, et al. The hypothalamo-pituitary-liver axis: a new hormonal system in control of hepatic steroid and drug metabolism. Biochem Action Horm 1980; 7:47.

98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156.

Gustafsson JA, Mode A, Norstedt G, et al. Sex steroid induced changes in hepatic enzymes. Annu Rev Physiol 1983; 45:51. Roy AK, Chatterjee B. Sexual dimorphism in the liver. Annu Rev Physiol 1983; 45:37. Eden S. Age- and sex-related differences in episodic growth hormone secretion in the rat. Endocrinology 1979; 105:555. Mode A, Norstedt G, Simic B, et al. Continuous infusion of growth hormone feminizes hepatic steroid metabolism in the rat. Endocrinology 1981; 108:2103. Mode A, Gustafsson JA, Jansson JD, et al. Association between plasma levels of growth hormone and sex differentiation of hepatic steroid metabolism in the rat. Endocrinology 1982; 111:1692. Norstedt G, Palmiter R. Secretory rhythm of growth hormone regulates sexual differentiation of mouse liver. Cell 1984; 36:805. Van Thiel DH, Gavaler JS. Sex steroids and the liver. In: Francavilla A, Panella C, DiLeo A, Van Thiel DH, eds. Liver and hormones. New York: Raven Press, 1987; 43:183. Gilnoer D, McGuire RA, Gershengorn MC, et al. Effects of estrogen on thyrotropin-binding globulin metabolism in rhesus monkeys. Endocrinology 1977; 100:9. Gilnoer DM, Gershengorn MC, DuBois A, Robbins J. Stimulation of thyroxine binding globulin synthesis by isolated rhesus monkey hepatocytes after in vitro b-estradiol administration. Endocrinology 1977; 100:807. Moore DE, Kawagoe S, Davajan V, et al. An in vivo system in man for quantitation of estrogenicity: II. Pharmacologic in binding capacity of serum corticosteroid-binding globulin induced by conjugated estrogens, menstranol, and ethinyl estradiol. Am J Obstet Gynecol 1978; 130:482. Song CS, Kappas A. Hormones and hepatic function. In: Schiff L, ed. Diseases of the liver, 4th ed. Philadelphia: JB Lippincott, 1975:163. Windler EET, Kovaner PT, Chao Y-S, et al. The estradiol-stimulated lipoprotein receptor of rat liver: a binding site that mediates the uptake of rat lipoproteins containing apoproteins B and E. J Biol Chem 1980; 255:10. Porter LE, Van Thiel DH, Eagon PK. Estrogens and progestins as tumor inducers. Semin Liver Dis 1987; 7:24. Francavilla A, Gavaler JS, Makowka L, et al. Estradiol and testosterone levels in patients undergoing partial hepatectomy: a possible signal for hepatic regeneration? Dig Dis Sci 1989; 34:818. Francavilla A, Eagon PK, Di Leo A, et al. Sex hormone-related functions in regenerating male rat liver. Gastroenterology 1986; 91:1263. Francavilla A, Polimeno L, Di Leo A, et al. The effect of estrogen and tamoxifen in hepatocyte proliferation in vivo and in vitro. Hepatology 1989; 9:614. Francavilla A, Di Leo A, Eagon PK, et al. Regenerating rat liver: correlation between estrogen receptor localization and deoxyribonucleic acid synthesis. Gastroenterology 1984; 86:552. Fisher B, Gunduz M, Saffer EA, Zheng S. Relation of estrogen and its receptors to rat liver growth and regeneration. Cancer Res 1984; 44:2410. Van Thiel DH, Gavaler JS. Pregnancy-associated sex steroids and their effects on the liver. Semin Liver Dis 1987; 7:1. Eagon PK, Francavilla A, Di Leo A, et al. Quantitation of estrogen and androgen receptors in hepatocellular carcinoma and adjacent normal human liver. Dig Dis Sci 1991; 36:1303. Eagon PK, Willett SM, Seguitti ML, et al. Androgen-responsive functions of male rat liver. Gastroenterology 1987; 93:1162. Van Thiel DH, Gavaler JS, Spero JA, et al. Patterns of hypothalamic-pituitary-gonadal dysfunction in men with liver disease due to differing etiologies. Hepatology 1981; 1:39. Van Thiel DH, Gavaler JS, Rosenblum E, Eagon PK. Effects of ethanol on endocrine cells: testicular effects. Ann N Y Acad Sci 1987; 492:287. Van Thiel DH, Gavaler JS. Hypothalamic-pituitary-gonadal function in liver disease with particular attention to the endocrine effects of chronic alcohol abuse. Prog Liver Dis 1986; 8:273. Bannister P, Oakes J, Sheridan P, Losowsky MS. Sex hormone changes in chronic liver disease: a matched study of alcoholic versus non-alcoholic liver disease. Q J Med 1987; 63:305. Rose LI, Underwood RH, Newmark SR, et al. Pathophysiology of spirono-lactone-induced gynecomastia. Ann Intern Med 1977; 87:398. Van Thiel DH, Gavaler JS, Schade RR. Liver disease and the hypothalamic pituitary gonadal axis. Semin Liver Dis 1985; 5:35. Cundy TF, Butler J, Pope RM, et al. Amenorrhoea in women with non-alcoholic chronic liver disease. Gut 1991; 32:202. Weinhouse S. Regulation of glucokinase in liver. In: Horecker BL, Stadtman ER, eds. Current topics in cellular regulation. New York: Academic Press, 1976:1. Felig P, Wahren J. Influence of endogenous insulin secretion on splanchnic glucose and amino acid metabolism in man. J Clin Invest 1971; 50:1702. Cherrington AD, Steiner KE. The effects of insulin on carbohydrate metabolism in vivo. J Clin Endocrinol Metab 1982; 11:307. Chupin M, Charbonnel B, Le Bodic L, et al. Glucose tolerance in viral hepatitis. A study of twenty patients during the acute phase and after recovery. Diabetes 1978; 27:661. Kaneko K, Arai M, Funatomi H, et al. Changes in immunoreactive insulin, C-peptide immunoreactivity, and immunoreactive glucagon in acute viral hepatitis. J Gastroenterol 1995; 30:624. Sullivan SN, Chase RA, Christofides ND, et al. The gut hormone profile of fulminant hepatic failure. Am J Gastroenterol 1981; 76:338. Samson RI, Trey C, Timme AH, Saunders SJ. Fulminating hepatitis with recurrent hypoglycemia and hemorrhage. Gastroenterology 1967; 53:291. Bucher ML, Swaffield MN. Regulation of hepatic regeneration in rats by synergistic action of insulin and glucagon. Proc Natl Acad Sci U S A 1975; 72:1157. Arky RA. Hypoglycemia associated with liver disease and ethanol. Endocrinol Metab Clin North Am 1989; 18:75. Van Thiel DH. The liver and the endocrine system. In: Arias IM, Jakoby WB, Popper H, et al., eds. The liver: biology and pathobiology. New York: Raven Press, 1988:1007. Tauton OD, Greene HC. Fructose-1,6-diphosphatase deficiency, hypoglycemia and a response to folate therapy in a mother and a daughter. Biochem Med 1978; 19:260. Hommes FA, Bendienik E. Two cases of phosphoenolpyruvate carboxykinase deficiency. Acta Pediatr Scand 1976; 65:233. Freinkel N, Cohn AK. Alcoholic hypoglycemia: a prototype of the hypoglycemias induced by fasting in diabetes. In: Ostram J, ed. New York: Elsevier, 1969:873. Felig P, Brown WV, Levine RA, Klatskin G. Glucose homeostasis in viral liver disease. N Engl J Med 1979; 283:1436. Jaspan JB, Huen AH, Morley CG, et al. The role of the liver in glucagon metabolism. J Clin Invest 1977; 60:421. Shroyer LA, Varandani PT. Purification and characterization of rat liver cytosol neutral thiol peptidase that degrades glucagon, insulin and isolated insulin A and B chains. Arch Biochem Biophys 1985; 236:205. Collins JR, Crofford OB. Glucose intolerance and insulin resistance in patients with liver disease. Arch Intern Med 1969; 124:142. Kingston ME, Ashraf AM, Atiyeh M, Donnelly RJ. Diabetes mellitus in chronic active hepatitis and cirrhosis. Gastroenterology 1984; 87:688. Conn HO, Schreiber W, Elkington SG. Cirrhosis and diabetes: association of impaired glucose intolerance with portal-systemic shunting in Laënnec's cirrhosis. Dig Dis Sci 1971; 16:227. Johnston DG, Alberti KGMM, Faber OK, Binder C. Hyperinsulinism of hepatic cirrhosis: diminished degradation or hypersecretion. Lancet 1977; I:10. Romijn JA, Endert E, Sauerwein HP. Glucose and fat metabolism during short-term starvation in cirrhosis. Gastroenterology 1991; 100:731. Collins JR, Crofford OB. Glucose intolerance and insulin resistance in patients with liver disease. Arch Intern Med 1969; 124:142. Kingston ME, Ali MA, Atiyeh M, Donnelly RJ. Diabetes mellitus in chronic active hepatitis and cirrhosis. Gastroenterology 1984; 87:688. Conn HO, Schreiber W, Elkington SG. Cirrhosis and diabetes. II. Association of impaired glucose tolerance with portal-systemic shunting in Laënnec's cirrhosis. Am J Dig Dis 1971; 16:227. Johnson DG, Alberti KG, Faber OK, Binder C. Hyperinsulinism of hepatic cirrhosis: diminished degradation or hypersecretion? Lancet 1977; 1:10. Kruszynska YT, Home PD, McIntyre N. Relationship between insulin sensitivity, insulin secretion and glucose tolerance in cirrhosis. Hepatology 1991; 14:103. Kasperska-Czyzykowa T, Heding LG, Czyzyk A. Serum levels of true insulin, C-peptide and proinsulin in peripheral blood of patients with cirrhosis. Diabetologia 1983; 25:506. Ballmann M, Hartmann H, Deacon CF, et al. Hypersecretion of proinsulin does not explain the hyperinsulinaemia of patients with liver cirrhosis. Clin Endocrinol (Oxf) 1986; 25:351. Proietto J, Dudley FJ, Aitken P, Alford FP. Hyperinsulinaemia and insulin resistance of cirrhosis: the importance of insulin hypersecretion. Clin Endocrinol (Oxf) 1984; 21:657. Nygren A, Adner N, Sundblad L, Wiechel KL. Insulin uptake by the human alcoholic cirrhotic liver. Metabolism 1985; 34:48. Antoniello S, La Rocca S, Cavalcanti E, et al. Insulin and glucagon degradation in liver are not affected by hepatic cirrhosis. Clin Chim Acta 1989; 183:343.

156a. Siegel EG, Seidenstüker A, Gallwitz B, et al. Insulin secretion defects in liver cirrhosis can be reversed by glucagon-like peptide-1. J Endocrinol 2000; 164:13. 157. 158. 159. 160. 161. 162. 163. 164. 165. 166. 167. 168. 169. 170. 171. 172. 173. 174. 175. 176. 177. 178. 179. 180. 181. 182. 183. 184. 185. 186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202.

Cavallo-Perin P, Bruno A, Nuccio P, et al. Feedback inhibition of insulin secretion is altered in cirrhosis. J Clin Endocrinol Metab 1986; 63:1023. Petrides AS, Schulze-Berge D, Vogt C, et al. Glucose resistance contributes to diabetes mellitus in cirrhosis. Hepatology 1993; 18:284. Taylor R, Heine RJ, Collins J, et al. Insulin action in cirrhosis. Hepatology 1985; 5:64. Cavallo-Perin P, Cassader M, Bozzo C, et al. Mechanism of insulin resistance in human liver cirrhosis. Evidence of a combined receptor and postreceptor defect. J Clin Invest 1985; 75:1659. Blei AT, Robbins DC, Drobny E, et al. Insulin resistance and insulin receptors in hepatic cirrhosis. Gastroenterology 1982; 83:1191. Harewood MS, Proietto J, Dudley F, Alford FP. Insulin action and cirrhosis: insulin binding and lipogenesis in isolated adipocytes. Metabolism 1982; 31:1241. Greco AV, Crucitti F, Ghirlanda G, et al. Insulin and glucagon concentrations in portal and peripheral veins in patients with hepatic cirrhosis. Diabetologia 1979; 17:23. Petrides AS, Vogt C, Schulze-Berge D, et al. Pathogenesis of glucose intolerance and diabetes mellitus in cirrhosis. Hepatology 1994; 19:616. Sherwin RS, Fisher M, Bessoff J, et al. Hyperglucagonemia in cirrhosis: altered secretion and sensitivity to glucagon. Gastroenterology 1978; 74:1224. Riggio O, Merli M, Cangiano C, et al. Glucose intolerance in liver cirrhosis. Metabolism 1982; 31:627. Vidal J, Ferrer JP, Esmatjes E, et al. Diabetes mellitus in patients with liver cirrhosis. Diabetes Res Clin Pract 1994; 25:19. Frier BM, Saudek CD. Cholesterol metabolism in diabetes: the effect of insulin on the kinetics of plasma squalene. J Clin Endocrinol Metab 1979; 49:824. Nakayama H, Nakagawa S. Influence of streptozotocin diabetes on intestinal 3-OH-3-methylglutaryl coenzyme A reductase activity in the rat. Diabetes 1977; 26:439. Goodman MW, Michels LD, Keane WF. Intestinal and hepatic cholesterol synthesis in the alloxan diabetic rat. Proc Soc Exp Biol Med 1982; 170:286. Turley SD, Dietschy JM. Cholesterol metabolism and excretion. In: Arias IM, Jakoby WB, Popper H, et al., eds. The liver: biology and pathobiology. New York: Raven Press, 1982:467. Turley SD, Dietschy JM. Regulation of biliary cholesterol output in the rat: dissociation from the rate of hepatic cholesterol synthesis, the size of the hepatic cholesteryl ester pool, and the hepatic uptake of chylomicron cholesterol. J Lipid Res 1979; 20:923. Tavill AS. Protein metabolism and the liver. In: Wright R, Alberti KGMM, Karan S, Millward-Sadler GH, eds. Liver and biliary disease. London: WB Saunders, 1979:83. Meinders AE, Van Berge Henegouwen GP, Willekens FLA, et al. Biliary lipid and bile acid composition in insulin-dependent diabetes mellitus. Arguments for increased intestinal bacterial bile acid degradation. Dig Dis Sci 1981; 26:402. John DW, Miller LL. Regulation of net biosynthesis of serum albumin and acute phase plasma proteins. J Biol Chem 1969; 244:6134. Key PH, Bonorris GG, Coyne MJ, et al. Hepatic cholesterol synthesis: a determinant of cholesterol secretion in gallstone patients. Gastroenterology 1977; 72:1182. Sarva RP, Shreiner DP, Van Thiel DH, Yingvorapant N. Gallbladder function: methods for measuring filling and emptying. J Nucl Med 1985; 26:140. Evans DF, Cussler EL. Physiochemical considerations in gallstone pathogenesis. Hosp Pract 1974; 9:133. Schneider HL, Hornback KD, Kniaz JL, Efrusy ME. Chlorpropamide hepatotoxicity: report of a case and review of the literature. Am J Gastroenterol 1984; 79:721. Smals AG, Kloppenborg PW. Alcohol-induced Cushingoid syndrome. (Letter). Lancet 1977; 1:1369. Cheah JS, Tan BY. Diabetes among different races in similar environment. In: Waldhausl WK, ed. Diabetes. Amsterdam: Excerpta Medica, 1979:326. Hano T. Pathohistological study on the liver cirrhosis in diabetes mellitus. Kobe J Med Sci 1968; 14:87. Bloodworth JMB, Hamwi GJ. Histopathologic lesions associated with sulfonylurea administration. Diabetes 1959; 10:90. Goldstein MJ, Rothenberg AJ. Jaundice in a patient receiving acetohexamide. N Engl J Med 1966; 275:97. Rittinghaus EF, Juppner H, Burdelski M, Hesch RD. Selective determination of C-terminal (70-84) hPTH: elevated concentrations in cholestatic liver disease. Acta Endocrinol 1986; 111:62. Bouillon R, Auwerx J, Dekeyser L, et al. Serum vitamin D metabolites and their binding proteins in patients with liver cirrhosis. J Clin Endocrinol Metab 1984; 59:86. Imawari M, Akanuma Y, Itakura H, et al. The effects of diseases of the liver on serum 25-OH vitamin D and on the serum binding protein for vitamin D and its metabolites. J Lab Clin Med 1979; 93:171. Diamond T, Stiel D, Mason R, et al. Serum vitamin D metabolites are not responsible for low turnover osteoporosis in chronic liver disease. J Clin Endocrinol Metab 1989; 69:1234. Diamond TH, Stiel D, Lunzer M, et al. Hepatic osteodystrophy: static and dynamic bone histomorphometry and serum bone Gla-protein in 80 patients with chronic liver disease. Gastroenterology 1989; 96:213. Crippin JS, Jorgensen RA, Dickson ER, Lindor KD. Hepatic osteodystrophy in primary biliary cirrhosis. Effects of medical treatment. Am J Gastroenterol 1994; 89:47. Janes C, Dickson ER, Bonde S, Riggs BL. Bilirubin inhibits proliferation in cultured normal human osteoblast-like cells: a possible mechanism for bone loss in primary biliary cirrhosis. (Abstract). Gastroenterology 1992; 102:A827. Eastell R, Dickson ER, Hodgson SF, et al. Rates of vertebral bone loss before and after liver transplantation in women with primary biliary cirrhosis. Hepatology 1991; 14:296. Van Thiel DH, Lester R, Sherins RJ. Hypogonadism in alcoholic liver disease: evidence for a double defect. Gastroenterology 1974; 67:1188. Gordon GG, Olivo J, Rafii F, Southern AL. Conversion of androgens to estrogens in cirrhosis of the liver. J Clin Endocrinol Metab 1975; 40:1018. Van Thiel DH, Gavaler JS, Lester R, Goodman MD. Alcohol induced testicular atrophy: an experimental model for hypogonadism occurring in chronic alcoholic men. Gastroenterology 1975; 69:326. Van Thiel DH, Loriaux DL. Evidence for adrenal origin of plasma estrogens in alcoholic men. Metabolism 1979; 28:536. Valimaki M, Salaspuro M, Harkonen M, Ylikahri R. Liver damage and sex hormones in chronic male alcoholics. Clin Endocrinol (Oxford) 1982; 17:469. Longcope C, Pratt JH, Schneider S, Fineberg E. Estrogen and androgen dynamics in liver disease. J Endocr Invest 1984; 7:629. Bannister P, Oakes J, Sheridan P, Losowsky MS. Sex hormone changes in liver disease. Q J Med 1987; 63:305. Badr FM, Bartke A. Effect of ethyl alcohol on plasma testosterone levels in mice. Steroids 1974; 23:921. Badr FM, Bartke A, Daiyerio S, Bugler W. Suppression of testosterone production by ethyl alcohol: possible mode of action. Steroids 1977; 30:647. Gavaler JS, Van Thiel DH, Lester RR. Ethanol, a gonadal toxin in the mature rat of both sexes: similarities and differences. Alcohol Clin Exp Res 1980; 4:271.

203. Van Thiel DH, Gavaler JS, Eagon PK, et al. Alcohol and sex function. Pharmacol Biochem Behav 1980; 13:125. 204. Van Thiel DH, Gavaler JS, Rosenblum E, Tarter RE. Ethanol, its metabolism and hepatotoxicity as well as its gonadal effects: effects of sex. Pharmacol Ther 1989; 41:27. 204a. Tadic SD, Elms MS, Subbotin VM, Eagon PK. Hypogonadism precedes feminization in chronic alcohol-fed male rats. Hepatology 2000; 37:1135. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220. 221. 222. 223. 224. 225. 226. 227. 228. 229. 230. 231. 232. 233. 234. 235. 236. 237. 238. 239. 240. 241. 242. 243. 244. 245. 246. 247. 248. 249. 250. 251. 252. 253. 254. 255. 256.

Cobb CF, Ennis MF, Van Thiel DH, et al. Acethaldehyde and ethanol and direct testicular toxins. Surg Forum 1978; 29:641. Cobb CF, Van Thiel DH, Gavaler JS, Lester R. Effects of ethanol and acetaldehyde on the rat adrenal. Metabolism 1981; 30:537. Van Thiel DH, Lester R, Vaitukaitis J. Evidence for a defect in pituitary secretion of luteinizing hormone in chronic alcoholic men. J Clin Endocrinol Metab 1978; 47:499. Van Thiel DH, Gavaler JS. Endocrine effects of chronic alcohol abuse. Hypothalamic-pituitary-gonadal axis. In: Tarter RE, Van Thiel DH, eds. Alcohol and the brain. New York: Plenum Publishing, 1985:69. Smanik EJ, Barkoukis H, Mullen KD, McCullogh AJ. The liver and its effect on endocrine function in health and disease. In: Shiff L, Shiff E, eds. Diseases of the liver. Philadelphia: JB Lippincott, 1993:1373. Gavaler JS, Van Thiel DH. Endocrine consequences of alcohol abuse. In: Nemeroff CB, Loosen PT, eds. Handbook of psychoneuroendocrinology. New York: Guilford Press, 1987. Carter JN, Tyson JE, Tolis G, et al. Prolactin-secreting tumors and hypogonadism in 22 men. N Engl J Med 1978; 299:847. Van Thiel DH. Ethanol: its adverse effects upon the hypothalamic-pituitary-gonadal axis. J Lab Clin Med 1983; 101:21. Olivo J, Gordon GG, Rafii F. Estrogen metabolism in hyperthyroidism and cirrhosis of the liver. Steroids 1975; 26:47. Ryback RS. Chronic alcohol consumption and menstruation. JAMA 1977; 238:2143. Jones-Saumty DJ, Fabian MS, Parsons OA. Medical status and cognitive functioning in alcoholic women. Alcohol Clin Exp Res 1981; 5:372. James VH. The endocrine status of postmenopausal cirrhotic women. In: Langer M, Chiandussi L, Chopra IJ, Martini L, eds. The endocrine and the liver. New York: Academic Press, 1983:57. Valimaki M, Pelkonen R, Salaspuro M, et al. Sex hormones in amenorreic women with alcoholic liver disease. J Clin Endocrinol Metab 1984; 59:133. Becker U. The influence of ethanol and liver disease on sex hormones and hepatic oestrogen receptors in women. Dan Med Bull 1993; 40:447. Gavaler JS, Van Thiel DH. The association between moderate alcoholic beverage consumption and serum levels of estradiol and testosterone levels in postmenopausal women: relationship to the literature. Alcohol Clin Exp Res 1992; 16:87. Gavaler JS, Deal SR, Van Thiel DH, et al. Alcohol and estrogen levels in postmenopausal women: the spectrum of effect. Alcohol Clin Exp Res 1993; 17:786. Gavaler JS. Alcohol effects on hormone levels in normal postmenopausal women and postmenopausal women with alcohol-induced cirrhosis. Recent Dev Alcohol 1995; 12:199. Hegedus L. Decreased thyroid gland volume in alcoholic cirrhosis of the liver. J Clin Endocrinol Metab 1984; 55:930. Hegedus L, Rasmussen N, Ravn V, et al. Independent effects of liver disease and chronic alcoholism on thyroid function and size: the possibility of a toxic effect of alcohol on the thyroid gland. Metabolism 1988; 37:229. Green JRB, Strichter EJ, Mowat AG. Thyroid function and thyroid regulation in euthyroid men with chronic liver disease. Evidence of multiple abnormalities. Clin Endocrinol 1977; 7:453. Yadav HS, Chandhurij BN, Mukherjee SK. Effect of ethyl alcohol on thyroidal iodine trapping and renal clearance of 131-I label in rats. Indian J Med Res 1978; 58:1421. Van Thiel DH, Lester R. The effect of chronic alcohol abuse on sexual function. Clin Endocrinol Metab 1979; 8:499. Israel Y, Walfish PG, Orrego H. Thyroid hormones in alcoholic liver disease. Effect of treatment with 6-N-propylthiouracil. Gastroenterology 1979; 76:116. Chopra IJ, Solomon DH, Chopra U, et al. Alterations in circulating thyroid hormones and thyrotropin in hepatic cirrhosis: evidence for euthyroidism despite subnormal serum triiodothyronine. J Clin Endocrinol Metabol 1974; 39:501. Orrego H, Kalant H, Israel Y. Effect of short-term therapy with propylthiouracil in patients with alcoholic liver disease. Gastroenterology 1979; 76:105. Israel Y, Videla L, MacDonald A, Bernstein J. Metabolic alterations produced in the liver by chronic ethanol administration. Comparison between effects produced by ethanol and by thyroid hormones. Biochem J 1973; 134:523. Nomura S, Pittman CS, Chamber JB. Reduced peripheral conversion of thyroxine to triiodothyronine in patients with hepatic cirrhosis. J Clin Invest 1975; 56:643. Szilagi A. Thyroid hormones and alcoholic liver disease. J Clin Gastroenterol 1987; 9:189. Knudsen M, Christensen H, Berlid D, et al. Hypothalamic-pituitary and thyroid function in chronic alcoholics with neurological complications. Alcohol Clin Exp Res 1990; 14:363. Agner T, Hegen C, Andersen BN, Hegedus L. Pituitary-thyroid function and thyrotropin, prolactin, and growth hormone responses to TRH in patients with chronic alcoholism. Acta Med Scand 1986; 220:57. Hasselbach HC, Bech K, Eskildsen PC. Serum prolactin and thyrotropin responses to thyrotropin-releasing hormone in men with alcoholic cirrhosis. Acta Med Scand 1981; 209:37. Van Thiel DH, Gavaler JS, Sanghvi A. Lack of dissociation of prolactin responses to thyrotropin releasing hormone and metoclopramide in chronic alcoholic men. J Endocrinol Invest 1982; 5:281. Loosen PT, Prange AJ. Alcohol and anterior pituitary secretion. Lancet 1977; 2:985. Casacchia M, Rossi A, Stratta S. Thyrotropin-releasing hormone test in recently abstinent alcoholics. Psychiatr Res 1985; 16:249. Dackis CA, Bailey J, Pottash ALC, et al. Specificity of the DST and the TRH test for major depression in alcoholics. Am J Psychiatry 1984; 141:680. Garbutt JC, Mayo JP, Gillette GM, et al. Dose-response studies with thyrotropin-releasing hormone (TRH) in abstinent male alcoholics: evidence for selective thyrotroph dysfunction? J Stud Alcohol 1991; 52:275. Clark OH, Gerend PL. Effect of ethyl alcohol on the TSH-receptor-cyclase system in thyroid and non-thyroid tissues. World J Surg 1986; 10:787. Van Thiel DH, Gavaler JS, Wight WI, Abuid J. Thyrotropin releasing hormone (TRH) induced growth hormone (hGH) responses in cirrhotic men. Gastroenterology 1978; 75:66. Franz AG. Prolactin. N Engl J Med 1978; 298:201. Torro G, Kolodny RC, Jacobs LS. Failure of alcohol to alter pituitary and target organ hormone levels. Clin Res 1973; 21:505. Van Thiel DH, McClain CJ, Elson MK, McMillin MJ. Hyperprolactinemia and thyrotropin releasing factor (TRH) responses in men with alcoholic liver disease. Alcoholism 1978; 2:344. Jung Y, Russfield AB. Prolactin cells in the hypophysis of cirrhotic patients. Arch Pathol 1972; 94:265. Weiner RI, Ganong WF. Role of brain monoamines and histamine in regulation of anterior pituitary secretion. Physiol Rev 1978; 58:905. Borzio M, Calderara R, Ferrari C, et al. Growth hormone and prolactin secretion in liver cirrhosis: evidence for dopaminergic dysfunction. Acta Endocrinol 1981; 97:441. Assad SN, Cunningham GR, Samaan NA. Abnormal growth hormone dynamics in chronic liver disease do not depend on severe parenchymal disease. Metabolism 1990; 39:349. Bauer AGC, Wilson JHP, Lamberts SWJ, Blom W. Hyperprolactinemia in hepatic encephalopathy: the effect of the infusion of an amino-acid mixture with excess branched chain amino acids. Hepatogastroenterology 1983; 30:174. Sato F, Nakamura K, Taguchi M, et al. Studies on the site of ethanol action in inducing prolactin release in male rats. Metab Clin Exp 1996; 45:1330. Zanoboni A, Zanoboni-Muciaccia W. Gynaecomastia in alcoholic cirrhosis. Lancet 1975; 2:876. Zanoboni A, Zanoboni-Muciaccia W. Elevated basal growth hormone levels and growth hormone response to TRH in alcoholic patients with cirrhosis. J Clin Endocrinol Metab 1977; 45:576. Panerai AE, Salerno F, Manneschi M, et al. Growth hormone and prolactin response of thyrotropin-releasing hormone in patients with severe liver disease. J Clin Endocrinol Metab 1977; 45:134. Stein JZ, Smith WO, Ginn HE. Hypophosphatemia in acute alcoholism. Am J Med Sci 1966; 252:78. Knockel JP. The pathophysiology and clinical characteristics of severe hypophosphatemia. Arch Intern Med 1977; 137:203.

256a. Legroux-Gerot I, Blanckaert F, Solau-Gervais E, et al. Cases of osteoporosis in males. A review of 160 cases. Rev Rhum Engl Ed 1999; 66:404. 257. 258. 259. 260. 261. 262. 263. 264. 265. 266. 267. 268. 269. 270. 271. 272. 273. 274.

Nilsson BE. Conditions contributing to fracture of the femoral neck. Acta Clin Scand 1970; 136:383. Baran DT, Teitelbaum SL, Bergfeld MA, et al. Effect of alcohol ingestion on bone and mineral metabolism in rats. Am J Physiol 1980; 238:E507. Bikle DD, Genant HK, Cann C, et al. Bone disease in alcohol abuse. Ann Intern Med 1985; 103:42. Kristensson H, Lunden A, Nilsson BE. Fracture incidence and diagnostic roentgen in alcoholics. Acta Orthop Scand 1995; 51:205. Spencer H, Rubio N, Rubio E, et al. Chronic alcoholism. Frequently overlooked cause of osteoporosis in men. Am J Med 1986; 80:393. Schnitzler CM, Solomon L. Bone changes after alcohol abuse. S Afr Med J 1984; 66:730. Crilly RG, Anderson C, Hogan D, et al. Bone hystomorphometry, bone mass and related parameters in alcoholic males. Calcif Tissue Int 1988; 43:269. Purohit V. Alcohol and osteoporosis. Introduction to the symposium. Alcohol Clin Exp Res 1997; 21:383. Soszynski PA, Forhman LA. Inhibitory effects of ethanol on the growth hormone (GH)-releasing hormone–GH–insulin like growth factor-I axis in the rat. Endocrinology 1992; 131:2603. Johnell O, Nilsson BE, Wiklund PE. Bone morphometry in alcoholics. Clin Orthop 1982; 165:253. Cheung RCY, Gray C, Boyde A, Jones SJ. Effects of ethanol on bone cells in vitro resulting in increased resorption. Bone 1995; 16:143. Krawitt EL. Effect of acute alcohol administration on duodenal calcium transport. Proc Soc Exp Biol Med 1974; 146:406. Krawitt EL. Effect of ethanol ingestion on duodenal calcium transport. J Lab Clin Med 1975; 85:665. Kalbfleisch JM, Lindenman RD, Ginn HE, Smith NO. Effects of ethanol administration on urinary excretion of magnesium and other electrolytes in alcoholic and normal subjects. J Clin Invest 1963; 42:1471. Jones JE, Shane SR, Jacob WH, Flink EB. Magnesium balance studies in chronic alcoholism. Ann N Y Acad Sci 1969; 162:934. Martin HE, McCuskey C, Tupikoua N. Electrolyte disturbance in acute alcoholism. Am J Clin Nutr 1959; 7:191. Fankushen D, Roskin D, Dimich A, Wallack S. The significance of hypomagnesemia in alcoholic patients. Am J Med 1964; 37:802. Rasmussen H. Parathyroid hormone calcitonin and the calciferols. In: Williams RH, ed. Textbook of endocrinology, 5th ed. Philadelphia: WB Saunders, 1974:686.

CHAPTER 206 EFFECTS OF NONRENAL HORMONES ON THE NORMAL KIDNEY Principles and Practice of Endocrinology and Metabolism

CHAPTER 206 EFFECTS OF NONRENAL HORMONES ON THE NORMAL KIDNEY PAUL L. KIMMEL, ANTONIO RIVERA, AND PARVEZ KHATRI Pathways of Hormonal Action on the Kidneys Mineralocorticoids Glucocorticoids Antidiuretic Hormone Parathyroid Hormone Calcitonin Insulin Glucagon Estrogen Progesterone Thyroid Hormones Prolactin Catecholamines Growth Hormone Atrial Natriuretic Hormone Chapter References

PATHWAYS OF HORMONAL ACTION ON THE KIDNEYS Generally, there are four pathways by which renal function may be modified by hormones: 1. 2. 3. 4.

Changes in systemic hemodynamics, which may induce secondary changes in renal function Changes in glomerular hemodynamics, such as afferent or efferent arteriolar tone Changes in the glomerular ultrafiltration coefficient Changes in the tubular handling of solutes and water

Many hormones profoundly affect normal renal function: mineralocorticoids, glucocorticoids, antidiuretic hormone (ADH), parathyroid hormone (PTH), calcitonin, insulin, glucagon, estrogen, progesterone, thyroid hormones, prolactin, catecholamines, growth hormone, and atrial natriuretic hormone (Table 206-1). Hormones regulate both active and passive transport processes in part through activation of channels, stimulating the synthesis of new channels, promoting redistribution of channels, or modulating the extent of active transport of various solutes at the level of the tubules. Such responses mediate the ability of the kidneys to respond to varying internal and external environmental stresses. These influences and responses are considered in this chapter.

TABLE 206-1. Kidney Responses to Nonrenal Hormones*

MINERALOCORTICOIDS Mineralocorticoid receptors1 are present in the cortical and medullary collecting and late distal convoluted tubules2 (Fig. 206-1). These receptors are nonspecific, because they equivalently bind both mineralocorticoids and glucocorticoids.3 However, glucocorticoids such as cortisol and corticosterone do not stimulate mineralocorticoid receptors in vivo because they are metabolized into inactive 11-keto congeners by the action of a specificity-conferring enzyme system, 11b-hydroxysteroid dehydrogenase (11-HSD).4 Aldosterone is protected from inactivation by dehydrogenation. 11-HSD activity is inhibited by licorice, potentially increasing mineralocorticoid-mediated cellular responses.5 Another area of current interest involves the interaction of aldosterone with membrane receptors that mediate physiologic changes through nongenomic mechanisms.6

FIGURE 206-1. This figure provides a schema of the anatomy of the glomerulus and the various tubular segments that comprise the nephron, the functional unit of the kidney. The glomerulus is a capillary network, lined by endothelial cells, supported by a central region of mesangial cells and matrix material, within the Bowman capsule. An afferent arteriole enters each glomerulus at its hilum and subsequently divides into lobules, which form the glomerular tuft. The capillaries rejoin to form the efferent arteriole, which exits the glomerulus at the vascular pole. The proximal tubule originates from the urinary pole of the Bowman capsule and is composed of a convoluted and a straight section. The loop of Henle is composed of a thin segment, with descending and ascending portions, and a thick segment, with medullary and cortical portions. The distal convoluted tubule is variably composed of bright, granular, and light segments. The region of transition between the distal convoluted tubule and the collecting duct is the connecting segment, which is composed of a mixture of cells, including distal convoluted tubule, and connecting, intercalated, and principal cells. The collecting duct extends from the connecting segment in the cortex to the papillary tip and is divided into the cortical collecting tubule and the outer and inner medullary collecting ducts. The segments of the collecting duct are composed of varying proportions of principal and intercalated cells. The cellular composition and length of the individual nephron segments may vary with the species and the anatomic location of the glomerulus.

Aldosterone mediates increased sodium reabsorption in distal nephron segments, which is characterized by increased luminal (apical) membrane sodium conductance, increased de novo synthesis of basolateral sodium-potassium adenosine triphosphatase (ATPase) and citrate synthase, and greater luminal electronegativity of the late distal convoluted, connecting, and collecting tubules.7,8,9 and 10 After mineralocorticoid treatment, there is an increase in the basolateral membrane surface area of the principal cells of the connecting and collecting tubules. The cortical collecting duct is the main site affected.11 The gene encoding the epithelial sodium channel (ENaC), a multimeric protein composed of several subunits,12 has been cloned. This channel is expressed in the

distal convoluted tubule, the cortical collecting tubule, and the collecting duct.12 Aldosterone induces activation of sodium channels within 1.5 to 3.0 hours as an early response, and synthesis of sodium channels over 6 to 24 hours as a late response. Aldosterone may also mediate changes in the subcellular localization of the channels.13 Mineralocorticoids do not directly affect glomerular filtration rate or renal blood flow.14 The short-term administration of mineralocorticoids causes a decrease in urinary sodium excretion. During prolonged treatment with mineralocorticoids, urinary sodium excretion increases to normal levels. This phenomenon is known as mineralocorticoid escape. The immediate effect probably is mediated by changes in tubular sodium handling. The long-term effect likely reflects the balance of systemic and tubular influences on sodium excretion. This sodium retention, which is mediated by tubular mechanisms, causes progressive extracellular fluid volume expansion and usually an increase in systemic blood pressure. The volume expansion and increased renal arterial pressure cause inhibition of proximal tubular fluid reabsorption, resulting in increased urinary sodium excretion, opposing the distal tubular effects of the hormone15,16 (see Table 206-1). Aldosterone is necessary for the maintenance of maximum potassium excretion by the kidney, especially in the case of adaptation to a high-potassium diet (see Chap. 79). Aldosterone increases renal potassium excretion by increasing distal tubular Na+ /K+-ATPase activity, by increasing its synthesis, and by targeting and insertion of new Na+ /K+-ATPase units into the basolateral membrane.17 Such changes mediate an increased renal tubular intracellular potassium concentration. Aldosterone increases basolateral potassium permeability as well as luminal cell membrane potassium conductance. Mineralocorticoids increase potassium secretion by the principal cells. A low-conductance potassium channel mediates potassium secretion across the apical membrane of principal cells in the cortical collecting tubule.17 Aldosterone also enhances apical principal-cell potassium-channel density.17 Increased potassium excretion results as a consequence of the increased negative transepithelial voltage potential difference, which varies directly with ambient mineralocorticoid levels. The net result of mineralocorticoid action is that urinary potassium excretion increases as the circulating potassium concentration falls. In the case of the physiologic response of mineralocorticoid synthesis to volume depletion, marked changes in potassium balance are limited since the urinary flow rate decreases. In instances of diuretic-induced volume depletion, the increased urinary flow rate in patients with secondary mineralocorticoid stimulation may be associated with marked increases in urinary potassium excretion and with the development of hypokalemia and a negative potassium balance. The time courses of functional responses in the nephron segments are different for sodium and potassium transport.17 It is unclear to what extent short-term changes in plasma aldosterone concentration alter renal potassium handling.18 The administration of mineralocorticoids to normal and mineralocorticoid-deficient persons increases urinary acidification by two mechanisms: increased hydrogen ion excretion and enhanced ammonia production.19,20 Aldosterone increases hydrogen ion secretion by type A intercalated cells in the collecting duct via two mechanisms: direct stimulation of the proton pump (hydrogen-translocating ATPase), and, indirectly, by stimulating sodium influx, which creates a lumen-negative potential difference.19 Studies in isolated collecting tubules demonstrate acute, non–sodium-dependent, in vitro influences on bicarbonate transport after mineralocorticoid treatment. Although bicarbonate reabsorption is stimulated by mineralocorticoids in the medullary collecting duct, an opposite effect occurs in the cortical collecting duct. Systemic acid-base balance may be a more important modulator of bicarbonate handling by the cortical collecting tubule than are mineralocorticoids.21

GLUCOCORTICOIDS Dexamethasone receptors have been identified in the proximal tubule and cortical collecting tubule, with a modest distribution in other nephron segments. The distribution of corticosterone receptors resembles that of mineralocorticoid receptors, although some also are located in the proximal tubule and thick ascending limb of Henle.2 Glucocorticoid hormones activate both glucocorticoid and mineralocorticoid receptors located in the cytoplasm. In cells expressing 11-HSD (such as renal epithelia), glucocorticoids are metabolized and inactivated before they stimulate mineralocorticoid receptors.4 When glucocorticoid or mineralocorticoid receptors are activated, they enter the nucleus and enhance transcription.4 Although exogenous glucocorticoids stimulate renal Na+/K+ -ATPase, this does not occur with physiologic doses. Glucocorticoids do not affect citrate synthase production or the basolateral membrane area. The direct effects of glucocorticoids on the kidney are difficult to evaluate, because glucocorticoid deficiency in humans is associated with alterations in cardiac output and changes in systemic hemodynamics. Moreover, animals that have undergone adrenalectomy or humans who have Addison disease have concurrent deficiencies in mineralocorticoid and catecholamine production. Glucocorticoids increase renal blood flow and glomerular filtration rate, without affecting glomerular or capillary pressure or permeability—essentially a vasodilatory effect.22 Cortisone infusion decreases sodium excretion without affecting the excretion of chloride. Glucocorticoids may increase renal potassium excretion. The onset of the kaliuresis is more rapid and of shorter duration than that induced by mineralocorticoids, and it is relatively independent of dietary factors such as sodium and potassium intake. Because a direct, immediate action of glucocorticoids is not observed in the distal nephron, these effects on electrolyte excretion may be the result of changes in luminal flow and renal hemodynamics. It is possible that glucocorticoids modulate the level of sensitivity of the distal nephron to other factors that regulate ion transport.18 Although glucocorticoids stimulate renal gluconeogenesis and ammoniagenesis, and may affect luminal sodium and hydrogen ion exchange in the proximal tubule, a net effect on hydrogen ion secretion has been difficult to document.2,23 Acidosis increases adrenal secretion of glucocorticoids, which in turn increase proximal tubule apical membrane Na+/H+ -antiporter activity in vivo; in vitro glucocorticoids enhance both the ability of acidosis to increase levels of cellular Na+/H+-exchanger 3 protein (likely through increasing synthesis) and the acidosis-induced trafficking of this protein to the apical membrane.24 In this manner, glucocorticoids facilitate the renal response to systemic acidemia by facilitiating net hydrogen ion excretion. In humans, the deficiency of glucocorticoids is associated with a renal diluting defect. The mechanism is controversial. Although in vitro data suggest that glucocorticoids directly influence distal tubular permeability to water, the impaired water excretory ability may result from concurrent alterations in glomerular filtration rate, plasma volume, ADH release, or distal tubular fluid delivery.14 Rapid administration of glucocorticoids does not change urinary calcium or magnesium excretion, although variable changes in urinary phosphate excretion have been reported. However, the long-term administration of glucocorticoids is associated with hypercalciuria and hypermagnesuria. The mechanism may be secondary to bone dissolution or an increase in plasma volume.25

ANTIDIURETIC HORMONE The action of ADH (arginine vasopressin [AVP]) is mediated through its binding to specific receptors.26 V1 receptors are found in the vascular smooth muscle and liver. V3 receptors (also known as V1b receptors) are present in the neurons of the adenohypophysis and may play a role in mediating AVP-induced corticotropin secretion.26 V2 receptors are found in renal epithelial cells. Although the presence of the V2 receptor in endothelial cells has been inferred from pharmacologic studies, expression of its messenger RNA (mRNA) and protein has not yet been documented. ADH receptors have been localized in cells of the late distal convoluted tubules, the connecting tubule, the cortical collecting and medullary collecting ducts, and the glomerulus.27,28 ADH-sensitive adenylate cyclase is found in isolated rabbit glomeruli, and contractile responses have been induced by ADH in cultured rat mesangial cells.29 This may explain the mechanism whereby ADH decreases the glomerular ultrafiltration coefficient in rats, although the single nephron glomerular filtration rate remains unchanged.30 V2-receptor transcripts are heavily expressed in cells of the renal connecting tubule and of the cortical, outer medullary, and inner medullary collecting ducts. Although the V2 receptor is expressed in thick ascending limbs of the loops of Henle of the rat, its presence in this segment of the human nephron is currently debated. V2 receptors on the basolateral membrane of the principal cells of the collecting ducts are activated by AVP binding. V2 receptors are coupled to an adenylate cyclase stimulatory G protein (Gs). Binding of AVP to V 2 receptors on renal epithelial cells results in increased synthesis of cyclic adenosine monophosphate (cAMP), in activation of protein kinase A (PKA), and in increased synthesis of water channels (aquaporins[AQP]) and their cAMP-mediated incorporation into the luminal surface of these cells.31,31a At the collecting duct cell membranes, wide variations in permeability are achieved by differential regulation of the synthesis, trafficking, insertion, and removal of different types of aquaporins.31 At least five AQPs have been identified. AQP-1, AQP-2, AQP-3, and AQP-4 are found in the kidneys. AQP-1, or channel-forming integral membrane protein (CHIP), a 28-kDa protein, was the first molecular water channel identified. AQP-1 is expressed in mammalian red cells, renal proximal tubules, thin descending limbs, and other water-permeable epithelia. AQP-1 mediates isosmotic volume reabsorption in the proximal tubule and facilitates the function of the countercurrent multiplier in the loop of Henle. AQP-1 is localized in both apical and basolateral plasma membranes, where it may function both as an entry and as an exit route for transepithelial water transport. In contrast to AQP-2, limited amounts of AQP-1 are localized in membranes of vesicles or vacuoles. Tubule segments that are not water permeable (such as

the thin and thick ascending limbs of the loop of Henle) do not express AQPs. AQP-2 is the vasopressin-regulated water channel in renal collecting ducts. It is found from the connecting tubule, through the cortical collecting duct, and through the entire inner medullary collecting duct. ADH-induced PKA-mediated phosphorylation of the carboxy terminus of AQP-2 results in insertion and fusion of vesicles containing AQP-2 into the apical membrane, causing increased water permeability. AQP-2 is diffusely distributed in the cytoplasm in states of hydration. In contrast, apical localization of AQP-2 is intensified in dehydration or after vasopressin administration. These observations are thought to represent the insertion of preformed water channels by exocytosis into the apical plasma membrane from intracellular vesicles (the “shuttle hypothesis”). After the ADH-induced insertion of intracellular vesicles containing water channels into the apical membrane,32 the tubular reabsorption of water is facilitated in the face of a renal osmotic gradient, decreasing free water excretion and producing relatively concentrated urine. In the absence of ADH, the water channels are transported into the cell for further processing or for subsequent reinsertion into the apical membrane29,32 (see Chap. 25). There are at least three ways in which AVP regulates the action of AQP-2: AVP regulates synthesis, controls trafficking from cytoplasm to the membrane, and regulates the balance between exocytosis and endocytosis of the water channels. Moreover, studies have provided evidence for the existence of vasopressin “escape,” a selective down-regulation of AQP-2 expression, decreased cAMP signaling, and decreased activity of intracellular mediators—a process that occurs in states of AVP excess.33,34 (See also ref. 34a.) AQP-3 and AQP-4 are expressed in the basolateral membrane of cortical and medullary collecting-duct cells. AQP-3 is the water channel in basolateral membranes of renal medullary collecting ducts, and AQP-4 plays a critical role in the inner medullary collecting duct. Basolateral AQP-3 and AQP-4 mediate the exit of cell water, ultimately into the extracellular fluid.31,35 Genetic abnormalities in the regulation of synthesis of functional V2 receptors and AQP-2 are associated with nephrogenic diabetes insipidus.36 Calcium binding to the luminal calcium receptor of the inner medullary collecting duct cells causes endocytosis and down-regulation of AQP-2, reducing water permeability.31,37 This may be one mechanism whereby hypercalcemia mediates decreased renal concentrating ability. A similar down-regulation has been noted in rats treated with lithium, and subjected to dietary potassium deprivation, perhaps explaining the concentration defect noted in patients with severe, prolonged hypokalemia.38 Increased sodium and potassium excretion are associated with the administration of ADH, but these effects are variable and may be species-specific. In humans, it is unclear whether such changes represent alterations in systemic hemodynamics secondary to mild volume expansion or a direct tubular influence of the hormone on electrolyte transport. ADH may stimulate sodium reabsorption in the loop of Henle, but may diminish it in the cortical collecting tubule.39,40 Other studies demonstrate that the activity of ENaCs in cortical collecting-duct principal cells is regulated by ADH. ADH-induced increases in intracellular levels of cAMP result in insertion of channels into the plasma membrane and increased sodium reabsorption. The ADH effect is greater in the presence of aldosterone, but it is detectable even when the levels of circulating aldosterone are very low.12 Discrepancies in findings between different studies may have resulted from species differences, from variations in cell types studied, or from dietary conditions or technical approaches. ADH stimulates potassium secretion in the distal convoluted tubule, but has little effect on potassium transport in the cortical collecting tubule40,41 Patients with the syndrome of inappropriate ADH secretion may have hypercalciuria (see Chap. 27). Although the hypercalciuria has been attributed to volume expansion and increased sodium excretion, an in vitro study suggests that the hormone directly inhibits both calcium and phosphate reabsorption by the cortical collecting tubule.42 AVP also increases urea permeability in the inner medullary collecting duct. The gene encoding a renal urea transporter, UT-2 (UT-A2), was cloned in 1993.43 Subsequently, several urea transporters have been identified in different tissues and species, some of which are vasopressin-responsive.44 AVP stimulates the expression of an inner medullary collecting-duct urea transporter (UT-1, UT-A1)45 and urea transport—which results in enhanced medullary tonicity (a phenomenon known as “urea trapping”)—thereby increasing the efficiency of the renal concentrating mechanism, especially in volume-depleted or dehydrated subjects. The physiologic roles and responses of this system in humans remain to be fully elucidated.

PARATHYROID HORMONE The PTH receptor (a member of the G protein–linked 7-membrane-spanning–receptor family) has been cloned in animals and humans,46,47 and 48 and exhibits homology with the calcitonin receptor.46 PTH receptors are found in the glomerulus, proximal tubule, medullary and cortical thick ascending limb of Henle, and early distal convoluted tubule.27 Molecular biology studies of microdissected nephron segments demonstrate the presence of mRNAs coding for the parathyroid hormone/parathyroid hormone–related protein (PTH/PTHrP) receptor and an extracellular, G protein–coupled Ca2+-sensing receptor (RaKCaR) in the glomeruli, proximal tubules, thick ascending limbs, distal convoluted tubules, and collecting ducts of rats and mice.49,50 The reason for the differences noted in the distribution of receptor mRNA seen in these two studies is unclear. The action of PTH probably is mediated primarily through cAMP (see Chap. 51). Renal PTH receptors are down-regulated when exposed to PTH. Binding of PTH to its receptor stimulates intracellular signaling by both cAMP and inositol-(1,4,5)-triphosphate (IP3)/diacylglycerol (DAG) pathways.51 After PTH binds to its receptor, Gs couples to adenylate cyclase and stimulates production of cAMP, which in turn activates PKA, while Gq couples to phospholipase C (PLC) to form IP3 and DAG from phosphatidylinositol-(4,5)-biphosphate (PIP2). IP3 releases calcium from intracellular stores, and DAG stimulates protein kinase C (PKC) activity, ultimately mediating effects at the cell membrane.52 In thyroparathyroidectomized animals, PTH infusion diminishes whole kidney and single nephron glomerular filtration rates, most likely by reducing the glomerular ultrafiltration coefficient. Although volume depletion is a factor, PTH may partially mediate the decreased glomerular filtration rate seen in hypercalcemic patients with hyperparathyroidism. The physiologic role of PTH in modulating the glomerular filtration rate in humans is unclear.53 In clearance studies in humans, PTH infusion increases renal excretion of phosphate, sodium, potassium, and bicarbonate, and reduces the excretion of calcium and magnesium. The hormone inhibits the reabsorption of fluid, sodium, chloride, calcium, and phosphate in the proximal tubule of the dog and rat. High levels of PTH stimulate calcium and magnesium reabsorption in the thick ascending limb of Henle in the absence of substantial changes in sodium and chloride transport.54,55 and 56 Microperfusion studies confirm that PTH stimulates calcium reabsorption in the distal convoluted tubule. Granular tubular epithelial cells are specifically involved. PTH stimulation of calcium uptake in the distal convoluted and connecting tubules occurs via apical calcium entry, most likely through the insertion of voltage-operated, dihydropyridine-sensitive calcium channels into the membrane. Stimulation of basolateral sodium/calcium exchanger and/or calcium/ATPase may also play a role in enhancing calcium reabsorption in these nephron segments.55 The hallmark of PTH action on the kidney is phosphaturia57 PTH stimulates gluconeogenesis and nonspecifically inhibits phosphate reabsorption, in the proximal straight tubule, via cAMP/PKA and IP3/PKC pathways.58 PTH decreases brush border membrane (BBM) sodium/phosphate cotransport and reduces the content of type II sodium/phosphate cotransporter protein. 59,60 Likewise, parathyroidectomy causes an increase in cotransporter protein levels.60 Most of the effect of PTH on phosphate balance occurs in the distal nephron. PTH inhibits phosphate reabsorption in the late segments of the distal convoluted tubule.58 Renal 1a-hydroxylase activity is increased by PTH, thus enhancing the production of 1,25-dihydroxyvitamin D [1,25(OH)2D] from 25-hydroxyvitamin D in the proximal convoluted tubule by a cAMP-mediated process.61 Findings suggest that in addition to the proximal convoluted tubule, the murine distal convoluted tubule expresses a PTH-responsive 25-hydroxyvitamin D3–24-hydroxylase, but the effects are mediated by different mechanisms.62 The role of PTH in modifying acid-base balance in humans is unclear. The immediate effect of the hormone is an inhibition of proximal tubular bicarbonate reabsorption. Although metabolic acidosis occurs with acute PTH infusion and in patients with hyperparathyroidism, the steady-state and long-term, direct renal effect of the hormone [in the absence of hypercalcemia and an increase of circulating 1,25(OH)2D] appears to be an increase in renal bicarbonate reabsorption or distal tubular hydrogen ion secretion, causing metabolic alkalosis. PTH also indirectly stimulates distal tubular hydrogen ion secretion and titratable acid excretion by increasing phosphate delivery to the distal nephron.63 The overall systemic effect may be modulated by time or by calcium and vitamin D homeostasis.64

CALCITONIN Calcitonin membrane receptors have been identified in renal cortical and medullary tissue.65 Calcitonin-receptor mRNA is present in the cortical and medullary thick ascending limbs and the cortical collecting duct.66 The calcitonin receptor is a G protein–coupled receptor with seven-membrane-spanning regions, coupled by Gs to adenylate cyclase. The PTH and calcitonin receptors are family members that have similar amino-acid sequences, although their ligands do not. Calcitonin-sensitive

adenylate cyclase activity is found primarily in the thick ascending limb of Henle, distal convoluted tubule, and collecting tubule in humans.67 Calcitonin infusion causes natriuresis and kaliuresis in humans, but does not affect the glomerular filtration rate. High-dose infusion causes phosphaturia in humans, although the contribution of circulating PTH cannot be excluded. Calcitonin increases phosphate excretion in thyroparathyroidectomized animals and in humans with hypoparathyroidism, suggesting a direct renal effect. The increase in urinary magnesium excretion has been variable. Although calcitonin infusion in humans increases urinary calcium excretion, these findings are controversial and may depend on dosage or the species of the hormone. A direct renal effect on calcium handling is difficult to dissociate completely from its skeletal effects,58 but studies suggest that calcitonin has a direct effect on distal convoluted tubule cells, leading to an increase in calcium reabsorption, which is mediated by calcium channels.68 In the absence of ADH, calcitonin may play a role in the urinary-concentrating mechanism by stimulating adenylate cyclase pools, which are similar to those activated by ADH. In the absence of ADH, calcitonin causes an increase in calcium, magnesium, potassium, sodium, and chloride reabsorption by the loop of Henle.39 Calcitonin stimulates 1a-hydroxylase activity in the proximal straight tubule through cAMP-independent mechanisms, although this effect may be species-specific.61

INSULIN Insulin binding occurs in the glomerulus, proximal tubule, distal convoluted tubule, and medullary ascending limb of Henle.69 These sites, however, may be involved in insulin metabolism rather than mediation of tubular responses.70 The functional significance of glomerular insulin receptors is unknown, because the administration of insulin to healthy humans does not affect the glomerular filtration rate.71,72 and 73 The influence of insulin on tubular function is more significant. The binding of insulin to its receptor on the basolateral membrane of the proximal tubule cell initiates the phosphorylation of a receptor subunit. This activated subunit mediates the subsequent phosphorylation of intracellular protein substrates, which, in turn, mediate the hormone's physiologic effect. Studies in animals and humans demonstrate that the administration of insulin decreases urinary sodium excretion in the absence of changes in plasma glucose, renal blood flow, or glomerular filtration rate.71 In humans, insulin administration sufficient to achieve circulating levels between 41 and 90 mU/mL reduced sodium excretion in a dose-dependent manner, suggesting a distal nephron site of action.72 This effect seems to be a direct one and has been documented in isolated perfused kidneys without changes in renal hemodynamics. Animal studies suggest that the response is dependent on ADH, but independent of the angiotensin and prostaglandin systems.71 Although sodium and hydrogen ion exchange may be increased, insulin decreases gluconeogenesis and sodium reabsorption by the proximal tubule.74 Physiologic doses of insulin stimulate electrogenic sodium transport in experimental models of the cortical collecting duct75 by mechanisms independent of transcription or protein synthesis. Aldosterone and insulin act independently to stimulate apical sodium entry into A6 epithelial cells by increasing the sodium channel density.76 Insulin induces an increase in apical cell membrane sodium permeability, by increasing the number of activated sodium channels and the length of time they remain open.77 The principal site of insulin action affecting urinary sodium excretion probably is the thick ascending limb of Henle or more distal segments of the nephron.78 An insulin-responsive glucose transporter (GLUT4) is expressed at low levels in the kidney, primarily in the renal vasculature and in glomerular epithelial and mesangial cells.79 Insulin administration may decrease urinary potassium excretion, but the site and mechanism of action are obscure. An important factor may be the hormone's influence on extrarenal potassium disposition. The administration of insulin causes potassium uptake by the intracellular fluid, decreasing the plasma potassium concentration and the total potassium load presented for glomerular filtration and renal excretion. On the other hand, insulin decreases potassium secretion by the isolated perfused kidney without changing renal hemodynamics. This suggests a direct tubular effect of the hormone on potassium transport or an indirect diminution in renal potassium secretion secondary to the concurrent reduction in sodium excretion.78 The potential interactions in patients between hormonal effects on circulating ion levels and tubular functional changes are delineated in a clinical study of insulin administration during water diuresis, in which lithium, glucose, and/or sodium or potassium chloride were administered.73 Potassium chloride infusion, to an extent that prevented the development of hypokalemia, but not sodium chloride repletion, abrogated the typical insulin-induced responses of urinary sodium and potassium excretion. Although the extent to which changes in potassium metabolism mediate insulin-induced renal tubular effects remains controversial, these results suggest that at least some of the findings in older clearance studies may have been affected by insulin-induced changes in circulating potassium levels, and that in vivo studies must be carefully designed and controlled to account for the multiple physiologic changes that occur in patients during hormone administration or dysregulation. Insulin administration increases urinary calcium excretion, presumably acting at the proximal tubule, although a distal tubular effect is possible in humans. Urinary phosphate excretion decreases after insulin administration, as a result of increases in proximal tubular reabsorption.78 Infusion of insulin in normal humans decreases both the renal clearance and fractional excretion of uric acid, perhaps as a result of the decreased sodium excretion.80

GLUCAGON Glucagon-induced adenylate cyclase activation has been documented in the thick ascending limb of Henle, distal convoluted tubule, cortical collecting tubule, and medullary collecting duct.81 The glucagon receptors of the rat and human 82 have been cloned. The glucagon receptor, which is expressed in the kidney,83,84 and 85 is a seven-transmembrane domain receptor with a conserved G protein–binding site and an aminoterminus domain that is involved in ligand binding.85 Glucagon infusion in supraphysiologic doses increases renal blood flow and glomerular filtration rate, associated with natriuresis, calciuria, and phosphaturia. Changes in electrolyte excretion depend on changes in renal hemodynamics, rather than direct tubular effects. No significant alterations in electrolyte transport have been documented in in vitro, isolated systems.78 The effect of physiologic changes in the level of circulating glucagon on renal function is unclear.86 In the absence of ADH, glucagon increases calcium, magnesium, and potassium reabsorption by the thick ascending limb of Henle.87 Different cell types or pools of cAMP may be activated by each hormone.

ESTROGEN Estrogen receptors have been identified in rat renal tissue.14,88,89 Estrogen receptors have been detected in renal tissue by direct methods, and their presence has been inferred as a result of responses after exposure to tamoxifen, a competitive inhibitor of estrogen.90 The a, but not the b, subtype estrogen receptor has been identified in the kidney.90,91 Interest has been directed toward understanding the effects of sex hormones in renal physiology since the progression of many renal diseases is slower in males than in females both in humans and in animal models92 17b-Estradiol increases growth of proximal tubular cells in culture, with a peak effect at 10–9 to 10–10 mol/L, but higher doses are inhibitory.90 It increases the proliferation of mesangial cells at 10- to 100-nmol/L concentrations, but suppresses their proliferation at 10-mmol/L concentrations.93 Moreover, it markedly suppresses mesangial-cell collagen synthesis at 1-and 10-mmol/L concentrations.93 These effects may be mediated by antagonizing the actions of transforming growth factor-b.94 Estrogen replacement therapy increases inducible nitric oxide synthase in the renal medulla of oophorectomized rats.95 Few data exist on the effects of estrogen on glomerular filtration rate and renal electrolyte handling.96,97 Although the administration of estrogen to humans in physiologic doses has no influence on renal blood flow, glomerular filtration rate,98 urine flow, or the tubular handling of glucose, estradiol administration causes sodium retention. The mechanism is unknown, but stimulation of the renin–angiotensin–aldosterone system or secondary systemic vasodilatation has been suggested to explain the diminution in sodium excretion.99 Long-term estrogen administration decreases renal calcium excretion and phosphate reabsorption.25 This may be the result of skeletal effects, rather than a direct renal effect.

PROGESTERONE Although definitive evidence of a progesterone receptor in the human kidney remains inconclusive, pharmacologic effects of this hormone may be manifested through interactions with other pathways. Progesterone binds to mineralocorticoid and glucocorticoid receptors and may act partly by competitively inhibiting aldosterone. Two isoforms of the progesterone receptor (hPR-A and hPR-B) have been identified.100 The A form inhibits glucocorticoid, androgen, and mineralocorticoid receptor-mediated gene transcription. Progesterone at concentrations of 10–9 to 10–10 mol/L inhibited growth of rabbit proximal tubular cells in culture.90 Pharmacologic doses of progesterone have displaced aldosterone from binding sites in the toad bladder and have blunted its physiologic effects, but physiologic doses have had little influence on this system. Increased renal blood flow and glomerular filtration rate occur after the administration of pharmacologic doses of progesterone.

Progesterone causes natriuresis and has been implicated in the blunting of mineralocorticoid-induced kaliuresis.99,101

THYROID HORMONES Hyperthyroidism has been associated with an increase in renal blood flow and glomerular filtration rate, and hypothyroidism has been associated with a diminution of these functions. These alterations have proved to be reversible after patients become euthyroid. Thyroxine supplementation in humans causes small increases in both glomerular filtration rate and renal blood flow. These changes may reflect effects on systemic hemodynamics. The maximum tubular reabsorption of glucose is increased in hyperthyroidism. Thyroid hormone supplementation enhances free water excretion, consistent with a renal concentrating defect. Both urinary diluting and, to a lesser extent, concentrating abilities are impaired in hypothyroidism. The mechanism is controversial: glucocorticoid deficiency, alterations in plasma volume and renal sodium reabsorption, increased ADH secretion, and decreased glomerular filtration rate and distal tubular fluid delivery have been implicated. Hyperthyroidism has been associated with increased urinary calcium and magnesium excretion, and with tubular reabsorption of phosphate. These changes may be secondary to changes in ionized calcium or PTH levels, rather than to direct renal effects.40 Thyroid hormones do, however, have a direct effect on phosphate transport in renal cells.102 Renal phosphate reabsorption in proximal tubular BBM vesicles was increased in a dose-dependent manner in animals treated with triiodothyronine (T3), while it was diminished in vesicles from hypothyroidrats.103 The membrane findings occurred concurrently with consonant changes in levels of the sodium-phosphate transporter protein, and expression of the NPT-2 gene. These findings suggest a role for thyroid hormone in long-term regulation of phosphate homeostasis. Animal studies have suggested a sodium reabsorptive defect in hyperthyroidism; however, the physiologic significance of this in humans is uncertain. Hypothyroidism may be associated with impaired renal tubular acidification. The mechanisms may involve abnormalities in sodium transport, carbonic anhydrase, or glutaminase activity.104 Thyroid hormone also modulates Na+/K+ -ATPase activity in the kidney.61

PROLACTIN Prolactin infusion results in an unchanged or slightly elevated glomerular filtration rate. Data on the effects of this hormone in humans remain scanty, but studies suggest decreases in urinary excretion of water, sodium, and potassium, and increases in urinary excretion of calcium.105 Animal studies suggest direct effects on the renal tubular handling of sodium, potassium, water, and calcium.101,106,107

CATECHOLAMINES The direct effects of catecholamines on renal function have been difficult to ascertain. Endogenous catecholamines with different levels of a-and b-agonism affect systemic determinants of renal function, such as blood pressure, peripheral resistance, renal blood flow, and cardiac output. Adrenergic infusion also may modify the secretion of other hormones, such as ADH, renin, and prostaglandin E2, which, in turn, affect renal function. Dopamine, and a-and b-adrenergic receptors, have been isolated from renal tissue.108 b-Adrenergic receptors have been localized in glomeruli, the ascending limb of Henle, the distal convoluted tubule, and the collecting tubule.109 Isoproterenol stimulates adenylate cyclase activity in the connecting and cortical collecting tubules.81 a-Adrenergic agents cause renal vasoconstriction, resulting in a dose-dependent decrease in glomerular filtration rate and renal blood flow, and an increase in the filtration fraction (i.e., glomerular filtration rate divided by renal plasma flow). b-Adrenergic agents have little effect on renal hemodynamics. The glomerular filtration rate tends to remain constant, in association with an increased renal plasma flow, resulting in a decreased filtration fraction.108 Although norepinephrine receptors have been localized in the glomerulus, infusion of the hormone does not necessarily directly change the single nephron glomerular filtration rate, despite preferential vasoconstriction of the afferent arteriole. 110 Dopamine infusion at relatively low doses may specifically increase the glomerular filtration rate and urinary sodium excretion, most likely by changing renal hemodynamics.103 Catecholamine infusions tend to decrease urinary sodium excretion, probably primarily as a result of an a-effect, perhaps mediated largely by changes in renal hemodynamics. In the isolated perfused kidney, this effect is blocked by propranolol, suggesting primarily a b-effect, but a2-agonists also may decrease sodium excretion in this system.111 Catecholamine infusions increase urinary calcium excretion in animals, an effect probably mediated by a-agonists. The mechanism may be unrelated to changes in renal sodium handling.25 Norepinephrine infusion increases free water excretion in humans and animals, in the absence of changes in the glomerular filtration rate, which is primarily an a-effect. b-Agonists usually cause antidiuresis, although this effect may be a result of ADH release (i.e., extrarenal).108 b-Agonists do not affect the osmotic permeability of the cortical collecting tubule.112 The a2-agonists specifically interfere with ADH-induced increases in osmotic permeability, presumably by interfering with the generation of cAMP.113 The a2-agonists inhibit cAMP formation by PTH in the proximal convoluted tubule and by ADH in the cortical collecting tubule, but not in the thick ascending limb of Henle114 a2-Agonists inhibit AVP-stimulated water and urea permeability in the rat inner medullary collecting duct.44 The a2-agonists may antagonize the renal effects of other hormones that activate adenylate cyclase. a2-Agonists stimulate Na+/K+ -ATPase activity through PKC-mediated pathways, involving an increase in intracellular levels of IP3 and DAG.115 In the proximal tubule, both a-and b-agonists stimulate fluid reabsorption.61 In the late proximal tubule, a-and b-agonists have little effect, but dopamine decreases fluid reabsorption by non–cAMP-mediated events.116 In the cortical collecting tubule, b-agonists decrease potassium secretion and increase chloride reabsorption, but have little effect on sodium reabsorption.41,112

GROWTH HORMONE Although growth hormone receptors have been identified in isolated proximal tubular basolateral membranes,117 most of the renal effects of growth hormone appear to be mediated by its ability to stimulate the synthesis of insulin-like growth factors (IGFs), especially IGF-I. Data concerning the role of growth hormone in renal function in normal humans are scanty and contradictory.104,118,119,120 and 121 The rapid infusion of growth hormone has little influence on glomerular filtration rate, and renal plasma flow remains unchanged or decreases.122 Animal studies have not suggested an acute effect of the hormone on glomerular filtration rate, renal blood flow, or the clearance of sodium, calcium, or phosphate.123 Studies in patients with acromegaly (see Chap. 12) and in normal persons during the sustained administration of growth hormone suggest increased glomerular filtration rate, renal blood flow, and calcium and magnesium excretion.104,118 The infusion of growth hormone in normal humans produces a delayed increase in glomerular filtration rate and renal plasma flow that correlates in time with increased levels of circulating IGF-I.124 These findings explain the previous contradiction between the lack of change in glomerular filtration rate and renal plasma flow during short-term infusions with growth hormone, and the increments in glomerular filtration rate and renal plasma flow found in patients with acromegaly and normal humans during sustained administration of growth hormone. It is difficult to separate the direct renal effects on electrolyte transport from the systemic anabolic effects of the hormone. Administration of growth hormone to subjects with ammonium chloride–induced acidosis increased urinary pH and net renal acid excretion, by increasing renal ammoniagenesis concurrently with increased sodium retention.121 Growth hormone enhances the reabsorption of phosphate in the proximal tubule.118,125 This effect may be mediated by IGF-I.117 The direct action of growth hormone to stimulate gluconeogenesis in the proximal tubule is not mediated by IGFs.126 An effect of growth hormone to increase renal calcitriol synthesis may also be mediated by IGF-I.127

ATRIAL NATRIURETIC HORMONE Since the initial description of atrial natriuretic factor,128 other similar peptides, secreted not only by the atria but also by the ventricles, brain, and kidneys, have been characterized.129 This group of peptides rapidly and reversibly induces marked diuresis, natriuresis, kaliuresis, and reduction in blood pressure and extracellular volume130,131 and 132 (see Chap. 178). Receptors for atrial natriuretic peptide (ANP) have been identified in the glomerulus and the inner medullary collecting duct.129 ANPs increase renal plasma flow and glomerular filtration rate. The latter is mediated by a decrease in the afferent and an increase in the efferent arteriolar resistance, although an increase in the glomerular capillary ultrafiltration coefficient is involved in the response in dehydrated animals.133 Simultaneous decreases in blood pressure and increases in glomerular filtration rate (in the absence of sustained changes in renal plasma flow or renal vascular resistance) cause a marked increase in

filtration fraction, and make it difficult to evaluate the direct tubular effects of the peptides in vivo. ANPs interfere with the renin–angiotensin–aldosterone system by decreasing the secretion of renin and aldosterone. Moreover, ANPs may antagonize the actions of such vasoconstrictors as angiotensin II, norepinephrine, and ADH.134 However, the action of ANPs may be offset in situations in which enhanced stimulation of the renin–angiotensin–aldosterone axis or the sympathetic nervous system occurs. Restoration of the natriuretic effects of ANPs may be seen in these settings after denervation or the administration of captopril.133 The mechanism of action of ANPs is unclear, although preferential vasoconstriction of the efferent arteriole, an increase in the glomerular ultrafiltration coefficient, a direct effect on tubular sodium and water transport, and redistribution of renal blood flow all are possible factors mediating the integrated response135,136 and 137 ANP exerts inhibitory effects on NaCl reabsorption in the cortical collecting duct, and on sodium, chloride, and water reabsorption in the medullary collecting ducts.138,139,140,141 and 142,142a ANPs inhibit the tubular reabsorption of calcium, phosphorus, and magnesium, and also interfere with water reabsorption mediated by ADH.143,144,145 and 146 The natriuretic effect probably is a combination of the increased glomerular filtration rate and changes in sodium reabsorption in the medullary collecting duct. These two mechanisms can be dissociated.133 Increased circulating levels of ANP may partially mediate aldosterone escape.15 Better understanding of the role of ANPs in sodium and fluid homeostasis may be achieved when specific antagonists become available for investigational use. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31.

White PC. Disorders of aldosterone biosynthesis and action. N Engl J Med 1994; 331:250. Marver D. Evidence of corticosteroid action along the nephron. Am J Physiol 1984; 246:F111. Funder J. Corticosteroid receptors and renal 11b-hydroxysteroid dehydrogenase activity. Semin Nephrol 1990; 10:311. Funder J. Mineralocorticoids, glucocorticoids, receptors and response elements. Science 1993; 259:1132. Whorwood CB, Sheppard MC, Stewart PM. Licorice inhibits 11-betahydroxysteroid dehydrogenase messenger ribonucleic acid levels and potentiates glucocorticoid hormone action. Endocrinology 1993; 132:2287. Wehling M. Specific, nongenomic actions of steroid hormones. Annu Rev Physiol 1997; 59:365. Gary H. Regulation of Na+ permeability by aldosterone. Semin Nephrol 1992; 12:24. Palmer LG, Frindt G. Regulation of apical membrane Na+ and K+ channels in rat renal collecting tubules by aldosterone. Semin Nephrol 1992; 12:37. O'Neil RG. Aldosterone regulation of Na+ and K+ transport in cortical collecting duct. Semin Nephrol 1990; 10:365. Marver D. Regulation of Na+/K+-ATPase by aldosterone. Semin Nephrol 1992; 12:56. Kokko JP. Primary acquired hypoaldosteronism. Kidney Int 1985; 27:690. Fyfe GK, Quinn A, Canessa CM. Structure and function of the Mec-ENaC family of ion channels. Semin Nephrol 1998; 18:138. Garty H, Palmer LG. Epithelial sodium channels: function, structure and regulation. Physiol Rev 1997; 77:359. Fanestil DD, Chun PS. Steroid hormones and the kidney. Annu Rev Physiol 1981; 43:637. Haas JA, Knox FG. Mechanisms for escape from the salt-retaining effects of mineralocorticoids: role of deep nephrons. Semin Nephrol 1990; 10:380. Stein JH. Hormones and the kidney. Hosp Pract (Off Ed) 1979; 14:91. Giebisch G. Renal potassium transport: mechanisms and regulation. Am J Physiol 1998; 274:F817. Field MJ, Giebisch GJ. Hormonal control of renal potassium excretion. Kidney Int 1985; 27:379. Stone DK, Crider BP, Xie X-S. Aldosterone and urinary acidification. Semin Nephrol 1990; 10:375. Perez GO, Oster JR. Acid-base pathophysiology in endocrine diseases. Mineral Electrolyte Metab 1985; 11:192. Garcia-Austt J, Good D, Burg M, Knepper M. Deoxycorticosterone-stimulated bicarbonate secretion in rabbit cortical collecting ducts. Am J Physiol 1985; 249:F205. Ichikawa I, Kon V. Hormonal regulation of glomerular filtration. Annu Rev Med 1985; 36:515. Welbourne TC, Givens G, Joshi S. Renal ammoniagenic response to chronic acid loading: role of glucocorticoids. Am J Physiol 1988; 254:F134. Ambuhl PM, Yang X, Peng Y, et al. Glucocorticoids enhance acid activation of the Na/H exchanger 3 (NHE3). J Clin Invest 1999; 103:429. Sutton RAL, Dirks JH. Calcium and magnesium: renal handling and disorders of metabolism. In: Brenner BM, Rector FC, eds. The kidney. Philadelphia: WB Saunders, 1986:551. Carmichael MC, Kumar R. Molecular biology of vasopressin receptors. Semin Nephrol 1994; 14:341. Chabardes D, Gagnan-Brunette M, Imbert-Teboul M, et al. Adenylate cyclase responsiveness to hormones in various portions of the human nephron. J Clin Invest 1980; 65:439. Abramow M, Beauwens R, Cogan E. Cellular events in vasopressin action. Kidney Int 1987; 32(Suppl 21):S56. Handler JS, Orloff J. Antidiuretic hormone. Annu Rev Physiol 1981; 43:611. Ichikawa I, Brenner BM. Evidence for glomerular actions of ADH and dibutyryl cyclic AMP in the rat. Am J Physiol 1977; 233:F102. Zeidel ML. Recent advances in water transport. Semin Nephrol 1998; 18:167.

31a. Klussmann E, Maric K, Rosenthal W. The mechanisms of aquaporin control in the renal collecting duct. Rev Physiol Biochem Pharmacol 2000; 141:33. 32. Verkman AS. Mechanisms and regulation of water permeability in renal epithelia. Am J Physiol 1989; 257:C837. 33. Knepper MA. Molecular physiology of urinary concentrating mechanism: regulation of aquaporin water channels by vasopressin. Am J Physiol 1997; 272:F3. 34. Nielsen S, Fror J, Knepper MA. Renal aquaporins: key roles in water balance and water balance disorders. Curr Opin Nephrol Hypertens 1998; 7:509. 34a. Gustaffson CE, Katsura T, McKee M, et al. Recycling of AQP2 occurs through a temperature- and bafilomycin-sensitive trans-Golgi-associated compartment. Am J Physiol Renal Physiol 2000; 278:F317. 35. Frokiaer J, Marples D, Knepper MA, Nielsen S. Pathophysiology of aquaporin-2 in water balance disorders. Am J Med Sci 1998; 316:291. 36. Scheinman SJ, Guay-Woodford LM, Thakker RV, Warnock DG. Genetic disorders of renal electrolyte transport. N Engl J Med 1999; 340:1177. 37. Earm JH, Christensen BM, Frokiaer J, et al. Decreased aquaporin-2 expression and apical plasma membrane delivery in kidney collecting ducts of polyuric hypercalcemic rats. J Am Soc Nephrol 1998; 9:2181. 38. Marples D, Frokiaer J, Dorup J, et al. Hypokalemia-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla and cortex. J Clin Invest 1996; 97:1960. 39. Elalouf JM, Roinel N, DeRouffignac C. ADH-like effects of calcitonin on electrolyte transport by Henles loop of rat kidney. Am J Physiol 1984; 246:F213. 40. Holt WF, Lechene C. ADH-PGE2 interactions in cortical collecting tubule. I. Depression of sodium transport. Am J Physiol 1981; 241:F452. 41. Kimmel PL, Goldfarb S. Effects of isoproterenol on potassium secretion by the cortical collecting tubule. Am J Physiol 1984; 246:F804. 42. Holt WF, Lechene C. ADH-PGE2 interactions in cortical collecting tubule. II. Inhibition of Ca and P reabsorption. Am J Physiol 1981; 241:F461. 43. You G, Smith CP, Kanai Y, et al. Cloning and characterization of the vasopressin-regulated urea transporter. Nature 1993; 365:844. 44. Sands JM. Regulation of renal urea transporters. J Am Soc Nephrol 1999; 10:635. 45. Terris J, Ecelbarger CA, Sands JM, Knepper MA. Long-term regulation of renal urea transporter protein expression in rat. J Am Soc Nephrol 1998; 9:729. 46. Juppner H, Abou-Samra AB, Freeman M, et al. A G protein-linked receptor for parathyroid hormone and parathyroid-hormone related peptide. Science 1991; 254:1024. 47. Bringhurst FR, Juppner H, Guo J, et al. Cloned, stably expressed parathyroid hormone/PTH related peptide receptors activate multiple messenger signals and biological responses in LLC-PK1 kidney cells. Endocrinology 1993; 132:2090. 48. Schneider H, Feyen JH, Seuwen K, Movva NR. Cloning and functional expression of a human parathyroid hormone receptor. Eur J Pharmacol 1993; 246:149. 49. Riccardi D, Lee WS, Lee K, et al. Localization of the extracellular Ca(2+)-sensing receptor and PTH/PTHrP receptor in rat kidney. Am J Physiol 1996; 271:F951. 50. Yang WT, Hassan S, Hyang YG, et al. Expression of PTHrP, PTH/PTHrP receptor and Ca2+-sensing receptor mRNAs along the rat nephron. Am J Physiol 1997; 272:F751. 51. Massry SG, Smogorzewski M. The mechanisms responsible for the PTH-induced rise in cytosolic calcium in various cells are not uniform. Mineral Electrolyte Metab 1995; 21:13. 52. Friedman PA, Coutermarsh BA, Kennedy SM, Gesek FA. Parathyroid hormone stimulation of calcium transport is mediated by dual signaling mechanisms involving protein kinase A and protein kinase C. Endocrinology 1996; 137:13. 53. Ichikawa I, Humes HD, Dousa TP, Brenner BM. Influence of parathyroid hormone on glomerular ultrafiltration in the rat. Am J Physiol 1978; 234:F393. 54. Friedman PA. Basal and hormone-activated calcium absorption in mouse renal thick ascending limbs. Am J Physiol 1988; 254:F62. 55. Friedman PA, Gesek F. Cellular calcium transport in renal epithelia: measurements, mechanisms and regulation. Physiol Rev 1995; 75:429. 56. Lau K, Bourdeau JE. Parathyroid hormone action in calcium transport in the distal nephron. Curr Opin Nephrol Hypertens 1995; 4:55. 57. Friedlander G. Autocrine/paracrine control of renal phosphate transport. Kidney Int 1998; 53:S65:S18. 58. Agus ZS, Wasserstein A, Goldfarb S. PTH, calcitonin, cyclic nucleotides and the kidney. Annu Rev Physiol 1981; 43:583. 59. Murer H, Lotscher M, Kaissling B, et al. Renal brush border membrane Na/Pi-cotransport: molecular aspects in PTH-dependent and dietary regulation. Kidney Int 1996; 49:1769. 60. Kempson SA, Lotscher M, Kaissling B, et al. Parathyroid hormone action on phosphate transporter mRNA and protein in rat renal proximal tubules. Am J Physiol 1995; 268:F784. 61. Kurokawa K. Cellular mechanisms and sites of hormone action in the kidney. In: Seldin DW, Giebisch G, eds. The kidney. New York: Raven Press, 1985:739. 62. Yang W, Friedman PA, Kumar R, et al. Expression of 25(OH)D3 and cAMP or 24-hydroxylase in distal nephron: coordinate regulation by 1,25(OH)2+D3 PTH. Am J Physiol 1999; 276:E793. 63. Stim JA, Bernardo AA, Arruda JA. The role of PTH and vitamin D in acid excretion and extrarenal buffer mobilization. Mineral Electrolyte Metab 1994; 20:60. 64. Hulter HN. Effects and interrelationships of PTH, Ca 2+, vitamin D, and Pi in acid-base homeostasis. Am J Physiol 1985; 248:F739. 65. Sexton PM, Adam WR, Mosely JM, et al. Localization and characterization of renal calcitonin receptors by in vitro autoradiography. Kidney Int 1987; 32:862. 66. Firsov D, Bellanger AC, Marsy S, Elaouf JM. Quantitative RT-PCR analysis of calcitonin mRNAs in the rat nephron. Am J Physiol 1995; 269:F702. 67. Morel F. Sites of hormone action in the mammalian nephron. Am J Physiol 1981; 240:F159. 68. Friedman PA, Gesek FA. Hormone-responsive Ca 2+ entry in distal convoluted tubules. J Am Soc Nephrol 1994; 4:1396. 69. Nakamura R, Emmanuel DS, Katz AI. Insulin binding sites in various segments of the rabbit nephron. J Clin Invest 1983; 72:388. 70. Nakamura R, Hayashi M, Emmanuel DS, Katz AI. Sites of insulin and glucagon metabolism in the rabbit nephron. Am J Physiol 1986; 250:F144. 71. Gupta AK, Clark RV, Kirchner KA. Effects of insulin on renal sodium excretion. Hypertension 1992; 19SI:I78. 72. Stenvinkel P, Bolinder J, Alvestrand A. Effects of insulin on renal hemodynamics and the proximal and distal tubular sodium handling in healthy subjects. Diabetologia 1992; 35:1042. 73. Friedberg CE, van Buren M, Bijlsma JA, Koomans HA. Insulin increases sodium reabsorption in diluting segment in humans: evidence for indirect mediation through hypokalemia. Kidney Int 1991; 40:251. 74. Hammerman M. Interaction of insulin with the renal proximal tubular cell. Am J Physiol 1985; 249:F1. 75. Rodriguez-Commes J, Isales C, Kalghati L, et al. Mechanism of insulin-stimulated electrogenic sodium transport. Kidney Int 1994; 46:666. 76. Blazer-Yost BL, Liu X, Helman SI. Hormonal regulation of EnaC: insulin and aldosterone. Am J Physiol 1998; 274:C1373. 77. Marunaka Y, Hagiwara N, Tohda H. Insulin activates single amiloride-blockable Na + channels in a distal nephron cell line (A6). Am J Physiol 1992; 263:F392. 78. Smith D, DeFronzo RA. Insulin, glucagon and thyroid hormone. In: Dunn MJ, ed. Renal endocrinology. Baltimore: Williams & Wilkins, 1983:367. 79. Brosius FC III, Briggs JP, Marcus RG, et al. Insulin-responsive glucose transporter expression in renal microvessels and glomeruli. Kidney Int 1992; 42:1086. 80. Quinones-Galvan A, Natali A, Baldi S, et al. Effect of insulin on uric acid excretion in humans. Am J Physiol 1995; 268:E1. 81. Morel F, Imbert-Teboul M, Chabardes D. Distribution of hormone-dependent adenylate cyclase in the nephron and its physiological significance. Annu Rev Physiol 1981; 43:569. 82. MacNeil DJ, Occi JL, Hey PJ, et al. Cloning and expression of a human glucagon receptor. Biochem Biophys Res Commun 1994; 198:328. 83. Dunphy JL, Taylor RG, Fuller PJ. Tissue distribution of rat glucagon receptor and GLP-1 receptor gene expression. Mol Cell Endocrinol 1998; 141:179. 84. Hansen LH, Abrahamsen N, Nishimura E. Glucagon receptor mRNA distribution in rat tissues. Peptides 1995; 16:1163.

85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142.

Burcelin R, Katz EB, Charron MJ. Molecular and cellular aspects of the glucagon receptor. Diabetes Metab 1996; 22:373. Premen AJ, Hall JE, Smith MJ Jr. Postprandial regulation of renal hemodynamics: role of pancreatic glucagon. Am J Physiol 1985; 248:F656. Bailly C, Roinel N, Amiel C. PTH-like glucagon stimulation of Ca and Mg reabsorption in Henles loop of the rat. Am J Physiol 1984; 246:F205. Hagenfeldt Y, Eriksson HA. The estrogen receptor in the rat kidney. Ontogeny, properties and effects of gonadectomy on its concentration. J Steroid Biochem 1988; 31:49. Davidoof M, Caffier H, Schiebler T. Steroid hormone binding receptors in the rat kidney. Histochemistry 1980; 69:39. Han HJ, Jung JC, Taub M. Response of primary rabbit kidney proximal tubule cells to estrogens. J Cell Physiol 1999; 178:35. Kuiper GG, Carlsson B, Grandien K, et al. Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors alpha and beta. Endocrinology 1997; 138:863. Silbiger SR, Neugarten J. The impact of gender on the progression of chronic renal disease. Am J Kidney Dis 1995; 25:515. Kwan G, Neugarten J, Sherman M, et al. Effects of sex hormones on mesangial cell proliferation and collagen synthesis. Kidney Int 1996; 50:1173. Lei J, Sibiger S, Ziyadeh F, Neugarten J. Serum-stimulated alpha I type IV collagen gene transcription is mediated by TGF-b and inhibited by estradiol. Am J Physiol 1998; 24:F252. Neugarten J, Ding Q, Friedman A, et al. Sex hormones and renal nitric oxide synthesis. J Am Soc Nephrol 1997; 8:1240. Preedy JRK, Aitken EH. The effects of estrogen on water and electrolyte metabolism. J Clin Invest 1956; 35:423. Johnson JA, Davis JO, Baumber S, Schineider EG. Effects of estrogen and progesterone on electrolytic balances in normal dogs. Am J Physiol 1970; 219:1691. Dignam WS, Voskian J, Assali NS. Effects of estrogens on renal hemodynamics and excretion of electrolytes in human subjects. J Clin Endocrinol 1956; 16:1032. Ferris TF, Francisco LL. Estrogen, progesterone and the kidney. In: Dunn MJ, ed. Renal endocrinology. Baltimore: Williams & Wilkins, 1983:462. Vegeto E, Shahbaz MM, Wen DX, et al. Human progesterone receptor A form is a cell- and promoter-specific repressor of human progesterone receptor B function. Mol Endocrinol 1993; 7:1241. Elkarib AO, Garland HO, Green R. Acute and chronic effects of progesterone and prolactin on renal function in the rat. J Physiol 1983; 337:389. Beers KW, Dousa P. Thyroid hormone stimulates the Na +-PO4 symporter but not the Na+-SO4 in renal brush border. Am J Physiol 1993; 265:F2323. Alcalde AI, Sarasa M, Raldua D, et al. Role of thyroid hormone in regulation of renal phosphate transport in young and aged rats. Endocrinology 1999; 140:1544. Katz AI, Lindheimer MD. Actions of hormones on the kidney. Annu Rev Physiol 1977; 39:97. Berl T, Better OS. Renal effects of prolactin, estrogen, and progesterone. In: Brenner BM, Stein JH, eds. Hormonal function and the kidney, contemporary issues in nephrology, vol 4. New York: Churchill Livingstone, 1979:194. Stier CT Jr, Cowden EA, Friesen HG, Allison MEM. Prolactin and the rat kidney: a clearance and micropuncture study. Endocrinology 1984; 115:362. Costanzo LS, Adler RA. Chronic prolactin excess causes hypercalciuria: a direct renal effect. Kidney Int 1986; 29:157. Shrier RW. Effects of adrenergic nervous system and catecholamines on systemic and renal hemodynamics, sodium and water excretion and renin secretion. Kidney Int 1974; 6:291. Munzel PA, Healy DP, Insel PA. Autoradiographic localization of b-adrenergic receptors in rat kidney slices using [ 125 I]-iodocyanopindolol. Am J Physiol 1984; 246:F240. Dworkin LD, Ichikawa I, Brenner BM. Hormonal modulation of glomerular function. Am J Physiol 1983; 244:F95. Besarab A, Silva P, Landsberg L, Epstein F. Effects of catecholamines on tubular function in the isolated perfused rat kidney. Am J Physiol 1977; 233:F39. Iino Y, Troy JL, Brenner BM. Effects of catecholamines on electrolyte transport in cortical collecting tubules. J Membr Biol 1981; 61:67. Krothapalli RK, Duffy BW, Senekjian HO, Suki WN. Modulation of the hydroosmotic effect of vasopressin on the rabbit cortical collecting tubule by adrenergic agents. J Clin Invest 1983; 72:287. Umemura S, Marver D, Smyth D, Pettinger W. a2-Adrenoceptors and cellular cAMP levels in single nephron segments from the rat. Am J Physiol 1985; 249:F28. Gesek FA. Alpha 2 adrenergic receptors activate phospholipase C in renal epithelial cells. Mol Pharmacol 1996; 50:407. DiBona GF. Catecholamines and neuroadrenergic control of renal function. In: Dunn MJ, ed. Renal endocrinology. Baltimore: Williams & Wilkins, 1983:323. Hammerman MR. The growth hormone-insulin-like growth factor axis in kidney. Am J Physiol 1989; 257:F503. Feld S, Hirschberg R. Growth hormone, the insulin-like growth factor system, and the kidney. Endocr Rev 1996; 17:423. Hirschberg R, Adler S. Insulin-like growth factor system and the kidney: physiology, pathophysiology, and therapeutic implications. Am J Kidney Dis 1998; 31:901. Hammerman MR, Miller SB. Effects of growth hormone and the insulin-like growth factor on renal growth and function. J Pediatr 1997; 131:S17. Sicuro A, Mahlbacher K, Hulter HN, Krapf R. Effects of growth hormone and systemic acid-base homeostasis in humans. Am J Physiol 1998; 24:F650. Parving HH, Noer I, Mogensen CE, Svendsen PA. Kidney function in normal man during short-term growth hormone infusion. Acta Endocrinol (Copenh) 1978; 89:796. Westby GR, Goldfarb S, Goldberg M, Agus ZS. Acute effects of bovine growth hormone on renal calcium and phosphate excretion. Metabolism 1977; 26:525. Hirschberg R, Rabb H, Bergamo R, Kopple JD. The delayed effect of growth hormone on renal function in humans. Kidney Int 1989; 35:865. Hammerman MR, Karl IE, Hruska KA. Regulation of canine renal vesicle Pi transport by growth hormone and parathyroid hormone. Biochim Biophys Acta 1980; 603:322. Rogers SA, Hammerman MR. Growth hormone directly stimulates gluconeogenesis in canine renal proximal tubule. Am J Physiol 1989; 257:E751. Wei S, Tanaka H, Seino Y. Local action of exogenous growth hormone and insulin-like growth factor-I on dihydroxyvitamin D production in LLC-PK1 cells. Eur J Endocrinol 1998; 139:454. de Bold AJ, Borenstein HB, Veress AT, Sonnenberg H. A rapid and potent natriuretic response to intravenous injection of atrial myocardial extracts in rats. Life Sci 1981; 28:89. Gunning ME, Brenner BM. Natriuretic peptides and the kidney: current concepts. Kidney Int 1992; 42(Suppl 38):S-127. Espiner EA. Physiology of natriuretic peptides. J Intern Med 1994; 235:527. Beland B, Tuchelt H, Bahr V, Oelkers W. The role of atrial natriuretic factor (alpha-human ANF-[99-126]) in the hormonal and renal adaptation to sodium deficiency. J Clin Endocrinol Metab 1994; 79:183. Nicholls MG. The natriuretic peptide hormones. (Editorial and historical review). J Intern Med 1994; 235:507. Awazu M, Ichikawa I. Biological significance of atrial natriuretic peptide in the kidney. Nephron 1993; 63:1. Laragh JH. Atrial natriuretic hormone, the renin-aldosterone axis, and blood pressure-electrolyte homeostasis. N Engl J Med 1985; 313:1330. Maack T, Camargo MJD, Kleinert HD, et al. Atrial natriuretic factor: structure and functional properties. Kidney Int 1985; 27:607. Goetz KL. Physiology and pathophysiology of atrial peptides. Am J Physiol 1988; 254:E1. Zeidel ML, Brenner BM. Action of atrial natriuretic peptides on the kidney. Semin Nephrol 1987; 7:91. Butlen D, Mistaoui M, Morel F. Atrial natriuretic peptide receptors along the rat and rabbit nephrons: [ 125 I] rat atrial natriuretic peptide binding in micro dissected glomeruli and segments of the rat and rabbit nephrons. Pflugers Arch 1987; 408:366. Nonoguchi H, Sands JM, Knepper MA. ANF inhibits NaCl and fluid absorption in cortical collecting duct of rat kidney. Am J Physiol 1989; 256:F179. Sonnenberg H, Honrath V, Chong CK, Wilson DR. Atrial natriuretic factor inhibits sodium transport in medullary collecting duct. Am J Physiol 1986; 250:F963. Van de Stolpe A, Jamison RL. Micropuncture study of the effect of ANP on the papillary collecting duct in the rat. Am J Physiol 1988; 254:F477. Ziedel ML. Medullary collecting duct sodium transport. Am J Physiol 1993; 295:F159.

142a. Holtback U, Kruse MS, Brismar H, Aperia A. Intrarenal dopamine coordinates the effect of antinatriuretic and natriuretic factors. Acta Physiol Scand 2000; 168:215. 143. 144. 145. 146.

Ortola FV, Ballermann BJ, Brenner BM. Endogenous ANP augments fractional excretion of Pi, Ca and Na in rats with reduced renal mass. Am J Physiol 1988; 255:F1091. Dunn BR, Ichikawa I, Pfeffer JM, et al. Renal and systemic hemodynamic effects of synthetic atrial natriuretic peptide in the anesthetized rat. Circ Res 1986; 59:237. Pollock DM, Arendshorst WJ. Effects of atrial natriuretic factor on renal hemodynamics in the rat. Am J Physiol 1986; 251:F795. Dillingham MA, Anderson RJ. Inhibition of vasopressin action by atrial natriuretic factor. Science 1986; 231:1572.

CHAPTER 207 RENAL METABOLISM OF HORMONES Principles and Practice of Endocrinology and Metabolism

CHAPTER 207 RENAL METABOLISM OF HORMONES RALPH RABKIN AND MICHAEL J. HAUSMANN Metabolism of Peptide Hormones Removal of Peptide Hormones by Glomerular Filtration Removal of Peptide Hormones from the Postglomerular Peritubular Circulation Metabolism of the Major Peptide Hormones Growth Hormone Prolactin Arginine Vasopressin Parathyroid Hormone Calcitonin Insulin Insulin-Like Growth Factors Glucagon Atrial Natriuretic Hormone Glycoprotein Hormones Thyroid Hormones Steroid Hormones Glucocorticoids Aldosterone Chapter References

Three patterns of interaction occur among the kidney and hormones. The kidney may be the site of action, of production, or of degradation of hormones. The last interaction is the one discussed in this chapter. For peptide hormones, the kidney is a key site of metabolism and, together with the liver, accounts for most of their destruction.1,2,3,4,5 and 6 A few hormones, however, including growth hormone (GH), calcitonin, and C peptide, are handled primarily by the kidney. The kidney plays a minor role in the metabolism of steroid hormones; these hormones circulate bound to large proteins and are not readily available for glomerular filtration. Hence, relatively small amounts are filtered, degraded by the kidney, or excreted in the urine. The importance of the kidney in hormone metabolism becomes apparent in renal failure. Under these circumstances, the metabolic clearance rate (MCR) of hormones metabolized by the kidney is prolonged, and this may play a role in the pathogenesis of some of the endocrine manifestations associated with uremia. The situation may be compounded by uremic depression of extrarenal sites of hormone degradation and also by changes in hormone secretion. Prohormones and hormone metabolites may accumulate. Because commonly used hormone assays frequently do not distinguish between the bioactive and inactive forms of hormone, this potential pitfall must be considered when interpreting measured hormone levels in renal failure (Fig. 207-1). The following discussion regarding the physiology of hormone metabolism refers to data obtained in humans whenever possible.

FIGURE 207-1. Gel filtration patterns of plasma from two patients with renal failure and elevated immunoreactive calcitonin levels. Samples of 2 mL were applied to a Bio Gel P-10 (1.5 ×100 cm) column, and 1-mL fractions were eluted with 0.2 M phosphate, pH = 7.4. Arrow indicates elution position of labeled monomer. Heterogeneity of molecular species is evident. (125I-HCT, iodine-125 human calcitonin.) (From Lee JC, Parthmore JG, Deftos LJ. Immunochemical heterogeneity of calcitonin in renal failure. J Clin Endocrinol Metab 1977; 45:528.)

METABOLISM OF PEPTIDE HORMONES The kidney extracts 16% to 45% of the various peptide hormones in the renal circulation. Occasionally, if the hormone is sensitive to enzymes in glomeruli and vascular endothelial cells (e.g., bradykinin, atrial natriuretic hormone), extraction may be as high as 90%.4 Depending on the contribution of extrarenal sites, the kidney may account for 30% to 80% of the total metabolism of a hormone (Table 207-1). An important feature of the renal removal process is that, unlike in the liver, saturation of uptake is difficult to achieve.1 This largely reflects the dominant role of glomerular filtration in clearing hormones. Consequently, the kidney extracts peptide hormones from a constant volume of plasma per unit time, and the proportion of plasma hormone removed by the kidney remains constant through a broad range of plasma hormone concentrations. Thus, this system tends to maintain plasma hormone concentrations at basal levels. After a secretory stimulus and rise in plasma hormone concentration, the absolute amount of hormone removed increases proportionately; with return to the unstimulated state, the amount of hormone removed falls. This is not a true feed-back regulatory mechanism, however. In severe renal failure, this system fails because renal blood flow is reduced, thereby decreasing delivery of hormone to the kidney. In addition, the kidney loses its ability to extract the hormone (Fig. 207-2).

TABLE 207-1. Contribution of the Kidney to the Metabolism of Polypeptide and Glycoprotein Hormones

FIGURE 207-2. Relationship between plasma arterial (ART.) insulin concentration and renal arteriovenous (A-V) insulin difference in patients with advanced chronic renal failure. Under these circumstances, the kidney loses its ability to extract insulin. The regression line represents the relationship obtained in normal kidneys. The closed circles and triangles represent results for two diabetic patients, and the open symbols those for two nondiabetic subjects. (From Rabkin R, Simon NM, Steiner S, Colwell JA. Effect of renal disease on renal uptake and excretion of insulin in man. N Engl J Med 1970; 282:182.)

Peptide hormones are removed from the renal circulation by two major pathways: glomerular filtration and extraction from the postglomerular peritubular circulation (Fig. 207-3).

FIGURE 207-3. Pathways of peptide hormone degradation. A, Local degradation by the glomerulus represents a minor pathway for small hormones, such as angiotensin, bradykinin, and possibly calcitonin. B, Glomerular filtration and tubular reabsorption is the major pathway for all peptide hormones. Small linear peptides are hydrolyzed by the luminal membrane, but complex peptides require internalization before degradation. C, Peritubular removal occurs by receptor-mediated and non–receptor-mediated mechanisms. Receptor binding initiates hormone action, but its relationship to degradation is unknown. Degradation may be membrane associated or may require internalization. Peritubular removal probably occurs along the length of the nephron. (Modified from Rabkin R, Glaser T, Petersen J. Renal peptide hormone metabolism. Kidney 1983; 16:25.)

REMOVAL OF PEPTIDE HORMONES BY GLOMERULAR FILTRATION Glomerular filtration serves as the predominant removal route for bioactive and inactive forms of peptide hormones.3 Factors such as the size, conformation, and charge of the molecule affect the filtration rate. Thus, polymerization or protein binding, as occurs with insulin-like growth factor I (IGF-I), reduces filtration. Large polypeptide hormones (e.g., GH) are filtered slowly, at ~70% the rate of freely filtered molecules, and small peptides (e.g., angiotensin) probably pass through the glomerular filtration barrier without hindrance. The local destruction of hormone by the glomerulus is trivial for complex peptides such as parathyroid hormone (PTH). For less complex hormones such as angiotensin, bradykinin, and calcitonin, however, local destruction may account for a minor, but significant, portion of their metabolism.5 Specific receptors for many peptide hormones have been described in the glomerulus. The importance of these receptors is related to their roles in mediating the glomerular actions of their respective hormones. Accordingly, peptide hormones such as angiotensin, atrial natriuretic hormone, IGF-I, PTH, and vasopressin appear to modulate the glomerular filtration rate, and hormones such as insulin have important metabolic effects on the glomerulus (see Chap. 206). After passage through the glomerular filtration barrier, peptide hormones enter the proximal tubular lumen, where, depending on their structure, they are hydrolyzed on contact with the luminal membrane or are internalized by the tubular cell (see Fig. 207-3). Both of these processes are highly efficient, because only a small percentage of the filtered load is excreted in the urine. Although the percentage excreted tends to be constant over physiologic plasma hormone levels, it rises in some instances. For example, at physiologic plasma levels, urinary arginine vasopressin (AVP) excretion represents 6% of the filtered load, but at higher levels, it reaches 24%.6 This must be considered when measurements of urinary peptide hormone excretion are used as an indirect measure of endogenous hormone secretion. An even greater pitfall in the use of urine excretion rates is the fact that tubular dysfunction is associated with impaired tubular uptake of filtered peptide hormones.1 The biologic importance of proximal tubular destruction is that it serves as a means of conserving the constituent amino acids and of deactivating hormones such as angiotensin, atrial natriuretic hormone, and bradykinin, which may act from the luminal aspect of the tubules. This may prevent the unregulated action of filtered hormones on more distal nephron segments. Small linear peptides (e.g., angiotensin and bradykinin) are hydrolyzed within the proximal tubules by the peptidase-rich brush border membrane of the proximal tubule, and the metabolic products, predominantly amino acids, then are reabsorbed.7 As is the case for filtered proteins such as albumin, large, complex hormones (e.g., insulin, PTH, and GH) undergo relatively little or no brush border membrane hydrolysis. Internalization by the proximal tubular cell is a prerequisite for the degradation of these hormones.1,3,8 Internalization occurs by endocytosis in the proximal renal tubule. This process is initiated by binding of the filtered hormone to the endocytic plasma membrane receptor megalin, originally named gp330.8,9 To a much lesser extent endocytosis may be initiated by a charge-related interaction. In selected instances, as occurs with insulin, the hormone also may bind to specific sites in the brush border membrane.10 The bound hormone is transported by endocytosis into the proximal tubular cell interior and delivered to the lysosomes. Although some degradation may occur in endosomes, most occurs in lysosomes. The released amino acids then are returned to the peritubular circulation. No evidence exists for significant transtubular transport of intact hormone. A small amount of internalized hormone, however, may be transported back to the cell surface in recycling endosomes to be released into the extracellular compartment, and certain hormones such as IGF-I are also transported to the nucleus.11 Although two major systems exist for degrading filtered peptide hormones—hydrolysis by brush border membranes and hydrolysis after internalization—these systems are not mutually exclusive and some overlap may occur. Degradation of some peptide hormones of intermediate complexity may involve both pathways, depending on their resistance to luminal hydrolysis. For example, luteinizing hormone–releasing hormone, a linear peptide with a C-terminal amide and an N-terminal pyroglutamyl residue, is incompletely degraded by the brush border membrane. Resultant peptide fragments and free amino acids, and perhaps some intact hormone, are absorbed, with further hydrolysis occurring within the cell. The liberated amino acids and some small peptide fragments then are released into the interstitial compartment. Calcitonin may be another hormone handled by both pathways. REMOVAL OF PEPTIDE HORMONES FROM THE POSTGLOMERULAR PERITUBULAR CIRCULATION The kidney also removes peptide hormones from the postglomerular peritubular circulation, which is followed by binding of hormone to specific receptors in the basolateral membrane and by degradation.1 Hormones such as insulin, PTH, calcitonin, vasopressin, and angiotensin are delivered to their receptors and initiate their actions through the peritubular route. The contribution of this pathway to the catabolism of each hormone varies. For example, estimates are that peritubular removal may account for ~40% of the total insulin extracted by the human kidney, but peritubular removal of GH is minimal.1,12 For complex peptide hormones such as insulin, peritubular removal involves receptor-mediated endocytosis with intracellular degradation. For smaller, less complex peptide hormones, it is unclear whether receptor binding is a prerequisite for peritubular hormone degradation, whether degradation is membrane associated, and whether degradation follows internalization. Unlike the luminal aspect of the proximal tubular cell, the antiluminal aspect has a poorly developed endocytotic system. Significantly, degradation of hormone may not

necessarily be complete; peritubular PTH and insulin catabolism is associated with the release of large fragments of unknown bioactivity along with the formation of products of complete degradation.3,10

METABOLISM OF THE MAJOR PEPTIDE HORMONES GROWTH HORMONE The kidney is the predominant site of GH metabolism. In rats, the kidneys account for 65% to 70% of the total MCR1; in humans, the kidneys are estimated to account for 25% to 53% of the total MCR. 12 The human data indicate that, when plasma GH levels are elevated, saturation of the extrarenal clearance pathways occurs, resulting in a greater contribution of the kidneys to the MCR. Renal clearance is achieved primarily by the glomerular filtration and proximal tubule degradative route; peritubular removal does occur, but it is a minor process. Some restriction of GH through the glomerular filtration barrier occurs, and indirect estimations suggest that the rate of GH filtration is 70% that of water. Restriction is partly the result of the complexing of GH to GH-binding proteins (GHBPs).13 The principal GHBP in humans is derived from proteolytic cleavage of the full-length GH receptor and is identical to the extracellular domain of the receptor. Serum GHBP levels are, therefore, believed to reflect GH-receptor number and usually are low in advanced renal failure. Indeed GH resistance in uremia may be caused partly by a fall in receptor number.13,14 Microperfusion and autoradiographic studies suggest that the filtered hormone is absorbed in the proximal tubule. This absorptive process is highly efficient, because 25,000 Da) and complex structure of glycoproteins. Furthermore, the absence of glycoprotein hormone–specific receptors excludes receptor-mediated removal. The glycoprotein hormones have a relatively high urinary excretion rate (see Table 207-1). For example, the urinary excretion of FSH averages 43% of the total FSH renal clearance, but the urinary excretion of nonglycosylated hormones is less than 1% to 2% of the filtered load.1 This indicates that the tubular absorption of filtered glycoproteins is less efficient. Despite the low renal clearance rate of glycoprotein hormones, the renal contribution to the total body MCR of glycoprotein hormones is

relatively large. This reflects the slower total MCR of glycoprotein hormones. In humans, urinary excretion accounts for ~5% and 25% of the total MCR of LH and human chorionic gonadotropin, respectively. This does not take into account the contribution of intrarenal degradation. In more detailed studies in the rat (see Table 207-1), the indirect estimates of the renal contribution to the MCR of LH, FSH, and erythropoietin averaged 94%, 78%, and 23%, respectively (see Chap. 16).39,40 THYROID HORMONES The thyroid hormones are released from the thyroid gland as the iodinated amino acids, thyroxine (3,5,3',5'-tetraiodothyronine; T4) and, to a minor extent, 3,5,3'-triiodothyronine (T3) (see Chap. 30). Like steroid hormones, they are transported in the plasma almost entirely in association with binding proteins. The metabolically active unbound T4 and T3 constitute 0.03% and 0.3% of total hormone, respectively. Most of the circulating T3 is derived from the peripheral conversion of T4. Other peripheral conversion products include the inactive 3,3',5'-triiodothyronine and 3,3'-diiodothyronine. The liver and kidneys are the major sites of conversion. The enzyme involved in this process, iodothyronine 5'-deiodinase, has been identified in microsomes.41,42 Both T4 and T3 are metabolized by the kidney and other organs through deamination and decarboxylation to form tetraiodothyroacetic acid and triiodothyroacetic acid. Approximately half of daily thyroid iodothyronine production is disposed of as urinary thyronine or as the acetic acid analog of thyronine.43 Both of these products are completely deiodinated. More than half of infused T4, however, is excreted in the urine as diiodometabolites. Because thyroid hormones are largely protein bound, the fraction extracted and excreted by the kidney is small; arteriovenous concentration differences are not detectable. The unbound T4 and T3, however, are freely filtered and excreted in the urine. During fasting, both mean T4 and T3 urinary excretion is decreased.43a Based on urine excretion rates, estimates are that ~65% of the filtered T4 is reabsorbed by the tubules. Conversely, T3 is added to the urine; this additional T3 is derived from the intrarenal conversion of T4. The kidney also contributes to the excretion of the glucuronide and sulfate conjugates that arise from liver metabolism.44 In addition, significant peritubular removal of T4 occurs; the T4 is then converted to T3 by 5'-deiodinase located on the cytoplasmic face of the proximal tubular basolateral membrane.45,46 Renal disease is accompanied by alterations in thyroid hormone metabolism (see Chap. 209).47 STEROID HORMONES The major role of the kidney in steroid hormone metabolism is the elimination of metabolites. Metabolic processing of active hormone also occurs in the kidney, but this is a minor process. Conversely, the liver is the central site of steroid hormone inactivation and a minor site of elimination. These organ differences may be explained partly by the binding of circulating steroid hormones to albumin and certain globulins with a high affinity for specific steroids. Depending on the steroid, as much as 98% may be protein bound. Protein binding effectively restricts glomerular filtration. In contrast, metabolites circulate in free form and are readily eliminated by the kidney. The kidney filters circulating intact hormone that is not plasma protein bound, and also may remove intact hormone from the postglomerular peritubular circulation. Lipid-soluble steroids enter the tubular cell, where they interact with cytoplasmic receptors and undergo metabolic transformations. GLUCOCORTICOIDS As with other steroids, the key role of the kidney in glucocorticoid metabolism is the excretion of metabolic products from extrarenal conversion. The liver is the major site of inactivation, accounting for 90% of the metabolism. The injection of radiolabeled cortisol is followed by the excretion of 90% of the radioactivity over 3 days. Less than 1% of the excreted radioactivity is associated with intact cortisol. In humans and several other species, cortisol is the most abundant corticosteroid synthesized by the adrenals, but in rabbits and rodents, the major product is corticosterone. In humans, 90% to 95% of the circulating cortisol is bound to the corticosteroid-binding globulin transcortin, or to albumin (see Chap. 72). The unbound form is freely filtered, and the lipid-soluble hormone is readily absorbed by the renal tubules, especially in the distal nephron. Altogether, between 60% and 90% of the filtered cortisol may be absorbed, and urinary excretion is low. Besides being filtered by the glomeruli, corticosterone may be extracted from the postglomerular peritubular circulation in the rat.48 Although the kidney serves mainly as a site of elimination of hepatic metabolites, it also is the most important site of the conversion of cortisol to inactive cortisone.49 This is significant because mineralocorticoid receptors bind cortisol and aldosterone with equal affinity, and circulating plasma cortisol levels are 500 times greater than plasma aldosterone levels. Thus, inactivation of cortisol is essential for aldosterone to mediate its specific regulatory actions. Abnormalities in conversion can result in pronounced clinical manifestations. Normally, cortisol is converted to cortisone by 11b-hydroxysteroid dehydrogenase (11b-OHSD) expressed in two forms, 11b-OHSD1 and 11b-OHSD2.49,50 The isoenzyme 11b-OHSD1 is located in proximal tubular cells and, by inactivating filtered glucocorticoids, limits the exposure of tubular cells to glucocorticoids. Low activity of this enzyme has been implicated in the development of renal cysts in cpk mice with hereditary cystic disease.51 The isoenzyme 11b -OHSD2 is located in the distal nephron in aldosterone-sensitive collecting duct cells and allows aldosterone to act independently of glucocorticoids in the kidney.52 When this enzyme is congenitally absent (syndrome of apparent mineralocorticoid excess53) or inactivated (as with glycyrrhetinic acid, a component of licorice), cortisol acts on renal mineralocorticoid receptors to stimulate potassium secretion, sodium avidity, and hypertension. Unidentified endogenous inhibitors of 11b -OHSD, which have been isolated from urine, are known as glycyrrhetinic acid–like factors or GALF.54 Evidence also exists for the renal conversion of cortisol to 6b-hydroxycortisol by the enzyme 6b-hydroxylase. When enhanced, this pathway also may be responsible for the generation of hypertension.55,56 and 57 Finally, the renal conversion of cortisol to 20-dehydrocorticosterone, which can inhibit AVP-stimulated water transport in toad bladder, has been reported. ALDOSTERONE The major role of the kidney in aldosterone metabolism is the excretion of metabolites, which is achieved by glomerular filtration and tubular secretion. In humans, the kidney is one of the major extrahepatic sites for aldosterone metabolism.58,59 Aldosterone circulates less firmly bound to plasma protein than do glucocorticoids; hence, it is cleared more rapidly from the plasma than are glucocorticoids, and the renal extraction is more prominent. The kidney extracts 10% of the hormone passing through it, but the liver extracts 92%. Free aldosterone is readily filtered at the glomerulus, and between 80% and 95% is absorbed by the proximal and distal nephron. The absorption is passive and is modulated by sodium and water movement. Aldosterone also is removed from the peritubular circulation. Thirty minutes after intravenous injection of [3H]-aldosterone in rats, 40% and 2.4% of the dose is present in the liver and kidney, respectively. Most of the radioactivity is present in the cytosol. The metabolic transformation of the hormone is rapid, and many metabolites formed locally and in the liver are found in the kidney. Hepatic and renal aldosterone metabolism is sex dependent and affected by salt intake. Both the nature and the amount of kidney metabolites produced vary according to gender and the dietary sodium ingested. Most of the products of aldosterone metabolism in the kidney are polar-neutral hydroxylated compounds (nonconjugated) and nonpolar-reduced compounds. Moreover, sulfates and carboxylic acid metabolites also are formed. Studies in animals suggest that the nonpolar, ring A– reduced metabolites may be important in modulating or mediating the action of aldosterone.58 Several of the reduced metabolites of aldosterone, although less active than the parent hormone, appear to possess some bioactivity. Whatever their role in the action of the hormone, the ultimate fate of the metabolites is excretion into the urine or entry into the circulation. CHAPTER REFERENCES 1. Rabkin R, Dahl DC. Renal uptake and disposal of proteins and peptides. In: Raub TJ, Audus KL, eds. Pharmaceutical biotechnology, vol. 5. Biological barriers to protein delivery. New York: Plenum Press, 1993:299. 2. Ardaillou R, Paillard F. Metabolism of polypeptide hormones by the kidney. Adv Nephrol Necker Hosp 1980; 9:247. 3. Maack T, Johnson V, Kau ST, et al. Renal filtration, transport, and metabolism of low-molecular-weight proteins: a review. Kidney Int 1979; 16(3):251. 4. Nasjlette A, Colessa-Chorerio J, McGiff JC. Disappearance of bradykinin in the renal circulation of dogs. Effects of kininase inhibition. Circ Res 1975; 37:59. 5. Thaiss F, Wolf G, Assad N, et al. Angiotensinase A gene expression and enzyme activity in isolated glomeruli of diabetic rats. Diabetologia 1996; 39(3):275. 6. Pruszczynski W, Caillens H, Drieu L, et al. Renal excretion of antidiuretic hormone in healthy subjects and patients with renal failure. Clin Sci 1984; 67:307. 7. Carone FA, Peterson DR. Hydrolysis and transport of small peptides by the proximal tubule. Am J Physiol 1980; 238:F151. 8. Christensen EI, Birn H, Verroust P, Moestrup S. Membrane receptors for endocytosis in the renal proximal tubule. Int Rev Cytol 1998; 180:237. 9. Orlando RA, Rader K, Authier F, et al. Megalin is an endocytic receptor for insulin. J Am Soc Nephrol 1998; 9(10):1759. 10. Rabkin R, Yagil C, Frank B. Basolateral and apical binding, internalization, and degradation of insulin by cultured kidney epithelial cells. Am J Physiol 1989; 257:E895. 11. Li W, Fawcett J, Widmer HR, et al. Nuclear transport of insulin-like growth factor-I and insulin-like growth factor binding protein-3 in opossum kidney cells. Endocrinology 1997; 138(4):1763. 12. Haffner D, Schaefer F, Girard J, et al. Metabolic clearance of recombinant human growth hormone in health and chronic renal failure. J Clin Invest 1994; 93:1163. 13. Baumann G. Growth hormone binding protein and free growth hormone in chronic renal failure. Pediatr Nephrol 1996; 10(3):328. 14. Tonshoff B, Blum WF, Mehls O. Derangements of the somatotropic hormone axis in chronic renal failure. Kidney Int Suppl 1997; 58:S106. 15. Emmanouel DS, Fong VS, Katz AI. Prolactin metabolism in the rat: role of the kidney in degradation of the hormone. Am J Physiol 1981; 249:F437. 16. Davison JM, Sheills EA, Philips PR, et al. Metabolic clearance of vasopressin and an analogue resistant to vasopressinase in human pregnancy. Am J Physiol 1993; 264:F348. 17. Claybaugh JR, Uyehara CF. Metabolism of neurohypophysial hormones. Ann N Y Acad Sci 1993; 689:250. 18. Moses AM, Steciak E. Urinary and metabolic clearances of arginine vasopressin in normal subjects. Am J Physiol 1986; 251:R365. 19. Hruska KA, Martin K, Mennes P, et al. Degradation of parathyroid hormone and fragment production by the isolated perfused dog kidney. J Clin Invest 1977; 60:501. 20. Hruska KA, Korkor A, Martin K, Slatopolsky E. Peripheral metabolism of intact parathyroid hormone. J Clin Invest 1981; 67:885.

21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.

Yamaguchi T, Fukase M, Nishikawa M, et al. Parathyroid hormone degradation by chymotrypsin-like endopeptidase in the opossum kidney cell. Endocrinology 1988; 123(6):2812. Simmons RE, Hjelle JT, Mahoney C, et al. Renal metabolism of calcitonin. Am J Physiol 1988; 254:F593. Chai SY, Christopoulos G, Cooper ME, Sexton PM. Characterization of binding sites for amylin, calcitonin, and CGRP in primate kidney. Am J Physiol 1998; 274(1 Pt 2):F51. Duckworth WC, Bennett RG, Hamel FG. Insulin degradation: progress and potential. Endocr Rev 1998; 19(5):608. Duckworth WC, Hamel FG, Liepnieks J, et al. Insulin degradation products from perfused rat kidney. Am J Physiol 1989; 256:E208. Nakamura R, Hayashi M, Emmanouel DS, Katz AI. Sites of insulin and glucagon degradation in the rabbit nephron. Am J Physiol 1986; 250:F144. Fawcett J, Rabkin R. Sequential processing of insulin by cultured kidney cells. Endocrinology 1995; 136:39. Dahl DC, Tsao T, Duckworth WC, et al. Retroendocytosis of insulin in a cultured kidney epithelial cell line. Am J Physiol 1989; 257:C190. Flyvbjerg A, Nielsen S, Sheikh I, et al. Luminal and basolateral uptake and receptor binding of IGF-I in rabbit renal proximal tubules. Am J Physiol 1993; 265:F624. Ballard FJ, Knowles SE, Walton PE, et al. Plasma clearance and tissue distribution of labelled insulin-like growth factor-I (IGF-I), IGF-II and des(1–3)IGF-1 in rats. J Endocrinol 1991; 128:197. Rabkin R, Fervenza FC, Maidment H, et al. Pharmacokinetics of insulin-like growth factor-1 in advanced chronic renal failure. Kidney Int 1996; 49(4):1134. Hirschberg R, Adler S. Insulin-like growth factor system and the kidney: physiology, pathophysiology, and therapeutic implications. Am J Kidney Dis 1998; 31(6):901. Talor Z, Emmanouel DS, Katz AI. Glucagon degradation by luminal and basolateral rabbit tubular membranes. Am J Physiol 1983; 244:F297. Vierhapper H, Gasic S, Nowotny P, Waldhausl W. Renal disposal of human atrial natriuretic peptide in man. Metabolism 1990; 39:341. Ruskoaho H. Atrial natriuretic peptide: synthesis, release, and metabolism. Pharmacol Rev 1992; 44(4):479. Zhao J, Ardaillou N, Lu CY, et al. Characterization of C-type natriuretic peptide receptors in human mesangial cells. Kidney Int 1994; 46(3):717. Walter M, Unwin R, Nortier J, Deschodt-Lanckman M. Enhancing endogenous effects of natriuretic peptides: inhibitors of neutral endopeptidase (EC.3.4.24.11) and phosphodiesteras e. Curr Opin Nephrol Hypertens 1997; 6(5):468. Lipkin GW, Dawnay AB, Harwood SM, et al. Enhanced natriuretic response to neutral endopeptidase inhibition in patients with moderate chronic renal failur e. Kidney Int 1997; 52(3):792. Emmanouel DS, Stavroyoulos T, Katz AI. Role of the kidney in metabolism of gonadotrophs in rats. Am J Physiol 1984; 247:E786. Emmanouel DS, Goldwasser E, Katz AI. Metabolism of pure human erythropoietin in the rat. Am J Physiol 1984; 247:F168. Goswami A, Rosenberg IN. Iodothyronine 5'-deiodinase in rat kidney microsomes. J Clin Invest 1984; 74:2097. Morreale de Escobar G, Calvo R, Escobar del Ray F, Obregon MJ. Thyroid hormones in tissue from fetal and adult rats. Endocrinology 1994; 134:2410. Chopra IJ, Boado RJ, Geffner DL, Solomon DH. A radioimmunoassay for measurement of thyronine and its acetic acid analog in urine. J Clin Endocrinol Metab 1988; 67:480.

43a.Rolleman EJ, Hennemann G, van Toor H, et al. Changes in renal triiodothyronine and thyroxine handling during fasting. Europ J Endocrinol 2000; 142:125. 44. Santini F, Hurd RE, Lee B, Chopra IJ. Sex related differences in iodothyronine metabolism in the rat: evidence for differential regulation among various tissues. Metabolism 1994; 43:793. 45. Lee WS, Berry MJ, Hediger MA, Larsen PR. The type I iodothyronine 5'-deiodinase messenger ribonucleic acid is localized to the S3 segment of the rat kidney proximal tubule. Endocrinology 1993; 132(5):2136. 46. Leonard JL, Ekenbarger DM, Frank SJ, et al. Localization of type I iodothyronine 5'-deiodinase to the basolateral plasma membrane in renal cortical epithelial cell s. J Biol Chem 1991; 266(17):11262. 47. Kaptein EM. Thyroid hormone metabolism and thyroid diseases in chronic renal failure. Endocr Rev 1996; 17(1):45. 48. Hierholzer K, Schoneshofer M, Siebe H, et al. Corticosteroid metabolism in isolated rat kidney in vitro. I. Formation of lipid soluble metabolites from corticosterone (B) in renal tissue from male rats. Pflugers Arch 1984; 400(4):363. 49. Whitworth JA, Stewart PM, Burt D, et al. The kidney is the major site of cortisone production in man. Clin Endocrinol (Oxf) 1989; 31:355. 50. Funder JW, Pearce PT, Smith R, Smith AI. Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science 1988; 242(4878):583. 51. Aziz N, Maxwell MM, Brenner BM. Coordinate regulation of 11 beta-HSD and Ke 6 genes in cpk mouse: implications for steroid metabolic defect in PKD. Am J Physiol 1994; 267(5 Pt 2):F791. 52. Escher G, Frey BM, Frey FJ. 11 beta-hydroxysteroid dehydrogenase—why is it important for the nephrologist? (Editorial). Nephrol Dial Transplant 1995; 10(9):1506. 53. White PC, Mune T, Agarwal AK. 11 b-Hydroxysteroid dehydrogenase and the syndrome of apparent mineralocorticoid excess. Endocr Rev 1997; 18(1):135. 54. Lo YH, Sheff MF, Latif SA, et al. Kidney 11 b-HSD2 is inhibited by glycyrrhetinic acid-like factors in human urine. Hypertension 1997; 29(1 Pt 2):500. 55. Clore J, Schoolwerth A, Watlington CO. When is cortisol a mineralocorticoid? Kidney Int 1992; 42:1297. 56. Morris DJ, Latif SA, Rokaw MD, et al. A second enzyme protecting mineralocorticoid receptors from glucocorticoid occupancy. Am J Physiol 1998; 274(5 Pt 1):C1245. 57. Ghosh SS, Basu AK, Ghosh S, et al. Renal and hepatic family 3A cytochromes P450 (CYP3A) in spontaneously hypertensive rats. Biochem Pharmacol 1995; 50(1):49. 58. Morris DJ, Brem AS. Metabolic derivatives of aldosterone. Am J Physiol 1987; 252:F365. 59. Egfjord M. Corticosteroid metabolism in isolated perfused rat liver and kidney. Experimental studies with emphasis on aldosterone. Acta Physiol Scand Suppl 1995; 627:1.

CHAPTER 208 EFFECTS OF ENDOCRINE DISEASE ON THE KIDNEY Principles and Practice of Endocrinology and Metabolism

CHAPTER 208 EFFECTS OF ENDOCRINE DISEASE ON THE KIDNEY ELLIE KELEPOURIS AND ZALMAN S. AGUS Pituitary Disease Acromegaly Hypopituitarism Thyroid Disorders Hyperthyroidism Hypothyroidism Parathyroid Disorders Primary Hyperparathyroidism Hypoparathyroidism Disorders of the Adrenal Cortex Hyperaldosteronism Hypoaldosteronism Cushing Syndrome Primary Adrenocortical Insufficiency Chapter References

Endocrine disorders may affect the kidney in various ways. Hormone excess or deficiency states can directly alter tubular transport. Moreover, changes can be produced indirectly by alterations in the circulation, which modify renal hemodynamics and sodium transport. Many of the effects produced are not manifest clinically, but a few can be life threatening; some effects are diagnostically useful.

PITUITARY DISEASE Several aspects of renal function are regulated directly by the anterior and posterior pituitary gland through the secretion of various hormones, the most important of which is vasopressin (see Chap. 23, Chap. 25 and Chap. 27). Two disorders of the anterior pituitary produce changes in renal function: acromegaly and hypopituitarism. ACROMEGALY As part of the generalized visceromegaly, the kidneys of patients with acromegaly are hypertrophied and sometimes achieve a remarkably large size. The glomeruli are enlarged, and all tubular dimensions are increased. The glomerular filtration rate (GFR) and renal blood flow (RBF) are elevated 20% to 50% above normal.1,2 Reversible prostatic enlargement is also common, even in men with hypogonadism.3 Generally, patients with acromegaly manifest a retention of water and electrolytes and an expanded plasma volume, which is appropriate for the degree of tissue growth and the consequent tissue perfusion requirements. The mechanisms by which this response occurs are unclear, but it is not the result of direct effects of growth hormone on renal sodium reabsorption. Hypertension occurs in 30% to 50% of patients with acromegaly. Although hypervolemia contributes to the increased blood pressure, the role of the renin–angiotensin–aldosterone system is unclear. Serum renin and aldosterone levels are normal in acromegaly. These normal values in the face of hypervolemia probably reflect a physiologic resetting at a higher circulating blood volume. Rarely, aldosterone-producing adenomas have been described in conjunction with acromegaly, and they contribute to the blood pressure increase. The electrolyte abnormality seen most frequently in acromegaly is hyperphosphatemia (~70% of patients). Serum phosphate often is elevated to levels of 5 to 6 mg/dL and has been used as a crude marker for the activity of the disease. The hyperphosphatemia reflects an increase in tubular phosphate reabsorption, which is independent of GFR and parathyroid hormone (PTH). The long-term, but not the short-term, administration of growth hormone increases tubular phosphate reabsorption; therefore, this effect may be mediated by a direct effect of insulin-like growth factor-I. 4 Hypercalciuria also is seen in acromegaly and may present with or without nephrocalcinosis. The serum calcium level usually is normal but is mildly elevated in 15% of patients. Hypercalcemia occurs in some patients due to an increase in the endogenous production of calcitriol5 but usually reflects coexisting primary hyperparathyroidism as part of multiple endocrine neoplasia syndrome, type 1 (see Chap. 188). HYPOPITUITARISM A defect in water excretion occurs in almost all patients with panhypopituitarism. The incidence of hyponatremia (usually mild) may be ~50%. Associated hypothyroidism or hyperprolactinemia may alter the renal handling of water, but the renal defect in these patients results primarily from glucocorticoid deficiency.1,6 On occasion, the hyponatremia may be severe and may be the presenting symptom in hypopituitarism.7,8 It is thought to be due to inappropriate secretion of antidiuretic hormone (ADH) that is caused by cortisol (not aldosterone) deficiency.6 The short-term parenteral administration of glucocorticoids is associated with the prompt restoration of normal diluting ability and a water diuresis in a dose-response fashion. The relationship between the plasma vasopressin level and plasma osmolality is also restored to normal. This observation can be used as a diagnostic maneuver in patients with unexplained hyponatremia. Although reduced GFR and increased permeability of the collecting duct in glucocorticoid deficiency contribute to the defect in water excretion, inappropriate secretion of ADH plays a major role (see Chap. 27). The rapid correction with hydrocortisone suggests that glucocorticoids modulate ADH secretion in response to changes in plasma osmolality. The hypersecretion of ADH and its lack of suppression by hypoosmolality may represent the loss of a normal feedback inhibition by cortisol. ADH is cosecreted with corticotropin-releasing hormone (CRH) from the hypothalamic paraventricular nuclei. With cortisol deficiency, an increase is seen in vasopressin messenger RNA as well as CRH and vasopressin immunoreactivity in these neurons.9

THYROID DISORDERS HYPERTHYROIDISM Although, in hyperthyroidism, alterations occur in renal hemodynamics and in the renal handling of water and electrolytes, hyperthyroidism seldom produces clinical manifestations of renal dysfunction.10,11,12,13 and 14 The RBF and GFR are increased because of renal vasodilation. The serum sodium concentration usually is normal, although studies demonstrate a mild impairment in maximal urinary concentration.15 The washout of renal medullary hypertonicity by increased blood flow may be the cause of this concentrating defect. Mild polyuria may occur in some patients, usually when hyperthyroidism is complicated by hypercalcemia or hyperglycemia. In patients with thyrotoxicosis, severe dehydration and hypernatremia may develop, mainly because of excessive water loss from the skin and lungs, and inadequate water intake. Mild proteinuria is common.16 Although serum potassium generally is normal in hyperthyroidism, total body potassium often is decreased concomitantly with the loss of lean body mass. With control of the hyperthyroid state, body weight and total body potassium levels normalize. In a distinct group of thyrotoxic patients, periodic muscular paralysis in conjunction with episodic hypokalemia can occur. Hyperthyroid Asian men are at particular risk (estimated at 15–20%) of developing the disease.17 The muscle function is improved with establishment of a euthyroid state and the administration of a b-adrenergic blocker. Use of b-blockers minimizes both the number and the severity of attacks, and often limits the fall in the plasma potassium concentration.18 The mechanism by which hyperthyroidism can produce hypokalemic periodic paralysis is not well understood. Thyroid hormone increases Na+ –K+–adenosine triphosphatase (ATPase) activity (potassium is driven into cells), and thyrotoxic patients with periodic paralysis have higher Na+–K+ –ATPase activity than those without paralytic episodes.19 Excess thyroid hormone may predispose to paralytic episodes by increasing the susceptibility to the hypokalemic action of epinephrine or insulin.18 Hyperthyroidism often is associated with disorders of divalent ion homeostasis. Hypercalcemia may occur due to the direct effect of thyroid hormone on bone resorption, although osteoporosis seldom is encountered.20 Total serum calcium is elevated in 10% to 20% of cases, and ionized serum calcium is increased in

approximately half.21,22 and 23 Usually, symptoms are minimal or absent because of the mild elevation, but with a greater increase, thirst and polyuria may occur. The hypercalcemia resolves after successful return to the euthyroid state and also can be controlled with b-adrenergic blocking drugs. If hypercalcemia does not resolve, the rare possibility of coexisting primary hyperparathyroidism should be considered. PTH levels should be measured, because thyrotoxicosis is associated with an increased incidence of parathyroid adenomas. Hypercalciuria is commonly seen in hyperthyroidism because of the increased filtered load of calcium resulting from increased GFR and the mobilization of bone calcium, and also from the suppression of PTH secretion caused by the increased serum calcium. Occasionally, the hypercalcemia and hypercalciuria are complicated by nephrocalcinosis and the development of nephrolithiasis.16 Rarely, patients with nephrocalcinosis may manifest an overt, distal, renal tubular acidosis, but this also has been reported in patients without nephrocalcinosis. The autoimmune nature of thyroid disease may underlie an immunologic renal injury manifested as an acidification defect. Elevation of the serum phosphate concentration is common in hyperthyroidism, as a result of an increased tubular reabsorption of phosphate secondary to decreased PTH secretion and a direct effect of thyroid hormone, which increases proximal tubular phosphate reabsorption.10 Hyperthyroidism also has been associated with mild hypomagnesemia caused by enhanced urinary excretion of magnesium resulting from the effects of the hypercalcemia or hypercalciuria, which inhibit magnesium transport in the loop of Henle.24 HYPOTHYROIDISM The kidney may undergo various structural and functional abnormalities in adult hypothyroidism. Anatomically, the most striking finding is the marked thickening of the glomerular and tubular basement membranes. Substances rich in mucopolysaccharides deposit in both these structures, as well as in renal blood vessels and the glomerular mesangium and renal interstitium. Cellular changes include vacuolization and periodic acid–Schiff–positive inclusion droplets. These changes are reversible after thyroid function is normalized. Mild proteinuria often is present, but the development of nephrotic syndrome is rare. An autoimmune glomerulonephritis associated with Hashimoto thyroiditis has been implicated in the development of nephrotic syndrome. In myxedema, RBF and GFR are consistently depressed 25% to 40% below normal values.25 These hemodynamic alterations reflect the hypodynamic state of the circulation and the renal vasoconstriction. The depressed cardiac output and, perhaps, the renal structural changes contribute to the fall in RBF and GFR. Tubular reabsorptive and secretory processes generally are decreased and reverse rapidly with the achievement of an euthyroid state. In hypothyroidism, the most significant manifestation of changes in renal function is mild hyponatremia in 20% of patients, which results from an impairment in renal diluting capacity leading to water retention. Thyroid function should be evaluated in any patient with an otherwise unexplained reduction in the plasma sodium concentration. Severe hyponatremia may occur with myxedema coma. The major factor contributing to the water retention appears to be the lack of suppression of ADH secretion.26 Measurements of circulating ADH levels have demonstrated elevated titers of the hormone that are not suppressed by water loading. Although a decrease in circulating blood volume may play a role, thyroid hormone deficiency probably causes impairment of the ability to respond to hypoosmolality similar to that which is observed in states of glucocorticoid deficiency.26 ADH secretion may be regulated at a lower serum osmolality, causing a “reset-osmostat” condition.27 Decreased delivery of tubular fluid to the distal diluting segments also has been implicated as a factor contributing to the diluting defect. This is caused by the hemodynamic alterations produced by a decreased cardiac output, which lead to a decrease in RBF and GFR, and to enhanced proximal tubular sodium and water reabsorption. Sodium chloride reabsorption in the diluting segments also may be diminished. In states of combined diabetes insipidus and hypothyroidism, the effect of thyroid deficiency in reducing free water clearance actually may be beneficial. In this situation, treatment with thyroid hormone may uncover a massive increase in urinary water losses. Regardless of the mechanism of the hyponatremia, normal water balance and correction can be rapidly achieved by the administration of thyroid hormone. Myxedema is characterized by changes in the distribution of salt and water in the various body compartments. Thus, although total body sodium usually is increased, circulating blood volume often is low. This may be related partly to the binding of sodium to interstitial mucopolysaccharides. Total exchangeable potassium is either normal or decreased, but serum potassium is normal. The renal tubular transport of sodium and potassium generally is normal, although mild defects have been described in some instances. Usually, the kidney adequately regulates the renal excretion of sodium and potassium over a wide range of dietary intake. Similarly, acid-base regulation generally is intact in hypothyroidism. Hyperuricemia in adult men and postmenopausal women with myxedema often is caused by a renal impairment in uric acid excretion (i.e., decreased renal tubular secretion of urate). Abnormalities in divalent ion metabolism in adult hypothyroidism often are mild and do not manifest as significant clinical problems. A generalized decrease in bone turnover is a common feature, associated with the decreased urinary and fecal excretion of calcium and phosphorus. Serum levels of calcium and phosphate are normal, but a loss of the circadian rhythm for serum and urinary phosphate has been reported, and modest hypermagnesemia frequently is encountered.24

PARATHYROID DISORDERS PRIMARY HYPERPARATHYROIDISM The renal effects of hyperparathyroidism are related principally to the degree and duration of hypercalcemia and the rate of onset of the elevation in the serum calcium concentration. Other features of hyperparathyroidism may play a role in the disturbances of water and electrolyte metabolism. These include renal phosphate wasting, stimulation of the renal production of 1,25-dihydroxyvitamin D3 (by hypophosphatemia and PTH), and direct effects of high levels of circulating PTH.28 Renal manifestations of primary hyperparathyroidism include decreased GFR, impaired urinary concentrating ability, reduced proximal phosphate reabsorption with consequent phosphaturia, hypercalciuria, nephrolithiasis, nephrocalcinosis, obstructive uropathy, distal renal tubular acidosis, hyperchloremic acidosis, hypertension, and normal or increased urinary excretion of magnesium. Hypercalciuria is a characteristic feature of primary hyperparathyroidism. The elevated serum calcium seen in this disorder significantly increases the fractional excretion of calcium in the urine. The hypercalciuria is a major factor contributing to the development of renal calculi and nephrocalcinosis in primary hyperparathyroidism. In older series of patients with hyperparathyroidism, the incidence of stone disease exceeded 50%, but the early diagnosis and treatment of asymptomatic cases of hyperparathyroidism have led to a much lower incidence of renal calculi (20%).29,30 In addition to calcium oxalate kidney stones, patients with primary hyperparathyroidism commonly have calcium phosphate stones because of the tendency of these patients to have an alkaline urine and the moderate enhancement of phosphate excretion that is present. Nephrocalcinosis, or renal parenchymal calcification, may be present; it is most commonly medullary but may be seen in the cortical areas of the kidney as well (Fig. 208-1). Nephrocalcinosis and nephrolithiasis are not necessarily associated. When detected on ordinary radiographic films, nephrocalcinosis usually is advanced and reflects severe renal parenchymal involvement. Earlier forms of the disease, not detected by conventional radiography, can be diagnosed by ultrasonography or computed tomography.31

FIGURE 208-1. Radiographic appearance of bilateral medullary nephrocalcinosis in a patient with primary hyperparathyroidism. Deposits of calcium phosphate are localized primarily in the medulla (probably beginning in collecting ducts and spreading to periductal tissue). This may also occur in patients with renal tubular acidosis, chronic urinary tract infection, and idiopathic hypercalciuria.

Renal insufficiency may occur in hyperparathyroidism and is related to the degree and duration of hypercalcemia. Mild hypercalcemia alone is only rarely associated with renal insufficiency. In a series of patients with asymptomatic hyperparathyroidism and mild hypercalcemia (70 dB. Speech reception threshold is the sound intensity at which the subject begins to hear words. Speech discrimination is the ability to understand a list of unrelated words. Sensorineural hearing loss is the most common type of loss. Treatment usually involves the use of hearing aids and aural rehabilitation. (ANSI, American National Standards Institute.)

PITUITARY GLAND In acromegaly, the tongue is increased in size. Also, most affected patients have a change in voice that consists of decreased pitch and a huskier sound. Occasionally, the recurrent laryngeal nerve is stretched, which causes vocal cord paralysis. If the condition does not improve spontaneously, the voice change can be treated by thyroplasty. A patient may develop diabetes insipidus as a result of trauma to the skull base secondary to an operation or a disease process that is otolaryngologic in origin. Large pituitary tumors can infiltrate the sinuses. The major interest of the otolaryngologist in pituitary disease has been patient management. The transseptal-transsphenoidal surgical approach is primarily a team procedure in which the rhinologist provides access for the neurosurgeon.2,3 This type of cooperative effort has markedly improved the management and outcome of these patients (see Chap. 23).

THYROID GLAND With hyperthyroidism (thyrotoxicosis, Graves disease), a patient may have a goiter and Graves ophthalmopathy. Often, the otolaryngologist is involved in the surgical management of the ophthalmopathy. Transantral orbital decompression is the most widely used surgical procedure for this condition.4 This procedure allows removal of the inferior and medial orbital walls to provide space for the excess tissue in the sinuses. No external incisions are necessary, and complications are minimal. Endoscopic decompression has been described and appears successful (see Chap. 43).5 A number of pertinent findings are associated with myxedema (hypothyroidism). A conductive hearing loss secondary to serous otitis media may be present (Fig. 216-2). Also, a sensorineural hearing loss may be present. The conductive loss usually resolves with treatment, but the sensorineural loss generally persists. Generalized mucosal edema produces nasal obstruction, thickened tongue, facial edema, hoarseness, and slowed speech. Although diagnosis is rarely a problem, if doubt arises, biopsy of the nasal mucosa can be performed and will reveal an increase in acid mucopolysaccharide content.

FIGURE 216-2. Audiogram demonstrates normal hearing in the left ear and conductive hearing loss in the right ear. Notice that the air conduction levels in the right ear are separated from the bone conduction markers. The speech reception threshold corresponds to the air conduction level. Speech discrimination scores are normal in both ears. This type of hearing loss usually is associated with some malfunction in the ossicular chain, tympanic membrane, or external ear canal. Surgical correction usually is possible. (ANSI, American National Standards Institute.)

Carcinoma of the thyroid may present as a mass in the gland, a neck mass of unknown cause, or a cause of vocal cord paralysis. Vocal cord paralysis secondary to recurrent laryngeal nerve involvement or surgical trauma often responds to thyroplasty, arytenoidectomy, or other appropriate therapy.6,7 When the tumor is under control, the customary practice is to wait 6 months before instituting therapy. If the prognosis is poor, however, and the patient is having trouble with aspiration or a weak voice, the therapy is performed immediately. Acute bacterial thyroiditis (see Chap. 46 and Chap. 213) may be associated with a large neck mass, hoarseness, vocal cord paralysis, or a compromised airway. In such cases, tracheostomy or vocal cord injection with fat may be indicated. An infant with congenital hypothyroidism (cretinism) usually presents with a severe sensorineural hearing loss (see Fig. 216-1), a broad flat nose, and a high-pitched cry8 (see Chap. 47). Correction of the hypothyroidism improves the voice, but the hearing loss remains. In Pendred syndrome9—a rare congenital syndrome associated with bilateral sensorineural hearing loss and a euthyroid goiter—the hearing loss involves high-frequency sound primarily, is of varying severity, and is nonreversible. The goiter results from a defect in the organification of thyroid hormone. Thyroid function tests are normal, and the diagnosis is confirmed using a perchlorate washout test. No treatment is available, but genetic counseling may be appropriate.10 Another nontreatable syndrome—Hollander syndrome—is characterized by progressive sensorineural hearing loss and euthyroid goiter.

CALCIUM AND PHOSPHATE METABOLISM Patients with hyperparathyroidism may have hearing loss, dysphagia, fasciculations of the tongue, tumors of the facial bones, and lesions of the oral mucosa. The hearing loss is sensorineural and nonreversible (see Fig. 216-1). The lesions seen in the facial bones are called brown tumors (osteitis fibrosa cystica). These lesions are benign, are most often located in the maxilla, and need not be excised unless they cause functional or cosmetic problems. The nodular lesion of the oral mucosa, called epulis, requires no therapy. Hyperparathyroidism is an important component of multiple endocrine neoplasia (MEN) (see Chap. 188). In MEN type 1, the findings may include those of hyperparathyroidism, pituitary tumor, and pancreatic tumor. In MEN type 2A, hyperparathyroidism, pheochromocytoma, and medullary thyroid carcinoma (occasionally presenting as a neck mass) are found. In MEN type 2B, the findings include medullary thyroid cancer, pheochromocytoma, and neuromas involving the mucosa lining the lips, oral cavity, nose, larynx, and eyes; hyperparathyroidism is rare. These neuromas are histologically benign, but their presence should alert the clinician to the possibility of a MEN syndrome. In any patient suspected of having MEN type 2A or 2B, the clinician must screen for pheochromocytoma preoperatively. If a secreting pheochromocytoma is present, anesthesia and surgery are exceedingly risky.11 Hypercalcemia that is unrelated to parathyroid malfunction may be seen with malignant lesions involving the head and neck region and with sarcoidosis (see Chap. 59). Careful head and neck examination should reveal any primary cancer. Sarcoidosis causes many characteristic findings in the head and neck region.12,13 Granular lesions of the nasal mucosa, ulcers of the larynx, neck masses, and swelling of the salivary glands are commonly seen. Hypoparathyroidism with hypocalcemia produces nerve irritability, which causes laryngeal stridor, laryngospasm, and a positive Chvostek sign (see Chap. 60). Frequently, this hypoparathyroidism is a sequela of a surgical procedure in the neck. Severe hypocalcemia and hypomagnesemia can cause bilateral vocal cord paralysis.14 Hypophosphatasia, as well as hyperphosphatasia, has been reported to occur in infants found to have an associated hearing loss.15

METABOLIC BONE DISEASE Several metabolic diseases of bone produce significant otolaryngologic findings, one of which is Paget disease of the bone (see Chap. 65). This is a disease process that has truly protean manifestations that are revealed in middle and old age. Skull changes are relatively common, as is involvement of the temporal bone. Clinically, a significant number of these patients have hearing loss. Classically, this loss begins as a mixed hearing loss and then becomes purely sensorineural (Fig. 216-3). Moreover, these patients can have tinnitus and vertigo, which seem to be related to the disease.16 Calcitonin therapy may halt the progress of the involvement.17

FIGURE 216-3. Audiogram reflects the type of hearing loss that is often seen with Paget disease of the bone. In the left ear, a mixed hearing loss is present: part of the loss is sensorineural and part is conductive. Notice the separation, in the low frequencies, between air conduction and bone conduction levels on the left side. The speech reception threshold corresponds to the air conduction level, but speech discrimination remains fairly good at this point. The hearing level in the right ear is fairly typical of end-stage Paget disease of the temporal bone. This is a fairly severe sensorineural hearing loss with poor speech discrimination. In such cases, the mixed hearing loss might occur relatively early in the course of the disease, whereas the sensorineural hearing loss might occur relatively late. (ANSI, American National Standards Institute.)

Osteogenesis imperfecta is a genetically induced disease with varying levels of involvement18 (see Chap. 66, Chap. 70 and Chap. 189). Affected patients may have a conductive hearing loss secondary to involvement of the ossicular chain. Clinically and histologically, the condition is identical to otosclerosis. Surgical treatment and use of hearing aids may be necessary. At operation, if the incus is involved, performing a stapedectomy may not be possible. In fibrous dysplasia (see Chap. 66), the maxilla is the bone most commonly involved in the head (Fig. 216-4). Mass lesions can form anywhere, leading to cosmetic and functional problems.19,20 These lesions can be debulked or removed, but they have a tendency to recur. Malignant transformation is rare but is more likely if the patient has been irradiated.21

FIGURE 216-4. Radiograph showing fibrous dysplasia of the left maxilla. Severe cystic involvement (arrow), overall expansion of the bone, and extreme thinning of the cortex (arrowhead) are evident. Notice the asymmetry of the orbits, as well as the gross deformity of the left orbit.

Osteopetrosis (Albers-Schönberg disease) may have otologic complications as well. The involvement varies greatly from patient to patient, but the temporal bone is commonly affected.22 Temporal bone disease causes sensorineural and conductive hearing losses. Usually, these are not treatable except with a hearing aid. Moreover, facial paralysis may be seen; this tends to be recurrent, and decompression often is advised. Dental caries occurs frequently and is severe. This condition may cause osteomyelitis of the mandible, which is difficult to control.

ADRENAL CORTEX Patients with adrenal insufficiency (Addison disease) may present with sunken eyes, dry tongue, and hyperpigmentation of the skin and tongue (see Chap. 76). Endolymphatic hydrops (Ménière syndrome), dysosmia, and dysgeusia all have been reported to occur with adrenal insufficiency. Adrenocortical hyperactivity (Cushing disease) causes moon facies and prominent supraclavicular fat pads.

PREGNANCY The most common otolaryngologic abnormality related to pregnancy is a severe vasomotor rhinitis. This usually arises in the second or third trimester. No predisposing factors have been identified. Treatment with decongestants is of limited benefit. Another unusual finding is hoarseness, which arises from vascular engorgement of the true and false vocal cords. An extremely dry throat—laryngitis sicca gravidarum—occasionally is seen. All of these conditions resolve after delivery. Finally, the incidence of facial paralysis is three times greater in pregnant women than in nonpregnant women; the cause for this is unknown, and treatment usually is limited to careful observation.23

DIABETES MELLITUS Sensorineural hearing loss has been reported to appear earlier and is more severe in diabetic individuals than in the normal population. When diabetic patients are compared with a general population of the same age and sex, however, no difference in hearing levels is found.24,25 Other conditions that have been thought to be more prevalent in diabetic persons are Ménière syndrome, facial nerve palsy (Bell palsy), and vocal cord paralysis, but none of these has been conclusively shown to be more common in these patients than in the population as a whole. Infections are a great problem in diabetic individuals. Two infectious processes are unique and are reviewed in some detail. The first is malignant (necrotizing) external otitis, a disease that affects elderly diabetic patients.26,27 It is a unilateral process that begins as a routine external otitis. Despite therapy, it evolves from a soft tissue infection into an osteomyelitis of the temporal bone and eventually involves the base of the skull. Pseudomonas aeruginosa is the major pathogen. The cardinal symptoms are severe otalgia and otorrhea. Granulation tissue is present at the anterior junction of the bony and the cartilaginous external auditory canal. The clinical appearance is deceptively mild, and the course of the disease is prolonged. Once the bone is involved, cranial nerve deficits develop. The facial nerve is the one most commonly affected, but cranial nerves VI through XII can be involved. If more than one cranial nerve is affected, the prognosis is poor. The treatment must be aggressive and multi-faceted. Local treatment consists of placing an antibiotic-soaked wick in the ear canal. Systemic therapy uses an aminoglycoside and one of the synthetic penicillins. Prolonged therapy with a cephalosporin or fluoroquinolone may be effective.28 Treatment must be continued for at least 4 weeks, and probably for 6 weeks. Surgical management is limited to debridement of the temporal bone, usually a radical mastoidectomy. Controlling the diabetes concurrently and monitoring renal function is imperative. Treatment must be continued until all clinical evidence of the disease disappears, the erythrocyte sedimentation rate normalizes, the gallium bone scan improves, and radiographic evidence of resolution is noted.29 At one time, the mortality rate associated with this disease was 50%, but it is now around 10% (see Chap. 152). The second infection of concern is mucormycosis. The two forms seen in the head and neck region are the rhinoorbital-cerebral form and the otic form.30,31,32 and 33,33a The nasal form presents with blindness, ophthalmoplegia, proptosis, facial swelling, palatal ulcer, or disorders of consciousness. Examination reveals brick red or black areas within the nasal cavity. Biopsy confirms the clinical impression of mucormycosis. The otic presentation usually is accompanied by otorrhea, followed by facial paralysis and then altered sensorium. Treatment must be prompt and aggressive. The diabetes must be controlled, because ketoacidosis and dehydration are invariably present at the onset of the disease. Systemic therapy is initiated with amphotericin B. Surgical debridement is more important in mucormycosis than in malignant otitis externa. Surgical management involves the removal of all necrotic tissue and the establishment of drainage from areas of infection. If the facial or optic nerve is involved, decompression is indicated. The mortality rate associated with this condition is 40%.

HYPOGLYCEMIA Reactive postprandial hypoglycemia34 has been associated with Ménière syndrome, fluctuating hearing loss, and episodic vertigo.35 Although some authors believe that hypoglycemia plays a major role in the evolution of these symptom complexes, the objective data are not impressive. The subjective nature of the symptoms and the tendency for remission to occur make scientific evaluation difficult.

LIPID METABOLISM Hyperlipidemia causes characteristic xanthomas of the face and may be associated with sensorineural hearing loss.36 The view once was that patients with Ménière syndrome, fluctuating hearing loss, episodic imbalance, and premature hearing loss often had hyperlipidemia.37 Most of those studies had flaws in their design, and hyperlipidemia probably is not a significant factor in these conditions. In abetalipoproteinemia, which is a rare recessive disorder, the presenting features include ataxia, acanthocytosis, and sensorineural hearing loss.38 CHAPTER REFERENCES 1. Ito H, Takamoto T, Nitta M, et al. DIDMOAD (diabetes insipidus, diabetes mellitus, optic atrophy and deafness) syndrome associated with myocardial disease. Jpn Heart J 1988; 29:371. 1a. Jenkin A, Renner D, Hahn F, Larsen J. A case of primary amenorrhea, diabetes and anosmia. Gynecol Endocrinol 2000; 14:65. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Laws ER Jr, Thapar K. Surgical management of pituitary adenomas. Baillières Clin Endocrinol Metab 1995; 9:391. Carrau RL, Jho HD, Ko Y. Transnasal-transsphenoidal endoscopic surgery of the pituitary gland. Laryngoscope 1996; 106:914. Harner SG. Orbital decompression techniques. In: Gorman CA, Waller RR, Dyer JA, eds. The eye and orbit in thyroid disease. New York: Raven Press, 1984:221. Metson R, Shore JW, Gliklich RE, Dallow RL. Endoscopic orbital decompression under local anesthesia. Otolaryngol Head Neck Surg 1995; 113:661. Bauer CA, Valentino J, Hoffman HT. Long-term result of vocal cord augmentation with autogenous fat. Ann Otol Rhinol Laryngol 1995; 104:871. Netterville JL, Aly A, Ossoff RH. Evaluation and treatment of complications of thyroid and parathyroid surgery. Otolaryngol Clin North Am 1990; 23:529. Chaouki ML, Maoui R, Benmiloud M. Comparative study of neurological and myxoedematous cretinism associated with severe iodine deficiency. Clin Endocrinol (Oxf) 1988; 28:399. Friis J, Johnsen T, Feldt-Rasmussen U, et al. Thyroid function in patients with Pendred's syndrome. J Endocrinol Invest 1988; 11:97. Maisel RH, Brown DR, Ritter FN. Endocrinology. In: Paparella MM, Shumrick DA, eds. Otolaryngology, vol 1. Philadelphia: WB Saunders, 1980:779. Werbel SS, Ober KP. Pheochromocytoma. Update on diagnosis, localization, and management. Med Clin North Am 1995; 79:131. Krespi YP, Kuriloff DB, Aner M. Sarcoidosis of the sinonasal tract: a new staging system. Otolaryngol Head Neck Surg 1995; 112:221. Benjamin B, Dalton C, Croxson G. Laryngoscopic diagnosis of laryngeal sarcoid. Ann Otol Rhinol Laryngol 1995; 104:529.

14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33.

Lye WC, Leong SO. Bilateral vocal cord paralysis secondary to treatment of severe hypophosphatemia in a continuous ambulatory peritoneal dialysis patient. Am J Kidney Dis 1994; 23:127. Schuknecht HF. Pathology of the ear. Cambridge, MA: Harvard University Press, 1974:172. Harner SG, Rose DE, Facer GW. Paget's disease and hearing loss. Otolaryngology 1978; 86:869. El Samma M, Linthicum FH Jr, House HP, House JW. Calcitonin as treatment for hearing loss in Paget's disease. Am J Otolaryngol 1986; 7:241. Bergstrom L. Osteogenesis imperfecta: otologic and maxillofacial aspects. Laryngoscope 1977; 87(Suppl):1. Feldman MD, Rao VM, Lowry LD, Kelly M. Fibrous dysplasia of the paranasal sinuses. Otolaryngol Head Neck Surg 1986; 95:222. Megerian CA, Sofferman RA, McKenna MJ, et al. Fibrous dysplasia of the temporal bone: ten new cases demonstrating the spectrum of otologic sequelae. Am J Otol 1995; 16:408. Sofferman RA. Cysts and bone dyscrasias of the paranasal sinuses. In: English GM, ed. Otolaryngology, vol 2. Philadelphia: Harper & Row, 1985:13. Stocks RM, Wang WC, Thompson JW, et al. Malignant infantile osteopetrosis: otolaryngological complications and management. Arch Otolaryngol Head Neck Surg 1998; 124:689. Hilsinger RL Jr, Adour KK. Idiopathic facial paralysis, pregnancy, and the menstrual cycle. Ann Otol Rhinol Laryngol 1975; 84:433. Harner SG. Hearing in adult-onset diabetes mellitus. Otolaryngol Head Neck Surg 1981; 89:322. Duck SW, Prazma J, Bennett PS, Pillsbury HC. Interaction between hypertension and diabetes mellitus in the pathogenesis of sensorineural hearing loss. Laryngoscope 1997; 107:1596. Chandler JR. Malignant external otitis and osteomyelitis of the base of the skull. Am J Otol 1989; 10:108. Slattery WH 3rd, Brackmann DE. Skull base osteomyelitis. Malignant external otitis. Otolaryngol Clin North Am 1996; 29:795. Levenson MJ, Parisier SC, Dolitsky J, Bindra G. Ciprofloxacin: drug of choice in the treatment of malignant external otitis (MEO). Laryngoscope 1991; 101:821. Grandis JR, Curtin HD, Yu VL. Necrotizing (malignant) external otitis: prospective comparison of CT and MR imaging in diagnosis and follow-up. Radiology 1995; 196:499. Vessely MB, Zitsch RP 3rd, Estrem SA, Renner G. Atypical presentations of mucormycosis in the head and neck. Otolaryngol Head Neck Surg 1996; 115:573. Sugar AM. Mucormycosis. (Review). Clin Infect Dis 1992; 1(Suppl 14):S126. Harril WC, Stewart MG, Lee AG, Cernoch P. Chronic rhinocerebral mucormycosis. Laryngoscope 1996; 106:1292. Gussen R, Canalis RF. Mucormycosis of the temporal bone. Ann Otol Rhinol Laryngol 1982; 91:27.

33a. Ferguson BJ. Mucormycosis of the nose and paranasal sinuses. Otolaryngol Clin North Am 2000; 33:349. 34. 35. 36. 37. 38.

Betteridge DJ. Reactive hypoglycemia. BMJ 1987; 295:286. de Vincentiis I, Ralli G. New pathogenetic and therapeutic aspects of Ménière's disease. Adv Otorhinolaryngol 1987; 37:97. Pulec JL, Pulec MB, Mendoza I. Progressive sensorineural hearing loss, subjective tinnitus and vertigo caused by elevated blood lipids. Ear Nose Throat J 1997; 76:716. Spencer JT Jr. Hyperlipoproteinemia and inner ear disease. Otolaryngol Clin North Am 1975; 8:483. Liston S, Meyerhoff WH. Metabolic hearing loss. In: English GM, ed. Otolaryngology, vol 1. Philadelphia: Harper & Row, 1985:9.

CHAPTER 217 DENTAL ASPECTS OF ENDOCRINOLOGY Principles and Practice of Endocrinology and Metabolism

CHAPTER 217 DENTAL ASPECTS OF ENDOCRINOLOGY ROBERT S. REDMAN Ontogeny of the Orofacial Structures Pituitary Gland Hypopituitarism Hyperpituitarism Thyroid Gland Hypothyroidism Hyperthyroidism Calcium and Phosphorus Metabolism Nutritional Factors Parathyroid Hormone Renal Function Adrenal Glands Hypoadrenocorticism Hyperadrenocorticism Adrenal Medulla Sex Hormones Women Men Pancreas: Diabetes Mellitus Chapter References

Most of the major circulating hormones are important in the normal growth and development of the orofacial region, including the teeth, and most of them also participate in the maintenance of the health and integrity of these structures. Consequently, hormonal abnormalities commonly have dental and oral manifestations. Many of these oral signs and symptoms, and the endocrinopathies and other disorders with which they may be associated, are summarized in Table 217-1 and Table 217-2.

TABLE 217-1. Dental and Orofacial Abnormalities That May Be Associated with Specific Metabolic or Endocrine Disorders

TABLE 217-2. Reported Effects of Various Endocrine Conditions on the Sense of Taste*

ONTOGENY OF THE OROFACIAL STRUCTURES When one considers the potential for hormonal effects on the development of mature oral structures, two phenomena must be examined. First, with regard to tooth development, the various stages of odontogenesis occur at different ages for each pair of teeth. The first primary teeth are initiated at ~6 weeks after conception, whereas the last secondary teeth (the third molars) finish root formation at ~20 years after birth.1 Thus, significant deviations from the normal chronology of tooth eruption (Fig. 217-1) can be an important sign of endocrinopathy, as well as of metabolic or nutritional problems. Furthermore, any part of a tooth that has completely calcified undergoes no significant further developmental changes in shape or composition. Thus, the duration, as well as the severity, of hormonal changes determine which parts of which teeth are affected. Once the teeth have fully formed and erupted, they can be altered only by destructive processes, such as periodontitis and caries, and their capacity for repair is limited. Therefore, abnormalities of shape or mineralization that are restricted to certain teeth also may serve as a permanent record, indicating when and for how long an endocrinopathy has been a factor (Fig. 217-2). Second, the maxilla, mandible, and most of the facial bones grow by intramembranous ossification, except for epiphyseal plate-like growth in a rim of hyaline cartilage on the head of each mandibular condyle. This cartilage persists in the adult.2 Thus, with appropriate hormonal stimulation, during adulthood, the facial bones and both jaws can enlarge by accretion, but the mandible also can elongate from the condyles.

FIGURE 217-1. Chronology of eruption of the permanent, or secondary, teeth. Eruption time, as depicted in this diagram, is defined as the time when the tooth first pierces the gingiva and becomes visible in the oral cavity. The numbers within the teeth designate the mean eruption times in years and months; example, the maxillary central incisors (next to the midline) erupt at the age of 7 years, 5 months in boys, and 7 years, 2 months in girls. The third molars (“wisdom teeth”) contain no numbers because their eruption times are extremely variable and are, therefore, of little use as a sign of metabolic or hormonal disturbance. Of the deciduous teeth (not shown),

the mandibular incisors usually erupt first, beginning at ~6 months after birth, followed by the first molars, canines, and second molars. Notice that when the deciduous molars are shed, they are replaced by the permanent premolars. All of the deciduous teeth usually have erupted by the age of 2 years. Like the permanent teeth, the deciduous teeth tend to erupt earlier in girls than in boys. (Modified from Sinclair D. Human growth after birth. London: Oxford University Press, 1978:103.)

FIGURE 217-2. Severe hypoplasia of enamel and dentin. The parents of this 7-year-old girl were unsure of the nature of her illness in early childhood. The limitation of the dental defect to the incisal third of the secondary teeth, of which only the central incisors (c) have erupted, would be compatible with the onset of an endocrinopathy, such as hypothyroidism, shortly after birth and its subsequent diagnosis and appropriate treatment at 14 to 16 months of age. (Courtesy of Douglas J. Sanders, DDS.)

PITUITARY GLAND Experiments in which the incisors continuously erupt have indicated the relative importance of the pituitary and several of its target endocrine gland hormones in the production of dental and alveolar bone abnormalities in hypopituitarism.3 In hypophysectomized rats, the rate of eruption progressively slows, then ceases, and the incisors become reduced in size and misshapen. Amelogenesis, morphogenesis, and rate of eruption are largely restored by thyroxine administration, whereas dentinogenesis and alveolar bone growth nearly normalize with growth hormone supplementation. Decreased levels of adrenocortical hormones also may participate in the abnormal morphogenesis of the incisors. HYPOPITUITARISM In pituitary dwarfism (growth hormone deficiency), eruption of both primary and secondary dentitions is delayed, and shedding of the primary teeth is delayed.3,4,5 and 6 The crowns of the teeth reportedly are smaller than normal, although some researchers have suggested that the crowns appear smaller only because they are incompletely erupted.3 Also, the roots of the teeth are noticeably stunted.3,4,5 and 6 The overall growth of the jaws is retarded, with the maxilla being less affected than the mandible.3 In this condition, the alveolar (tooth-supporting) regions of both jaws grow at a disproportionately reduced rate. Consequently, the dental arches are too small to accommodate all of the teeth, causing crowding and malocclusion.3,5,6 Hypofunction of the salivary glands also may occur, leading to increased dental caries and periodontitis.6 Adult-onset panhypopituitarism has no specific effects on the teeth, but characteristic orofacial changes include thinning of the mucosa of the lips and an immobile facial expression.3 HYPERPITUITARISM Growth hormone excess that occurs before puberty (gigantism) causes progressive, symmetric enlargement of the jaws, tongue, and teeth,3,6 and the eruption of the teeth is accelerated.3 Orofacial features of acromegaly emerge when the hyperpituitarism continues past, or begins after, 8 to 10 years of age.3,4 and 5 The jaws and facial bones enlarge disproportionately in relation to most of the bones of the skull because resumption of osseous growth is more vigorous in the intramembranous bones. Moreover, endochondral ossification resumes in the hyaline cartilage of the heads of the mandibular condyles, and the mandibular angles become flattened. This causes progressively more severe mandibular prognathism. The palatal arch is flattened, and panoramic dental radiographs may demonstrate enlarged maxillary sinuses. The periosteum of the jaws may become ossified at points of attachment of the muscles and tendons. The crowns of the teeth are not enlarged, but often excessive deposition of cementum on the roots (hypercementosis) occurs. The tongue may become so large that indentations form where it encroaches on the teeth. Partly from this pressure, and partly from the enlargement of the jaws, the teeth become spaced and outwardly tipped (Fig. 217-3). The nose and lips also are enlarged, adding to the general coarsening of the facial features (see Chap. 12).

FIGURE 217-3. The maxillary anterior teeth of a 52-year-old man with acromegaly are pictured. Notice the wide spaces between the teeth, as well as their anterior inclination.

Enlargement of the bones of the face and skull, and spacing and hypercementosis of the teeth can occur in osteitis deformans (Paget disease of bone).3,4 In contrast to that associated with acromegaly, however, the enlargement in this disease is limited to the bones. The lips may become thinner because of stretching, and the maxilla tends to enlarge disproportionately to the mandible, producing an anterior open bite and a maxillary prognathism. Also, the jaws and skull bones are especially inclined to exhibit radiolucency (osteoclastic stage) and fuzzy (cotton-wool) radiodensity (osteoblastic stage).

THYROID GLAND The serum concentration of thyroid hormone is low in neonatal rats and mice but increases to adult levels between 2 and 3 weeks of age. When this rise is prevented by thyroidectomy, the weaning process, tooth eruption, and maturation of the salivary glands are all retarded but not prevented.7,8 and 9 Rodent parathyroid glands are enclosed in the poles of the thyroid gland, and surgical removal of the thyroid eliminates both glands. Similar developmental retardation occurs when hypothyroidism is induced by propylthiouracil, however, and is prevented with timely replacement of thyroxine.7 This indicates that the effects are attributable to the lack of thyroxine and not to any disturbance of parathyroid hormone. Furthermore, the administration of thyroxine to suckling mice and rats induces precocious development of salivary glands8 as well as of the teeth.3 Also, in mature rats and mice, deficiency of thyroxine causes up to 50% reductions in salivary flow and of gland stores of amylase, protease, and other salivary proteins.9 These findings suggest that decreased salivary function, as well as the previously observed enamel hypocalcification, may contribute to the increased dental caries that occur in juvenile hypothyroidism. Of note, the major salivary glands actively concentrate iodine from body fluids.9 This does not complicate radioiodine uptake or scanning tests of the thyroid because of

the relatively small amounts used, the lesser uptake by the salivary glands, and the anatomic separation of these organs and the thyroid gland. If radioiodine is used to destroy thyroid tissue in a patient with thyroid hyperplasia, adenoma, or carcinoma, however, the salivary glands also may be seriously damaged.10,11 The resulting xerostomia may be permanent, and rampant caries and periodontal destruction follow unless long-term, specific dental care is instituted in a timely manner. Appropriate care includes scrupulous oral hygiene, frequent professional dental cleanings, use of saliva substitutes for oral moistness and comfort, and the daily topical application of a fluoride gel onto the teeth. HYPOTHYROIDISM Cretinism is characterized by maxillary prognathism because the underdevelopment of the maxilla is less severe than that of the mandible.3,6 Radiographic examination often reveals hypocalcification of the jaws, and sometimes abnormal development of the sinuses or nonunion of the mandibular symphysis is seen. The characteristic facies consists of a concave nasal bridge and flared alae nasi; stiff facial expression, thickened lips and enlarged tongue, owing to doughy, nonpitting edema; and a mouth held partly open because of the lack of room for the tongue inside the underdeveloped mandible.3,6 The somewhat similar facies in Down syndrome arises from a disproportionately underdeveloped maxilla and a relative macroglossia that is not associated with edema.3 In both juvenile myxedema and cretinism, development of the teeth is retarded, frequently with hypocalcification, enamel hypoplasia (see Fig. 217-2), persistence of large pulp chambers, and open apical foramina.5 Eruption of both dentitions and shedding of the primary dentition are generally (but erratically) delayed.3,4,5 and 6 This, and the underdevelopment of the jaws, causes a malocclusion that may be severe and also may be complicated by spreading and flaring of the teeth secondary to pressure from the enlarged tongue. Hypothyroidism at any age seems to predispose affected patients to excessive dental caries, as well as to accelerated alveolar bone loss in both dentulous areas (periodontitis) and edentulous ridges (atrophy).5,6 The increased caries and periodontitis are related to hyposalivation and to the drying effects of mouth breathing caused by the enlarged, protruding tongue.6 Adults with myxedema also develop thickened lips and a swollen tongue (from edema). Pressure from the latter, in conjunction with an exacerbation of periodontitis, may cause spreading and splaying of the teeth.3,4 and 5 Generally, the earlier that childhood hypothyroidism is treated, the greater is the success in preventing or reversing orofacial maldevelopment, except for the affected parts of dentin and enamel that have completed all phases of development. Hypothyroid patients often are unable to tolerate prolonged dental procedures. Also, they usually have an exaggerated response to premedication with narcotics or barbiturates.5 HYPERTHYROIDISM In children, hyperthyroidism accelerates the development of the teeth and jaws, but maldevelopment is unusual.3,5 Malocclusion results occasionally when shedding of primary teeth and eruption of secondary teeth are disproportionately precocious to jaw growth. Usually, the teeth are normal in terms of size, morphology, and calcification. Periodontitis may begin at an unusually early age, and both caries and periodontitis reportedly are exacerbated in hyperthyroidism at any age. In severe hyperthyroidism, rapid bone demineralization, manifested radiographically as osteoporosis of the jaws and loss of alveolar bone in both dentulous and edentulous areas. People with hyperthyroidism are likely to be poor dental patients because they are unable to hold still for dental procedures and are likely to develop cardiac arrhythmias.3,4,6 The anxiety, stress, and trauma associated with dental treatment thus may precipitate a medical emergency in the dental office. In particular, the use of epinephrine and other vasoconstrictors in local anesthetics is contraindicated.6

CALCIUM AND PHOSPHORUS METABOLISM Hypoplasia of enamel and dentin, marked chronologic deviations in eruption and exfoliation of teeth, loss of radiodensity of the jawbones (especially the lamina dura), and the presence of giant cell lesions in the jaws may alert the dentist to previously undiagnosed disorders of calcium and phosphorus metabolism. Abnormalities of mineral metabolism that produce oral signs and symptoms include disturbances of particular nutritional factors (calcium, fluoride, vitamin D) or hormones (parathyroid hormones, adrenal corticosteroids), and renal function. NUTRITIONAL FACTORS Calcium deficiency may be an important factor in osteoporosis, which may accentuate alveolar bone loss in edentulous ridges.3 Fluoride deficiency has only one overt oral effect: a greatly increased susceptibility to dental caries. It also may contribute to osteoporosis when calcium intake is inadequate. Dental fluorosis, or mottled (hypoplastic) enamel, occurs with increasing frequency and severity when fluoride concentrations exceed 1.0 ppm in the drinking water if it is ingested while tooth development and calcification are in progress3 (see Chap. 131). When the water supply contains 75% of affected patients have six or more with a diameter of at least 1.5 cm), soft pedunculated or firmer nonpedunculated neurofibromas, and, rarely, large lobulated masses (plexiform neuromas) containing tumors arranged along peripheral nerves.32

FIGURE 218-16. Discrete, hyperpigmented macules (arrows), referred to as café-au-lait spots, are visible on the backs of the legs of a patient with neurofibromatosis.

Werner syndrome33 is an autosomal recessive condition that is associated with endocrine abnormalities, such as diabetes mellitus, hypogonadism, osteoporosis, metastatic subcutaneous calcifications, and impotence. The skin typically shows patches of scleroderma-like changes and marked, premature wrinkling (Fig. 218-17).

FIGURE 218-17. A 45-year-old man with Werner syndrome. Note the wrinkled, thin skin, graying and thinned hair, and overall aged appearance.

Many other metabolic disorders, both acquired and inherited, have prominent cutaneous manifestations. Although a detailed discussion of each of these conditions is beyond the scope of this chapter, salient features of several of these diseases are summarized in Table 218-1.34,35,36 and 37

TABLE 218-1. Selected Metabolic Disorders with Cutaneous Manifestations

PIGMENTARY ALTERATIONS AND THE ENDOCRINE SYSTEM As previously discussed, many endocrine disorders are characterized by generalized or focal increased or decreased pigmentation of the skin. Techniques are available to quantify skin pigmentation.38 For example, hyperpigmentation may occur with Addison disease, ACTH-producing tumors of the pituitary gland, paraneoplastic ACTH secretion, and POEMS syndrome (peripheral neuropathy, organomegaly, endocrine dysfunction, monoclonal gammopathy, and skin pigmentation).32,39 Hypopigmentation may accompany panhypopituitarism, hypogonadism (particularly in the male patient), and vitiligo (in polyglandular autoimmune deficiency). CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

Fine JD, Moschella SL. Diseases of nutrition and metabolism. In: Moschella SL, Hurley HJ, eds. Dermatology, 2nd ed. Philadelphia: WB Saunders, 1985:1422. Costello MJ. Eruptions of pregnancy. N Y State J Med 1941; 41:849. Katz SI, Hertz KC, Yaoita H. Immunopathology and characterization of the herpes gestationis factor. J Clin Invest 1976; 57:1434. Lawley TJ, Stingl G, Katz SI. Fetal and maternal risk factors in herpes gestationis. Arch Dermatol 1978; 114:552. Shornick JK, Stastny P, Gilliam JM. High frequency of histocompatibility antigens HLA-DR3 and DR4 in herpes gestationis. J Clin Invest 1981; 68:553. Lawley TJ, Hertz KC, Wade TR, et al. Pruritic urticarial papules and plaques of pregnancy. JAMA 1979; 241:1696. Bierman SM, Ackerman AB, Katz SI. Autoimmune progesterone dermatitis of pregnancy. Arch Dermatol 1973; 107:896. Spangler AS, Emerson K Jr. Estrogen levels and estrogen therapy in papular dermatitis of pregnancy. Am J Obstet Gynecol 1971; 110:534. Lang PC. Cutaneous manifestations of thyroid disease. Cutis 1978; 21:862.

9a.Pujol RM, Monmany J, Bague S, Alomar A. Graves' disease presenting as localized myxoedematous infiltration in a smallpox vaccination scar. Clin Exp Dermatol 2000; 25:132. 10. Cheung H, Nicoloff JT, Kamiel MB, et al. Stimulation of fibroblast biosynthetic activity by serum of patients with pretibial myxedema. J Invest Dermatol 1978; 71:12. 11. Jolliffe DS, Gaylarde PM, Brock AP, Sarkany I. Pretibial myxedema: stimulation of mucopolysaccharide production of fibroblasts by serum. Br J Dermatol 1979; 100:557. 11a.Suzuki H, Shimura H, Haraguchi K, et al. Exophthalmos, pretibial myxedema, osteoarthropathy syndrome associated with papillary fibroelastoma in the left ventricle. Thyroid 1999; 9:1257. 12. 13. 14. 15. 16.

Reunala T, Collin P. Diseases associated with dermatitis herpetiformis. Br J Dermatol 1997; 136:315. Zone JJ, Petersen MJ. Dermatitis herpetiformis. In: Thiers BH, Dobson RL, eds. Pathogenesis of skin disease. New York: Churchill Livingstone, 1986:159. August M, Wang J, Plante D, Wang CC. Complications associated with therapeutic neck radiation. J Oral Maxillofac Surg 1996; 54:1409. DePadova-Elder SM, Ditre CM, Kantor GR, et al. Candidiasis endocrinopathy syndrome. Arch Dermatol 1994; 130:19. Carney JA, Gordon H, Carpenter PC, et al. The complex of myxomas, spotty pigmentation and endocrine overactivity. Medicine 1985; 64:270.

16a.Watson JC, Stratakis CA, Bryant-Greenwood PK, et al. Neurosurgical implications of Carney complex. J Neurosurg 2000; 92:413. 17. 18. 19. 20. 21. 22. 23.

Kahan RS, Perez-Figaredo RA, Neimanis A. Necrolytic migratory erythema. Distinctive dermatosis of the glucagonoma syndrome. Arch Dermatol 1977; 113:792. Bassett ML, Haliday JW, Powell LW. Hemochromatosis—newer concepts: diagnosis and management. Disease-a-Month Series 1980; 26:1. Feder JN, Gnirke A, Thomas W, et al. A novel MHC class I-like gene is mutated in patients with hereditary hemochromatosis. Nat Genet 1996; 13:399. Bonkovsky HL, Poh-Fitzpatrick M, Pimstone N, et al. Porphyria cutanea tarda, hepatitis C, and HFE gene mutations in North America. Hepatology 1998; 27:1661. Matsuoka LY, Wortsman J, Gavin JR, Goldman J. Spectrum of endocrine abnormalities associated with acanthosis nigricans. Am J Med 1987; 83:719. Darling TN, Skarulis MC, Steinberg SM, et al. Multiple facial angiofibromas and collagenomas in patients with multiple endocrine neoplasia type 1. Arch Derm 1997; 133:853. Pack S, Turner ML, Zhuang Z, et al. Cutaneous tumors in patients with multiple endocrine neoplasia type 1 show allelic deletion of the MEN1 gene. J Invest Derm 1998; 110:438.

23a.Grudeva-Popuva J, Dobrev H. Biomechanical measurement of skin distensibility in scleredema of Buschke associated with multiple myeloma. Clin Exp Dermatol 2000; 25:247. 24. Gipstein RM, Coburn JW, Adams DA, et al. Calciphylaxis in man: a syndrome of tissue necrosis and vascular calcification in 11 patients with chronic renal failure. Arch Intern Med 1976; 136:1273. 25. Mehregan DA, Winkelmann RK. Cutaneous gangrene, vascular calcification, and hyperparathyroidism. Mayo Clin Proc 1989; 64:211. 26. De Graaf P, Ruiter DJ, Scheffer E, et al. Metastatic skin calcification: a rare phenomenon in dialysis patients. Dermatologica 1980; 161:28. 27. Angelis M, Wong LL, Myers SA, Wong LM. Calciphylaxis in patients on hemodialysis: a prevalence study. Surgery 1997; 122:1083. 28. Zouboulis CC, Blume-Peytavi U, Lennert T, et al. Fulminant metastatic calcinosis with cutaneous necrosis in a child with endstage renal disease and tertiary hyperparathyroidism. Br J Dermatol 1996; 135:617. 29. Van Dop C, Bourne HR. Pseudohypoparathyroidism. Annu Rev Med 1983; 34:259. 30. Doty RL. Olfactory dysfunction in type I pseudohypoparathyroidism: dissociation from Gsa protein deficiency. J Clin Endocrinol Metab 1997; 82:247. 31. Hofman KJ. Diffusion of information about neurofibromatosis type 1 DNA Testing. Am J Med Genet 1994; 49:299. 32. Martuza RL, Eldridge R. Neurofibromatosis 2. N Engl J Med 1988; 318:684. 33. Bauer EA, Uitto J, Tau ML, Holbrook KA. Werner's syndrome: evidence for preferential regional expression of a generalized mesenchymal cell defect. Arch Dermatol 1988; 124:90. 34. Goldsmith LA. Tyrosine-induced skin disease. Br J Dermatol 1978; 98:119. 35. Fine JD, Wise TG, Falchuk KH. Zinc in cutaneous disease and dermatologic therapeutics. In: Moschella SL, ed. Dermatology update: reviews for physicians. New York: Elsevier, 1982:299. 36. Steinbach HL, Russell W. Measurement of the heel pad as an aid to diagnosis of acromegaly. Radiology 1964; 82:418. 37. Case Records of the Massachusetts General Hospital. Case 10-1987. N Engl J Med 1987; 316:606. 38. Bech-Thomsen N, Angelo HR, Wulf HG. Arch Dermatol 1994; 130:464. 39. Schulz W, Domenico D, Nand S. POEMS syndrome associated with polycythemia vera. Cancer 1989; 63:1175.

CHAPTER 219 PARANEOPLASTIC ENDOCRINE SYNDROMES Principles and Practice of Endocrinology and Metabolism

CHAPTER 219 PARANEOPLASTIC ENDOCRINE SYNDROMES KENNETH L. BECKER AND OMEGA L. SILVA Definition Pathogenesis Criteria General Principles Specific Paraneoplastic Endocrine Syndromes Paraneoplastic Growth Hormone–Releasing Hormone Syndrome Paraneoplastic Corticotropin-Releasing Hormone Syndrome Paraneoplastic Adrenocorticotropic Hormone Syndrome Paraneoplastic Arginine Vasopressin Syndrome Paraneoplastic Hypercalcemia Paraneoplastic Osteomalacia Paraneoplastic Secretion of Human Chorionic Gonadotropin Paraneoplastic Hypoglycemia Paraneoplastic Hyperreninism Paraneoplastic Erythrocytosis Combined Paraneoplastic Endocrine Syndromes Paraneoplastic Secretion of Other Known Peptide Hormones with no Associated Endocrine Syndrome Suspected Paraneoplastic Endocrine Syndromes Hypertrophic Osteoarthropathy Acanthosis Nigricans Miscellaneous Possibly Paraneoplastic Endocrine Syndromes Hormone-Mediated Effect of Neoplasia on the Body Chapter References

DEFINITION In addition to being a local disorder of tissue growth and a source of potential metastases, cancers often have important systemic metabolic manifestations. These remote biologic effects sometimes can dominate the other clinical effects of the malignant process.1,2 As discussed in other chapters of this text, many humoral manifestations are attributable to a neoplasm of a tissue that normally is the predominant site of production of a hormone (e.g., pituitary prolactinoma and galactorrhea; adrenal cortical adenoma and Cushing syndrome; thyroid adenoma and hyperthyroidism). However, if clinical manifestations are the result of hormones secreted by tumors emanating from tissues that normally do not secrete them into the blood at significant levels, they are termed paraneoplastic endocrine syndromes. Despite the extraordinary cellular differentiation and organization seen in humans, most normal tissues retain the ability to secrete many hormones, albeit some more efficiently than others (see Chap. 175). This inherent secretory capacity is shared by the neoplasms derived from these tissues. A cancer often reflects the innate humoral characteristics and secretory potentials of its cell of origin. Thus, conceptually, the elaboration of a hormone by a neoplasm is not really “ectopic”; it is “eutopic,” but quantitatively abnormal.

PATHOGENESIS Some tumors that produce paraneoplastic endocrine syndromes (e.g., small cell lung cancer,3 carcinoid tumor, Merkel cell tumor) arise from the diffuse neuroendocrine system (see Chap. 175). Although the primordial tissues engendering such neoplastic lesions initially were thought to have a common embryonic origin, this theory is no longer tenable. The neuroendocrine cancers do not share common embryonic precursors. Also, many paraneoplastic syndromes originate from tumors that are not neuroendocrine. In attempting to explain why a tumor produces hormones, investigators have stressed the fact that the DNA complement within each normal somatic cell is identical; that is, in each specific cell, some functions normally are repressed (e.g., a gene that codes for a certain hormone). Therefore, hormone-secreting cancers are viewed as exhibiting selective derepression of the genome, allowing synthesis of the hormone. Others attribute the hormonal elaboration by cancer to a failure of differentiation of undifferentiated, totipotential stem cells.

CRITERIA Certain traditional criteria have been proposed for accepting a group of symptoms and signs as a paraneoplastic endocrine syndrome4 (Table 219-1). These idealized criteria have not been met for several bona fide paraneoplastic syndromes, however, and often are unnecessary for making the diagnosis.

TABLE 219-1. Criteria for Defining a Paraneoplastic Endocrine Syndrome*

GENERAL PRINCIPLES A common feature of most paraneoplastic endocrine syndromes is the elaboration of peptide hormones. De novo steroid synthesis by cancer usually requires adrenal, gonadal, or placental tissue, and de novo thyroid hormone synthesis requires thyroidal or teratomatous tissue (see Chap. 204). Moreover, biogenic amines (i.e., histamine, serotonin) play significant roles in some paraneoplastic manifestations. Prostaglandin secretion also may be important (see Chap. 172). When encountering most paraneoplastic endocrine syndromes, the physician already is conscious of the associated neoplasm. Nonetheless, sometimes the syndrome precedes the diagnosis of the tumor; indeed, marked endocrine effects may mask the tumor, averting attention to inappropriate therapeutic approaches. Although a paraneoplastic endocrine syndrome might constitute an interesting but relatively harmless occurrence (e.g., acanthosis nigricans, hypertrophic osteoarthropathy), it also might contribute directly to the patient's premature demise (e.g., hypercalcemia, syndrome of inappropriate antidiuresis [SIAD]). Although tumors sometimes produce the same hormone as related normal cells, they often produce and secrete much greater quantities.5 Moreover, such tumors frequently secrete other hormones that may be secreted by their normal cellular counterparts in minute amounts. The clinical effects of the poly-hormonal potential of many of these tumors are unknown.

Generally, hormones secreted by cancers are not unique in chemical structure. Nonetheless, the distribution of the molecular forms of the secreted peptide frequently is abnormal. In particular, precursor forms with high molecular weight often predominate. This probably reflects a deficient or incomplete posttranslational modification; or it may signify an alternative means of secretion.6 Because these precursors often have less or no bioactivity, a clinical syndrome may not occur, despite an extraordinarily high level of radioimmunoassayable hormone in the serum. Most tumors producing such peptides remain clinically silent. Furthermore, the finding of high serum hormone levels emanating from a tissue other than a classic endocrine gland is not a specific indicator of neoplasia. Irritated, hyperplastic, or premalignant tissues or lesions also may secrete increased levels of peptide hormones (e.g., in chronic obstructive pulmonary disease, chronic bronchitis of smokers, regional ileitis, and ulcerative colitis). Morphologically, in contrast to normal, anatomically discrete endocrine organs, hormone-secreting neoplasms often do not possess a highly structured and coordinated neural control. Metabolically, the neoplastic cells in most paraneoplastic syndromes often do not respond to the usual physiologic control mechanisms that modulate normal hormone secretion (i.e., physiologic secretagogues, feedback suppression). Although exceptions exist, this relative autonomy may be of considerable diagnostic use. Last, the hormonal secretion by cancer may serve as an important biomarker for its presence, its response to therapy, and its relapse.

SPECIFIC PARANEOPLASTIC ENDOCRINE SYNDROMES This chapter discusses the paraneoplastic endocrine syndromes of certain cancers, including those in which the humoral mediator is suspected but not yet identified. Brief mention also is made of possible humoral syndromes that arise as a result of the effect of neoplasia on noncancerous tissue. Hormone-secreting tumors of the pancreas and of the neuroendocrine cells of the gut (e.g., insulinoma, gastrinoma) are discussed elsewhere (see Chap. 158 and Chap. 220). Although several humoral syndromes associated with other neuroendocrine tumors are mentioned, they also are discussed in other chapters. PARANEOPLASTIC GROWTH HORMONE–RELEASING HORMONE SYNDROME Although far less common than hypothalamic-pituitary acromegaly, the syndrome of acromegaly that is secondary to the tumoral secretion of growth hormone–releasing hormone (GHRH) is a fascinating and instructive example of a paraneoplastic endocrine syndrome.7,8 Normally, GHRH circulates at low or undetectable levels, although appreciable levels are found in the hypothalamic-hypophysial portal system. Some tumors secrete large amounts of GHRH, causing acromegaly. The secreted GHRH is the same as that in the human hypothalamus; both the 44- and 40-amino-acid hormones are found.9 Two lesions, in particular, may produce the syndrome: bronchial carcinoid and pancreatic islet cell tumor. In addition, GHRH is found in some carcinoid tumors involving the gastrointestinal tract and thymus, as well as in pheochromocytoma, medullary thyroid cancer, and small cell lung cancer. Although it is also in the blood of some patients with these tumors, however, it does not necessarily cause symptoms and signs of acromegaly. Patients with the full-blown syndrome of paraneoplastic GHRH secretion have the classic appearance of acromegaly (see Chap. 12), but often it is of more rapid onset. The serum levels of GHRH frequently exceed 200 pg/mL. The high serum levels of the hormone cause hyperplasia of the pituitary somatotropes, with consequent hypersecretion of growth hormone. As in hypothalamic-pituitary acromegaly, serum growth hormone levels are not suppressed after the administration of oral glucose; however, in contrast to classic acromegaly, patients with the paraneoplastic GHRH syndrome may have a markedly increased serum growth hormone response to insulininduced hypoglycemia.10 Usually, the sella turcica is not enlarged, and examination by computed tomography or nuclear magnetic resonance imaging yields normal results. Occasionally, however, an enlarged sella is seen11; some cases apparently eventuate in pituitary tumors. Pituitary surgery is not indicated in this syndrome. If feasible, the cancerous lesion should be removed. If surgery is successful, it should cause regression of the acromegalic syndrome. Medical treatment with bromocriptine often successfully suppresses the growth hormone level; the somatostatin analog octreotide acetate also has been useful in this regard.12 Paraneoplastic secretion of growth hormone is very rare. A case has been described in which acromegaly was apparently due to secretion of this hormone by a non-Hodgkin lymphoma.12a PARANEOPLASTIC CORTICOTROPIN-RELEASING HORMONE SYNDROME Corticotropin-releasing hormone (CRH) has been found in bronchial carcinoid tumors and in small cell carcinoma of the lung. Rarely, high serum levels of this hormone stimulate the pituitary gland to produce excess adrenocorticotropic hormone (ACTH). In such cases, the increased ACTH levels cause bilateral adrenal hyperplasia, and the resultant hypercortisolism results in Cushing syndrome13,14 (see Chap. 75). PARANEOPLASTIC ADRENOCORTICOTROPIC HORMONE SYNDROME The paraneoplastic ACTH syndrome (or ectopic ACTH syndrome) is caused by the secretion of ACTH by a nonpituitary neoplasm, which results in bilateral adrenal hyperplasia and manifestations of Cushing syndrome. This condition is more common than Cushing disease. Two-thirds of the cases of paraneoplastic ACTH syndrome are attributable to bronchogenic cancer. The principal offender is small cell cancer of the lung; occasionally, pulmonary adenocarcinoma also has been implicated. Another common cause is the relatively benign bronchial carcinoid tumor.15 Other, less frequent causes are thymic tumor (usually benign, and often of carcinoid histology), islet cell carcinoma of the pancreas,16 medullary thyroid cancer, pheochromocytoma, and colon carcinoma. Manifestations of the syndrome are influenced not only by the level of ACTH secretion but also by the bioactivity of the hormone. Many nonpituitary tumors that secrete ACTH typically remain biologically silent because of the secretion of bioinactive precursors of ACTH—big ACTH—that make up preproopiomelanocortin and its related products, including multiple immunologic forms of b-endorphin.17 Simultaneous assays of receptor-active ACTH and immunoactive ACTH have confirmed the inactivity of many of the secreted hormones. Thus, although approximately one-third of patients with small cell cancer of the lung have increased serum ACTH levels by radioimmunoassay, only 1% to 2% have hypercortisolism. Clinically, because of the close association with small cell lung cancer, the fact that many patients with the syndrome are men older than 40 years who abuse tobacco is not surprising. (This is in direct contrast to Cushing disease, which occurs in younger persons and has a strong female predilection.) The tumor type is important; in patients with slow-growing bronchial carcinoid often the disease onset is gradual and insidious, the duration of symptoms may span months to several years, and the classic cushingoid features (facial plethora, moon facies, easy bruising, hirsutism, truncal obesity, and atrophy of the extremities) develop over a prolonged period.18 In such patients, the tumor may be occult and difficult to localize; however, computerized tomography often is helpful.19,20 Because of the increased secretion of ACTH and its precursor (proopiomelanocortin), the level of melanocyte-stimulating hormone (MSH) activity (a-MSH within the ACTH molecule, and b-MSH within the b-lipotropin molecule) increases (see Chap. 14). Consequently, these patients may manifest marked hyperpigmentation that is similar in appearance to that encountered in Addison disease (Fig. 219-1).

FIGURE 219-1. Patient with paraneoplastic adrenocorticotropic hormone syndrome secondary to small cell cancer of the lung. Note the darkening of the skin, the

fullness of the cheeks and the supraclavicular fossae, and the cervicodorsal hump.

Patients with more aggressive malignant disease (e.g., small cell lung cancer [see Fig. 219-1], pancreatic adenocarcinoma) often do not have centripetal obesity. Commonly, the onset is acute, the symptoms have been present for several weeks at the time of examination, and obvious, rapid weight loss occurs. Often, the serum ACTH level is even higher than that in patients with carcinoid tumor, and the resultant hyperpigmentation is more pronounced. Other manifestations may include edema, muscle weakness, hypertension, severe hypokalemic alkalosis, and hyperglycemia. A clearcut parallelism between clinical findings, tumor type, and duration of the neoplasm is not always present, however. Laboratory studies of patients with the paraneoplastic ACTH syndrome may reveal normal or decreased serum potassium levels. Hypokalemia occurs much more commonly in this syndrome than in either pituitary Cushing disease or Cushing syndrome secondary to adrenal adenoma. Any unprovoked hypokalemia (i.e., in a patient not taking diuretics or laxatives) merits further investigation. Metabolic alkalosis and hyperglycemia may be present. The demonstration of hypercortisolism is essential to the diagnosis. Commonly, the serum cortisol level is extremely high—higher than that seen in most patients with Cushing disease (although the serum cortisol elevation in patients with carcinoid tumor often is moderate). The normal diurnal cortisol rhythmicity (see Chap. 6) is abolished, and a large day-to-day variability in the serum cortisol level often is seen. Rarely, periodic hypersecretion may be encountered, with intervals of days to weeks of normal levels interspersed among periods marked by increased values. Increased levels of serum dehydroepiandrosterone sulfate and urinary 17-ketosteroids may help to distinguish between Cushing syndrome caused by paraneoplastic ACTH production and Cushing syndrome caused by unilateral adrenal adenoma, the latter of which is characterized by normal values. Although exceptions exist, many patients with the paraneoplastic ACTH syndrome do not exhibit suppression of serum cortisol or urinary 17-hydroxycorticosteroid levels with daily administration of 8 mg of dexamethasone orally (2 mg every 6 hours for 2 days). Most patients with Cushing disease, however, do exhibit such suppression (see Chap. 74 and Chap. 75). In contrast to the marked responsiveness of serum 11-deoxycortisol and urinary 17-hydroxycorticosteroid levels to the administration of metyrapone that usually occurs in patients with Cushing disease, those with the paraneoplastic ACTH syndrome generally are unresponsive. In the paraneoplastic ACTH syndrome, serum ACTH levels usually are high, and several other hormones, such as calcitonin, arginine vasopressin (AVP), somatostatin, or vasoactive intestinal peptide, may simultaneously be produced. (The b-MSH–secreting tumor does not constitute a clinical entity that is distinct from the paraneoplastic ACTH syndrome.) The injection of CRH may induce a further increase in serum ACTH levels in Cushing disease, but not in the paraneoplastic ACTH syndrome.21 Localization studies are essential to the workup of these patients. Chest radiography, sputum cytology, and bronchoscopy often facilitate the diagnosis of small cell lung cancer. Results of computed tomographic and magnetic resonance imaging studies of the pituitary are normal. Similar studies of the chest or abdomen may reveal the tumor, and imaging of the abdomen often demonstrates bilateral adrenal hyperplasia. (In one instance, a pheochromocytoma that secreted ACTH was diagnosed in a patient with bilateral enlarged adrenal glands, one of which also contained a focal mass.) Bilateral, simultaneous inferior petrosal sinus sampling of patients with paraneoplastic ACTH syndrome generally reveals no unilateral ACTH gradient and demonstrates ACTH levels that are less than those derived from venous catheterization samples of the tumor effluent (e.g., mediastinal veins). The simultaneous administration of CRH further increases the utility of the procedure.22 Somatostatin-receptor scintigraphy may be used to demonstrate the location of the tumor-producing lesion.23 Occasionally, percutaneous needle aspiration of tumor tissue, with assay of the intracellular ACTH, has confirmed the diagnosis.24 The ideal therapy for the paraneoplastic ACTH syndrome is extirpation of the neoplasm. If this is not feasible, some patients respond, albeit transiently, to chemotherapy. If the neoplasm cannot be treated successfully, adrenocortical hyperactivity can be mitigated with drugs, such as mitotane (o,p'-DDD), aminoglutethimide, metyrapone, or ketoconazole.25 Such therapy may lead to prolonged control of the hypercortisolemia.26 In addition, such drugs can be administered before surgery in patients whose primary neoplasms are resectable, but whose Cushing syndrome initially might make an operation too hazardous. PARANEOPLASTIC ARGININE VASOPRESSIN SYNDROME SIAD is characterized by the excretion of a hypertonic urine despite an expanded extracellular volume. SIAD can arise from centralmechanisms (acute or chronic disorders of the central nervous system, drugs), or from peripheralmechanisms, in which case the secretion of AVP by a neoplasm is a common offender.27 This paraneoplastic AVP syndrome also has been termed tumoral hyponatremia28 (see Chap. 25 and Chap. 27). AVP and oxytocin, as well as the carrier protein neurophysin, are found within such cancers, and cell cultures of the tumor also secrete the precursor peptide, propressophysin. In 80% of cases, the tumor causing SIAD is small cell cancer of the lung29; although it often is an incidental finding, as many as one-third of patients with this type of cancer have some degree of the syndrome. Other tumors that have been known to cause SIAD include those involving the pancreas, thymus, and breast. Occasionally, a cancerous lesion that does not directly secrete AVP can produce this syndrome through a central mechanism (e.g., metastases to the brain). The symptoms and signs of SIAD are attributable to water intoxication and hyponatremia, and may range in severity from having no effect to being life-threatening. They may include nausea, weakness, and central nervous system effects, such as confusion or obtundation, all of which may be easily misconstrued as being caused by the malignant disease. The central nervous system effects may progress to convulsions and, sometimes, frank coma. Laboratory studies reveal hyponatremia (and consequent hypoosmolality: 200 pg/mL). On the whole, a stool volume of 70% of the patients have metastatic disease (usually in the liver and regional lymph nodes). The most characteristic feature of the glucagonoma syndrome is the presence of a necrolytic migratory erythematous rash (Fig. 220-2).14 The rash usually starts in the groin and perineum and then gradually migrates to the distal extremities. Typically, the initial lesions are erythematous macules, which become raised and bullous. Then the lesions break down and heal, often leaving a residual area of hyperpigmentation. The rash is intensely painful and pruritic, and secondary bacterial and fungal infections are common. The underlying cause of the rash is unknown, but several factors such as direct action of glucagon on the skin, amino-acid and fatty acid deficiency, and zinc deficiency have been implicated in its etiology. The rash is commonly associated with mucosal involvement, which results in stomatitis, cheilitis, and glossitis. Cachexia is a common feature of glucagonoma and may mislead the physician into believing that a more aggressive tumor, such as pancreatic

carcinoma, is responsible for the patient's symptoms. As glucagon opposes the effects of insulin on blood glucose homeostasis, impaired glucose tolerance, which usually results in mild diabetes, is also a manifestation of the glucagonoma syndrome. Other manifestations of glucagonoma include normocytic normochromic anemia, dystrophy of the nails, diarrhea, a tendency to venous thrombosis and pulmonary embolism, and neuropsychiatric symptoms. Paraneoplastic syndromes, such as optic atrophy, have also been reported in association with glucagonoma.

FIGURE 220-2. Truncal rash in a patient with a glucagonoma. (From Bloom SR, Polak JM. Glucagonoma syndrome. Am J Med 1987; 82[Suppl 5B]:25.)

The diagnosis of glucagonoma depends greatly on clinical suspicion. Once the diagnosis is considered, measurement of a highly elevated plasma glucagon value after an overnight fast is confirmatory. Plasma glucagon levels may be increased in patients with prolonged fasting, renal and hepatic failure, diabetic ketoacidosis, or therapy with oral contraceptives or danazol, and in those with trauma, burns, sepsis, or Cushing syndrome. These conditions are easily distinguishable from glucagonoma, however, and rarely result in markedly raised glucagon levels. A rare condition associated with elevated glucagon levels is familial hyperglucagonemia.15 Cosecretion of a second gut hormone is common, and one-fifth of glucagonoma patients also have elevated gastrin levels, which may cause acid hypersecretion. Elevated plasma insulin, pancreatic polypeptide, VIP, and urinary 5-hydroxyindoleacetic acid levels have all been observed with glucagonoma. Secretion of other cleavage products of preproglucagon by these tumors may result in gastrointestinal mucosal hypertrophy. The majority of glucagonomas are large and metastatic at presentation. Tumor localization can be achieved with ultrasonography, CT, or visceral angiography (see Chap. 159). Somatostatin-receptor scintigraphy is most useful for determining the extent of metastatic disease.15a Endoscopic ultrasonography is sensitive in the detection of pancreatic primary tumors, but limited penetration reduces the detection of distant spread. Localized solitary tumors may require multiple techniques for localization. Localized, solitary glucagonomas should be surgically excised to aim for a cure, whereas palliative measures are used for metastatic disease. The glucagonoma rash responds to oral and topical zinc, and somatostatin analogs. Longer acting somatostatin analogs and longer lasting preparations of octreotide are likely to be increasingly used.16,17 With time, the tumor may become less responsive to such therapy, and increasing doses and/or surgery or induced embolization to reduce tumor bulk may be required. Control of cachexia is extremely difficult and may require dietary intervention such as consumption of a high-protein diet. Glucose intolerance may require insulin therapy. Aspirin has been advocated for prevention of thrombotic episodes, and anticoagulant therapy is used for patients with proven thrombosis. Psychiatric symptoms, such as psychosis or depression, require appropriate psychiatric assessment and treatment. The available palliative measures directed at the tumor and its metastases include surgical debulking of the tumor, hepatic tumor embolization, and cytotoxic chemotherapy. Glucagonomas are relatively insensitive to chemotherapeutic agents, but occasionally a patient may benefit from this treatment modality.

SOMATOSTATINOMA Somatostatinomas18 are extremely rare tumors with an annual incidence of ~1 in 40 million. They occur mostly in the pancreas but can also arise in the duodenum. Duodenal somatostatinomas may be associated with neurofibromatosis type 1 (von Reckling-hausen disease) and pheochromocytoma.19 Duodenal somatostatinomas usually present early with obstructive symptoms and only rarely result in the tumor syndrome. Pancreatic tumors present late, with hepatic metastases, and often with the tumor syndrome characterized by the triad of cholelithiasis, diabetes,and steatorrhea. Hypoglycemia may occasionally occur and is likely to be caused by larger molecular forms of somatostatin. 20,21 Other features include anemia, hypochlorhydria, postprandial fullness, and weight loss. Occasionally, somatostatinomas may secrete adrenocorticotropic hormone (ACTH), resulting in Cushing syndrome. The diagnosis of somatostatinoma is confirmed by detection of highly elevated plasma somatostatin levels. Tumor localization uses the same techniques as described for other endocrine tumors of the gut. Treatment is mainly surgical. The palliative measures used are similar to those used for other gastrointestinal endocrine tumors.

PPoma Many types of pancreatic endocrine tumors also secrete pancreatic polypeptide (PP).22,23 and 24 Elevated plasma PP levels can, therefore, be used as a marker of gastrointestinal endocrine tumors, particularly VIPomas (see Chap 182). Rare tumors produce only PP, but this does not result in any unique clinical or biochemical features. Histologically, the tumors are usually composed of mixed cell types, one of which can be immunocytochemically identified as producing PP.

NEUROTENSINOMA Neurotensin is a 13-amino-acid peptide released from N cells in the small intestine. Neurotensin immunoreactivity has also been detected in enteric neurons. The hormone is produced by a small number of pancreatic endocrine tumors, which usually also produce VIP.25 When infused in humans, neurotensin causes increased watery secretions from the small intestine and frequent defecation.26 Neurotensin may therefore contribute to the clinical features of the VIPoma syndrome. When neurotensin is the sole product of an endocrine tumor, however, no clinical features occur. (Presumably, escape occurs from long-continued elevations of plasma neurotensin.) Neurotensin is also detected in non-endocrine tumors in the gastrointestinal tract, suggesting that it may have growth regulatory functions in the pancreas and colon.

ENTEROGLUCAGONOMA In the intestinal mucosa, proglucagon-derived peptides are synthesized and released by L cells of the terminal ileum and colon.27 Proglucagon undergoes tissue-specific posttranslational processing in the pancreas and intestine. In the pancreas, the end products of proglucagon processing are glucagon, glicentin-related pancreatic polypeptide (GRPP), and a large major fragment. In the intestine, the products are glucagon-like peptide-1 (GLP-1), glucagon-like peptide-2 (GLP-2), and glicentin (or enteroglucagon). Interest in these peptides as intestinal growth factors began with the observation that intestinal villous hypertrophy occurred in a patient with a tumor that was secreting proglucagon-derived peptides. Resection of the tumor resulted in the normalization of the previously increased blood levels of these peptides and regression of intestinal villous hypertrophy (see Chap. 160).28 Further evidence was provided by the fact that, after small bowel resection, proglucagon messenger RNA expression was increased in the remnant intestine. Other studies have reported that overexpression of the glucagon gene or exogenously administered proglucagon-derived peptides in rodents are associated with bowel growth and regeneration.29 Although initially the belief was that enteroglucagon was the growth factor released by the tumor described (hence, the name “enteroglucagonoma”), GLP-2, a 33-amino-acid peptide, appears to be the mediator of epithelial cell proliferation in the gut.30

OTHER ENDOCRINE TUMORS OF THE GUT Gastrointestinal endocrine tumors may secrete PTHrP, resulting in hypercalcemia. The hypercalcemia can be controlled by regular infusion of bisphosphonates, hepatic tumor embolization, and surgical debulking. Tumors have been described that secrete growth hormone–releasing hormone (GHRH), causing acromegaly, and ACTH, causing Cushing syndrome. These may occur either alone or in association with other gastrointestinal hormone tumor syndromes. Numerous other peptides have been detected either in the plasma or in tissue from endocrine tumors of the gut. These include neuropeptide Y (NPY), neuromedin-B, calcitonin gene–related peptide (CGRP), motilin, and bombesin, but these are not associated with specific clinical syndromes.

TUMOR MARKERS

No generally useful tumor marker exists for gastrointestinal endocrine tumors. Some interest has been shown in the use of chromogranins and neuron-specific enolase (NSE) as markers. Histochemically, chromogranins and NSE have been used for some time in the immunocytochemical identification of neuroendocrine tumors.31 In neuroendocrine tumors, a correlation is seen between tumor burden and circulating chromogranin A levels; the highest levels of chromogranin A are seen in metastatic disease.32 Small gastrinomas can result in high plasma chromogranin A levels, because gastrin causes hyperplasia of enterochromaffin-like cells that also secrete chromogranin A. Severe renal failure can result in elevated plasma chromogranin A levels comparable to levels seen with neuroendocrine tumors. Chromogranin A is most useful as a marker of nonfunctioning tumors. The peptide GAWK,33 the product of chromogranin B, is also a useful marker for neuroendocrine tumors, as are secretoneurins.33a The role of the above tumor markers remains to be defined, but at present they are not routinely used in the diagnosis and follow-up of patients.

SOMATOSTATIN THERAPY Somatostatin analogs are commonly used in the treatment of gut endocrine tumors. Somatostatin has widespread inhibitory effects in the gastrointestinal tract.34 It is used to reduce the secretion of peptides from endocrine tumors and also has anti-neoplastic actions.35 Somatostatin has a half-life of 50% of the liver parenchyma is replaced by tumor, embolization may precipitate fulminant hepatic failure. Other contraindications to tumor embolization include blood coagulation disorders, intercurrent infection, and end-stage disease. Embolization can result in the release of numerous vasoactive peptides from the tumor and, therefore, a hypotensive crisis. Patients are prepared for the procedure by prehydration. An octreotide infusion is used to block the effects of the peptides. Intravenous aprotinin and prophylactic broad-spectrum antibiotics are also administered to minimize the risks of the procedure. The procedure may result in fever, malaise, nausea, vomiting, abdominal pain, and paralytic ileus. Fever necessitates appropriate microbiologic investigations; in such cases, the development of a hepatic abscess should be excluded by abdominal ultrasonographic examination. Fortunately, complications of tumor embolization are rare in the hands of an experienced operator. Chemoembolization, using agents such as doxorubicin and iopamidol, has been advocated in the treatment of metastatic neuroendocrine tumors. Morbidity is reported to be less than for cytotoxic chemotherapy or for embolization alone. CHAPTER REFERENCES 1. Zollinger RM, Ellison EH. Primary peptic ulceration of the jejunum associated with islet cell tumors of the pancreas. Ann Surg 1955; 142:709. 2. Polak JM, Bloom SR. Review: The enterochromaffin-like cell, intragastric acidity and the trophic effects of plasma gastrin. Aliment Pharmacol Ther 1988; 2:291. 3. Walsh JH. Gastrin. In: Walsh JH, Dockray GJ, eds. Gut peptides. New York: Raven Press, 1994:75. 3a. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Goebel SU, Heppner C, Burns AL, et al. Genotype/phenotype correlation of multiple endocrine neoplasia type 1 gene mutations in sporadic gastrinomas. J Clin Endocrinol Metab 2000; 85:116. Jensen RT. Gastrointestinal endocrine tumours. Gastrinoma. Baillière's Clin Gastroenterol 1996; 10(4):673. Modlin IM, Tang LH. Approaches to the diagnosis of gut neuroendocrine tumors: the last word (today). Gastroenterology 1997; 112:583. Hammond PJ, Jackson JA, Bloom SR. Localization of pancreatic endocrine tumors. Clin Endocrinol 1994; 40:3. Jensen RT. Management of the Zollinger-Ellison syndrome in patients with multiple endocrine neoplasia type 1. J Intern Med 1998; 243:477. Verner JV, Morrison AB. Islet cell tumor and a syndrome of refractory diarrhea and hypokalemia. Am J Med 1958; 25:374. Park SK, O'Dorisio MS, O'Dorisio TM. Gastrointestinal endocrine tumours. Vasoactive intestinal polypeptide-secreting tumours: biology and therapy. Baillières Clin Gastroenterol 1996; 10(4):673. Bloom SR, Polak JM, Pearce AGE. Vasoactive intestinal peptide and watery-diarrhoea syndrome. Lancet 1973; 2:14. Dockray G. Vasoactive intestinal polypeptide and related peptides. In: Walsh JH, Dockray GJ, eds. Gut peptides. New York: Raven Press, 1994:447. O'Dorisio TM. VIP and watery diarrhea. In Bloom SR, ed. Gut hormones. Edinburgh: Churchill Livingstone, 1978:581. Frankton S, Bloom SR. Gastrointestinal endocrine tumours. Glucagonomas. Baillières Clin Gastroenterol 1996; 10(4):673. Bloom SR, Polak JM. Glucagonoma syndrome. Am J Med 1987; 82:25. Boden G, Owen OE. Familial hyperglucagonemia: an autosomal dominant disorder. N Engl J Med 1977; 296:534.

15a. Johnson DS, Coel MN, Bornemann M. Current imaging and possible therapeutic management of glucagonoma tumors. Clin Nuclear Med 2000; 25:120. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.

Arnold R, Frank M. Gastrointestinal endocrine tumours. Gastrointestinal endocrine tumours: medical management. Baillières Clin Gastroenterol 1996; 10(4):737. Trautman ME, Neuhaus C, Lenze H, et al. The role of somatostatin analogs in the treatment of endocrine gastrointestinal tumors. Horm Metab Res 1995; 27:24. Kreijs GJ, Orci L, Conlon JM, et al. Somatostatinoma syndrome. N Engl J Med 1979; 301:285. Griffiths DF, Williams GT, Williams ED. Duodenal carcinoid tumors, phaeochromocytoma and neurofibromatosis: islet cell tumor, phaeochromocytoma and the von Hippel-Lindau complex: two distinctive neuroendocrine syndromes. Q J Med 1987; 64:769. Penman E, Lowry PJ, Wass JAH, et al. Molecular forms of somatostatin in normal subjects and patients with pancreatic somatostatinomas. Clin Endocrinol (Oxf) 1980; 12:611. Bloom SR, Polak JM. Somatostatin. BMJ 1987; 295:288. Vinik AI, Strodel WE, Eckhauser FE, et al. Somatostatinomas, PPomas, neurotensinomas. Semin Oncol 1987; 14:263. Adrian TE, Lettenthal LO, Williams SJ, Bloom SR. Secretion of pancreatic polypeptide in patients with pancreatic endocrine tumors. N Engl J Med 1986; 315:287. Polak JM, Bloom SR, Adrian TE, et al. Pancreatic polypeptide in insulinomas, gastrinomas, VIPomas and glucagonomas. Lancet 1976; 1:328. Blackburn AM, Bryant MG, Adrian TE, Bloom SR. Pancreatic tumors producing neurotensin. J Clin Endocrinol Metab 1981; 52:820. Calam J, Unwin R, Peart WS. Neurotensin stimulates defaecation. Lancet 1983; 1:737. Holst JJ. Enteroglucagon. Ann Rev Physiol 1997; 59:257. Bloom SR. An enteroglucagon tumor. Gut 1972; 13(7):520. Drucker DJ, Erlich, P, Asa SL, et al. Induction of intestinal epithelial proliferation by glucagon-like peptide 2. Proc Natl Acad Sci U S A 1996; 93 (15):7911.

30. 31. 32. 33.

Hussain MA. A biological function for glucagon-like peptide 2. Eur J Endocr 1998; 139:265. Bishop AE, Polak JM. Gastrointestinal endocrine tumours. Pathology. Baillières Clin Gastroenterol 1996; 10(4):555. Nobels FRE, Kwekkeboom DJ, Bouillon R, et al. Chromogranin A: its clinical value as marker of neuroendocrine tumors. Eur J Clin Invest 1998; 28:431. Sekiya K, Ghatei MA, Salahuddin MJ, et al. Production of GAWK (Chromogranin-B 420-493)-like immunoreactivity by endocrine tumors and its possible diagnostic value. J Clin Invest 1989; 83:1834.

33a. Ischia R, Gasser RW, Fischer-Collorie R, et al. Levels and molecular properties of secretoneurin-immunoreactivity in the serum and urine of control and neuroendocrine tumor patients. J Clin Endocrinol Metab 2000; 85:355. 34. 35. 36. 37. 38. 39.

Chiba T, Yamada T. Gut somatostatin. In: Walsh JH, Dockray GJ, eds. Gut Peptides. New York: Raven Press, 1994:123. Pollack MN, Schally AV. Mechanisms of antineoplastic action of somatostatin analogs. Proc Soc Exp Biol Med 1998; 217:143. Lamberts SWJ, Van der Lely A-J, De Herder WW, et al. Octreotide. N Engl J Med 1996; 334(4):246. Patel YC, Greenwood MT, Panetta R, et al. The somatostatin receptor family. Life Sci 1995; 57(13):1249. De Herder WW, Hofland LJ, Van der Lely AJ, Lamberts SW. Gastrointestinal endocrine tumours. Peptide receptors in gut endocrine tumors. Baillières Clin Gastroenterol 1996; 10(4):571. Virgolini I. Mark Forster Award Lecture. Receptor nuclear medicine: vasointestinal peptide and somatostatin receptor scintigraphy for diagnosis and treatment of tumour patients. Eur J Clin Invest 1997; 27 (10):793.

CHAPTER 221 CARCINOID TUMOR AND THE CARCINOID SYNDROME Principles and Practice of Endocrinology and Metabolism

CHAPTER 221 CARCINOID TUMOR AND THE CARCINOID SYNDROME PAUL N. MATON Cell of Origin Pathology Clinical Features Carcinoids without Systemic Features Carcinoid Syndrome Pathogenesis of the Carcinoid Syndrome Diagnosis of the Carcinoid Syndrome Prognosis Treatment Control of the Tumor Control of Symptoms Chapter References

Carcinoids are found in 1% of autopsies, but clinical data suggest an incidence of 2 cases per 100,000 per year.1 Although the incidence of subclinical gastric carcinoids has probably been underestimated,2,3 most carcinoids remain localized and are not clinically significant. The major interest in these tumors relates to the few that produce 5-hydroxytryptamine and other substances and cause flushing, diarrhea, heart disease, and asthma—the carcinoid syndrome.3,4

CELL OF ORIGIN Carcinoids are neuroendocrine tumors that usually arise from enterochromaffin (EC) cells, which are found scattered throughout the body but occur principally in the submucosa of the intestine and main bronchi.3,4 The EC cell population is heterogeneous, which may explain the variety of features associated with carcinoid tumors (Table 221-1). Some EC cells are argentaffinic, whereas others are argyrophilic (see Chap. 195). Furthermore, some EC cells contain peptides, such as substance P, enkephalins, or motilin.3 Most gastric carcinoids arise not from EC cells but from enterochromaffinlike (ECL) cells, which may be important in the production of histamine.2,5

TABLE 221-1. Characteristics and Embryologic Derivation of Carcinoid Tumors

PATHOLOGY Carcinoid tumors are benign or of low-grade malignancy. Some authors, however, have included more aggressive “atypical” carcinoids and even highly malignant neuroendocrine carcinomas in accounts of carcinoids.6 The relative distribution of carcinoid tumors, their propensity to metastasize, and their ability to cause the carcinoid syndrome are given in Table 221-2. One analysis of 8000 carcinoids1 showed appendiceal carcinoids to be the most common. Overall, 75% of carcinoids occur in the gut, 24% in the lungs, and only 1% in other sites. Eighty percent of small intestinal carcinoids (which may be multiple) occur within 60 cm (2 ft) of the ileocecal valve, and 85% of bronchial carcinoids are in the main bronchi. The primary tumors tend to remain small and to extend outward, away from the lumen. They then spread to local lymph nodes. A marked fibrotic reaction may occur, which, with midgut carcinoids, may distort the gut and mesentery, and sometimes cause intestinal obstruction or vascular occlusion. Further spread occurs to the peritoneum or liver, and distant metastases may occur at almost any site; such distant lesions may include osteolytic and osteoblastic bone metastases.3,7,8 Most gastric ECL cell carcinoids are under the influence of gastrin. In patients with hypergastrinemia due to gastric atrophy (with or without pernicious anemia) or Zollinger-Ellison syndrome with multiple endocrine neoplasia type 1 (MEN1), a generalized hypertrophy of ECL cells is present that in some cases leads to the formation of multiple carcinoids.2,5 The many small polyps (sometimes numbering in the hundreds) that are seen initially depend on gastrin but may become autonomous. Metastases to lymph nodes are rare6 but are more common in sporadic (non–hypergastrinemia-associated) ECL cell carcinoids (see Table 221-2).

TABLE 221-2. Ability of Carcinoid Tumors to Metastasize and Produce the Carcinoid Syndrome

Carcinoids may contain and secrete various peptides and amines, and some carcinoids, particularly those of the foregut, can produce other clinical syndromes with or without the carcinoid syndrome. Carcinoids that secrete insulin, growth hormone, corticotropin, b-melanocyte–stimulating hormone, gastrin, calcitonin, substance P, growth hormone–releasing hormone, and bombesin-like peptides have been described.3 Many carcinoids secrete chromogranin, a peptide common to many neuroendocrine tumors.9 Carcinoid tumors of the stomach can occur in MEN1. The MEN1 gene, a tumor suppressor gene, is on chromosome 11 (11q13). In MEN1 patients gastric carcinoids (like parathyroid and islet cell tumors) exhibit loss of heterozygosity at 11q13 with deletion of the wild-type allele.10 This developmental mechanism is also true for lung carcinoids (and possibly for sporadic gut carcinoids,10,11 but not for thymic carcinoids12), irrespective of MEN1.13,14 The carcinoid syndrome is much rarer than carcinoid tumors: the estimated incidence is approximately three cases per million population per year. The syndrome occurs only when vasoactive substances reach the systemic circulation; therefore, most gut carcinoids cause the syndrome only when hepatic metastases are present.

Even then, the carcinoid syndrome occurs only in a few cases (see Table 221-2), and the histamine-type syndrome caused by gastric carcinoids has been described in less than ten cases (all carcinoids were sporadic ECL cell carcinoids). Ovarian and bronchial tumors, which drain directly into the systemic circulation, can cause the carcinoid syndrome without metastases; however, with bronchial carcinoids, metastases are usually present. Rarely, medullary carcinoma of the thyroid and small cell tumors of the lung cause the syndrome.

CLINICAL FEATURES CARCINOIDS WITHOUT SYSTEMIC FEATURES Carcinoids without systemic features occur most commonly in the appendix and small intestine1 (see Table 221-2) and usually are found incidentally at surgery.3,4,7,8 Small intestinal carcinoids are usually symptomless but, occasionally, cause intestinal obstruction or vascular occlusion. Ileal tumors generally are not demonstrated by simple radiology; therefore, barium infusion studies or angiography is required.15 Duodenal and gastric carcinoids most often are found incidentally at endoscopy. Colonic, rectal, and esophageal carcinoids also may be found incidentally or may cause obstruction.16 Bronchial carcinoids may be discovered as a coin lesion on chest radiographs or may be seen at bronchoscopy. They also may present with cough, wheeze, hemoptysis, or segmental obstruction and infection. Mediastinal and ovarian carcinoids appear as masses.17 Most carcinoids occur as an isolated disease, but associations are seen between foregut carcinoids and MEN1; between gastric carcinoids and hypergastrinemia, whether due to achlorhydria or to Zollinger-Ellison syndrome, especially as part of MEN118; between ampullary carcinoids and von Reck-linghausen disease3; and between renal carcinoids and horseshoe kidney.19 The diagnosis of all carcinoids without systemic features depends on the histologic structure and staining. CARCINOID SYNDROME The carcinoid syndrome3,4,7,8 is characterized by flushing, diarrhea, and heart disease, although the relative importance of the symptoms varies in different patients, reflecting differences in tumor origin, bulk, tumor products, and length of history. The primary tumor may have been removed many years before the development of the syndrome, or it may never have become clinically evident. Most patients with the carcinoid syndrome, however, have an ileal tumor, with evident hepatic metastases at the time of presentation (see Table 221-2). Carcinoid flushing is erythematous, and principally affects the upper part of the body. Some patients are unaware of the flushing, whereas others are distressed by it. Flushes may be brief or prolonged. They often are spontaneous, but they may be precipitated by alcohol, certain foods, abdominal palpation, or anxiety. Several patterns of flushing have been described,7 but only two are clinically distinctive. Gastric carcinoids may produce a bright red, geographic flush, often precipitated by food consumption.20 Bronchial carcinoids may cause severe prolonged flushes with salivation, lacrimation, sweating, facial edema, palpitations, diarrhea, and hypotension. After many months of flushing, a fixed, facial telangiectasia, edema, and cyanotic plethora may occur. Diarrhea occurs in most cases, although it is less evident with gastric carcinoids. In some patients, diarrhea is related to episodes of flushing; in others, the two seem independent. Watery diarrhea is more frequent than malabsorption and is due to increased motility and possibly intestinal secretion, but intermittent intestinal obstruction, cholorrheic diarrhea after previous intestinal resection for tumor, or vascular insufficiency or lymphatic obstruction may occur in some patients. Abdominal pain may be due to these abnormalities or to necrosis of hepatic metastases. Heart disease occurs in ~30% of patients. Insidious right heart failure often worsens during periods of flushing, and, in such cases, tricuspid regurgitation or stenosis is typical; less commonly, pulmonary stenosis may occur. The left side of the heart also may be involved, usually in association with bronchial carcinoids. The heart disease is caused by a unique form of fibrosis that involves the endocardium and valves.21,21a Fibrosis in other sites can cause constrictive pericarditis, retroperitoneal fibrosis, pleural thickening,22 and Peyronie disease. Wheezing occurs in 10% of patients and can be the presenting feature. A pellagra-like syndrome may occur, and confusional states have been described with foregut carcinoids. Rarely, patients experience arthralgia or a myopathy.23

PATHOGENESIS OF THE CARCINOID SYNDROME The flushing is poorly understood but may be due to the release of kinins in some patients.24,25 Carcinoids contain kallikrein, an enzyme that, when released into the circulation spontaneously or stimulated by alcohol or catecholamines, acts on plasma kininogens to generate bradykinin3,7 (see Chap. 170). In gastric carcinoids, the distinctive flush is mediated by histamine (see Chap. 181). The diarrhea is caused largely by 5-hydroxytryptamine (serotonin) through its effects on gut motility. The 5-hydroxytryptamine also contributes to the asthma and is probably implicated in the cardiac fibrosis. The diversion of tryptophan to the tumor for 5-hydroxytryptamine synthesis can lead to reduced protein synthesis, with hypoalbuminemia, and to nicotinic acid deficiency, with pellagra.26 The prostaglandins27 and many gut peptides28 probably are not mediators of the flushing or diarrhea in most patients. The role of other peptides such as substance P or other tachykinins (neurokinin A, neuropeptide K) has yet to be fully evaluated.25

DIAGNOSIS OF THE CARCINOID SYNDROME Once the carcinoid syndrome has been considered, confirmation of the diagnosis is often not difficult and rests on clinical features, measurement of the principal 5-hydroxytryptamine metabolite in urine (5-hydroxyindoleacetic acid [5-HIAA]) and, occasionally, the provocation of flushing with epinephrine.3,4,7,8 In a patient who flushes, who has an enlarged liver, and in whom urinary 5-HIAA excretion is >30 mg per day (normal is 75% for tumors from all sites. In patients with distant metastases, the 5-year survival was 30% or less.32 One analysis indicates an overall 5-year survival of 50%, with 45% of carcinoids having metastasized at the time of diagnosis.1 In the largest reported series of patients with the carcinoid syndrome, the median survival from first flush was 3 years but survival ranged up to 17 years.8 The median survival in patients with heart disease was 14 months, and in patients with a large tumor burden (5-HIAA level of >150 mg per day), it was 11 months.

TREATMENT CONTROL OF THE TUMOR Except in the case of gastric ECL cell carcinoids that are of low malignancy and can be observed for some years, surgery should be considered in all patients because the resection of local disease can result in cure of carcinoid tumors, and in cure of the carcinoid syndrome due to some bronchial and ovarian tumors.33 Interestingly, the appropriate surgery for multiple gastric carcinoids in patients with gastric atrophy is not removal of the tumors, but removal of the gastric antrum, leading to normalization of plasma gastrin and tumor regression.5 Resection of isolated hepatic metastases detected by computed tomographic scan or somatostatin analog scintigraphy34,35 and 36 also may be markedly beneficial in selected cases7,8; however, in the presence of extensive metastases, partial hepatic resection is not warranted, nor is removal of the primary tumor unless it is causing local problems. Hepatic transplantation has been beneficial on occasion.37 Chemotherapy for carcinoid tumors has been disappointing, with responses occurring in a minority and lasting only a mean of 7 months.8 Single agents have produced responses in up to 30%, streptozocin being the most effective agent. Various combinations of streptozocin with 5-fluorouracil, cyclophosphamide, and doxorubicin have produced response rates of up to 35%.8 Given the variation in tumor growth, questionable efficacy, and undoubted toxicity of chemotherapy, as well as the availability of other symptomatic

therapy, chemotherapy should be reserved for advanced tumors that are actively growing. Administration of interferon-a;, 3 to 6 million IU per day subcutaneously, reduces tumor size in ~15% of cases and stabilizes tumor size in another 30% to 40%.8,38,39,40 and 41 Octreotide acetate, used mainly for its effect on symptoms (see later), reduces tumor bulk in ~5% and stabilizes tumor size in another 20%.42,43 and 44 Radiotherapy is useful only for symptomatic therapy of bone and skin metastases. Hepatic artery occlusion leads to selective necrosis of hepatic metastases. Surgical ligation of the hepatic artery has been used to necrose the hepatic tumor in the carcinoid syndrome,8 but percutaneous arterial embolization is less traumatic, more selective, and can be repeated.45,46 and 47 Embolization, alone or in combination with chemotherapy,8,48 can produce a striking relief of symptoms and reduction in urinary 5-HIAA levels, even in patients with symptoms that are resistant to other modes of therapy. Complete remissions of up to 30 months have been reported. Second remissions may follow repeat embolizations, and survival may be prolonged. CONTROL OF SYMPTOMS Many patients have considerable hepatic tumor, yet they remain well, apart from occasional flushing or diarrhea.3,7 They should be advised to avoid precipitants of flushing and to supplement their diet with nicotinamide. Heart failure can be treated with diuretics, asthma with albuterol (salbutamol) (which does not precipitate flushing), and diarrhea with loperamide. If patients require further therapy, various other agents may be tried. For diarrhea, cyproheptadine, 4 to 8 mg every 6 hours, is the best oral agent.49 In the rare patient with carcinoid syndrome due to a gastric carcinoid, a combination of diphenhydramine 50 mg every 6 hours, together with an H2 antagonist (e.g., cimetidine 300 mg every 6 hours), has proved effective for flushes.50 For most patients, however, the most effective agent for both diarrhea and flushes is the long-acting somatostatin analog octreotide acetate (100–500 µg every 8–12 hours or the long-acting formulation every 24 hours subcutaneously), which has produced responses in >80% of patients.43,44,51 Another somatostatin analog, lanreotide, is also effective. Interferon-a, used principally for its effect on the tumor, reduces flushing and diarrhea in ~50% of patients8,38,39 and 40 (see Chap. 169). If drugs fail to control symptoms, hepatic embolization should be considered. Progressive cardiac disease can be halted only by removal of the tumor and cure of the carcinoid syndrome, but the occasional, carefully selected patient may benefit from tricuspid valve replacement.7,52 Anesthetics, surgery, chemotherapy, and hepatic artery occlusion can precipitate extremely severe flushing with hypotension—a carcinoid crisis. The risk of developing such a crisis can be reduced by appropriate premedication, careful monitoring, the judicious use of anesthetic drugs and techniques, and avoidance of flush-provoking agents, such as catecholamines.45,46,53 Should a crisis occur, hypotension should be treated with octreotide acetate 100 µg intravenously, which should be available whenever patients with the carcinoid syndrome undergo procedures.54 If octreotide acetate is not available, intravenous methoxamine, 3 to 5 mg, can be used. Other pressor agents should be avoided. CHAPTER REFERENCES 1. Modlin IM, Sandor A. An analysis of 8305 cases of carcinoid tumors. Cancer 1997; 79:813. 2. Rindi G, Luinetto O, Cornaggia M, et al. Three subtypes of gastric argyrophil carcinoid and the gastric neuroendocrine carcinoma: a clinical pathologic study. Gastroenterology 1993; 104:994. 3. Maton PN, Hodgson HJF. Carcinoid tumours and the carcinoid syndrome. In: Bouchier IAD, Allan RN, Hodgson HJF, Keighly MRB, eds. Textbook of gastroenterology. London: Bailliere-Tindall, 1984:620. 4. Feldman JM, Zakin D, Dannenberg AJ. Carcinoid tumors and syndrome. Semin Oncol 1987; 14:237. 5. Maton PN, Dayal Y. Clinical implications of hypergastrinemia. In: Zakim DH, Dannenberg AJ, eds. Peptic ulcer disease and other acid-related disorders. New York: Academic Research Associates, 1991:213. 6. Bordi C, Falchetti A, Azzoni C, et al. Aggressive forms of gastric neuroendocrine tumors in multiple endocrine neoplasia type 1. Am J Surg Pathol 1997; 21:1075. 7. Grahame-Smith DG. The carcinoid syndrome. London: William Heinemann, 1972. 8. Moertel CG. An odyssey in the land of small tumors. J Clin Oncol 1987; 5:1503. 8a.

Nuttall KL, Pingree SS. The incidence of elevations in urine 5-hydroxyin-doleacetic acid. Ann Clin Lab Sci 1998; 28:167.

9. Nobels FR, Kwekkeboom DJ, Coopmans W. Chromogranin A as a serum marker for neuroendocrine neoplasia: comparison with neuron–specific enolase and the alphasubunit of glycoprotein hormones. J Clin Endocrinol Metab 1997; 82:2622. 10. Debelenko LV, Emmert-Buck MR, Zhuang Z, et al. The multiple endocrine neoplasia type 1 gene locus is involved in the pathogenesis of type II gastric carcinoids. Gastroenterology 1997; 113:773. 11. Jakobovitz O, Nass D, De Marco L, et al. Carcinoid tumors frequently display genetic abnormalities involving chromosome 11. J Clin Endocrinol Metab 1996; 81:164. 12. Teh BT. Thymic carcinoids in multiple endocrine neoplasia type 1. J Intern Med 1998; 243:501. 13. Dong Q, Debelenko LV, Chandrasekharappa SC, et al. Loss of heterozygosity at 11q13: analysis of pituitary tumors, lung carcinoids, lipomas, and other uncommon tumors in subjects with familial multiple endocrine neoplasia type 1. J Clin Endocrinol Metab 1997; 82:1416. 14. Walch AK, Zitzelsberger HF, Aubele MM, et al. Typical and atypical carcinoid tumors of the lung are characterized by 11q deletions as detected by comparative genomic hybridization. Am J Pathol 1998; 153:1089. 15. Jeffree MA, Barter SJ, Hemingway AP, Nolan DJ. Primary carcinoid tumors of the ileum: the radiological appearances. Clin Radiol 1984; 35:451. 16. Spread C, Berkel H, Jewell L, et al. Colon carcinoid tumors: a population-based study. Dis Colon Rectum 1994;37:482. 17. Wang DY, Chang D-B, Kuo S-H, et al. Carcinoid tumors of the thymus. Thorax 1994; 49:357. 18. Hakanson R, Sundler F, eds. Mechanisms for the development of gastric carcinoids: proceedings of an international symposium. Digestion 1986; 35(Suppl 1):1. 19. Krishnan B, Truong LD, Saleh G, et al. Horseshoe kidney is associated with an increased relative risk of primary renal carcinoid tumor. J Urol 1997; 157:2059. 20. Roberts LJ, Marney SR, Oates JA. Blockade of the flush associated with metastatic gastric carcinoid by combined H 1 and H2 receptor antagonists: evidence for an important role of H 2 receptors in human vasculature. N Engl J Med 1979; 300:236. 21. Wikowske MA, Hartman LC, Mullaney CJ, et al. Progressive carcinoid heart disease after resection of primary ovarian carcinoid. Cancer 1994; 73:1889. 21a. Sakai D, Mukakami M, Kasazoe K, Tsutsumi Y. Ileal carcinoid tumor complicating carcinoid heart disease and secondary retroperitoneal fibrosis. Pathol Int 2000; 50:404. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54.

Moss SF, Lehner PJ, Gilbey SG, et al. Pleural involvement in the carcinoid syndrome. Q J Med 1993; 86:49. Lederman RJ, Bukowski RM, Nickelson P. Carcinoid myopathy. Cleve Clin J Med 1987; 54:299. Lucas KJ, Feldman JM. Flushing in the carcinoid syndrome and plasma kallikrein. Cancer 1986; 58:2290. Grahame-Smith DG. What is the cause of the carcinoid flush? Gut 1987; 28:1413. Swain CP, Tavill AS, Neale G. Studies of tryptophan and albumin metabolism in a patient with carcinoid syndrome, pellagra and hypoproteinemia. Gastroenterology 1976; 74:484. Metz SA, McRae JR, Robertson PR. Prostaglandins as mediators of paraneoplastic syndromes: review and update. Metabolism 1981; 30:299. Long RG, Peters JR, Bloom SR, et al. Somatostatin, gastrointestinal peptides and the carcinoid syndrome. Gut 1981; 22:549. Wilkin JK. Flushing reactions: consequences and mechanisms. Ann Intern Med 1981; 95:468. Aldrich LB, Moattari R, Vinik AI. Distinguishing features of idiopathic flushing and carcinoid syndrome. Arch Intern Med 1988; 148:2614. Young DS, Pestaner LC, Gibberman V. Effects of drugs on clinical laboratory tests. Clin Chem 1975; 21:398D. Godwin JD. Carcinoid tumors: an analysis of 2837 cases. Cancer 1975; 36:560. Norton JA. Neuroendocrine tumors of the pancreas and duodenum. Curr Probl Surg 1994; 31:77. Kwekkeboom DJ, Krenning EP, Bakker WH, et al. Somatostatin analog scintigraphy in carcinoid tumors. Eur J Nucl Med 1993; 20:283. Kwekkeboom DJ, Krenning EP. Somatostatin receptor scintigraphy in patients with carcinoid tumors. World J Surg 1996; 20:157. Kisker O, Weinel RJ, Geks J, et al. Value of somatostatin receptor scintigraphy for preoperative localization of carcinoids. World J Surg 1996; 20:162. Le-Treut YP, Delpero JR, Dousset B, et al. Results of liver transplantation in the treatment of metastatic neuroendocrine tumors. A 31–case French multicentric report. Ann Surg 1997; 225:355. Oberg K, Eriksson B. Role of interferons in the management of carcinoid tumors. Br J Hematol 1991; 79(Suppl 1):74. Janson ET, Oberg K. Long-term management of the carcinoid syndrome: treatment with octreotide alone and in combination with alpha interferon. Acta Oncol 1993; 32:225. öberg K, Norheim I, Lind E, et al. Treatment of malignant carcinoid tumors with human leukocyte interferon. Cancer Treat Rep 1986; 70:1296. Oberg K. Advances in chemotherapy and biotherapy of endocrine tumors. Curr Opin Oncol 1998; 10:58. Arnold R, Benning R, Neuhaus C, et al. Gastroenteropancreatic endocrine tumors: effect of Sandostatin on tumor growth. The German Sandostatin Study Group. Metabolism 1992; 41(Suppl 2):116. Kvols LK, Moertel CG, O'Connell MJ, et al. Treatment of the malignant carcinoid syndrome. N Engl J Med 1986; 315:663. Gorden P, Comi RJ, Maton PN, Go VLW. Somatostatin and somatostatin analogue (SMS 201–995) in treatment of hormone-secreting tumors of the pituitary and gastrointestinal tract and non-neoplastic diseases of the gut. Ann Intern Med 1989; 110:35. Maton PN, Camilleri M, Griffin G, et al. The role of hepatic arterial embolisation in the carcinoid syndrome. BMJ 1983; 287:932. Martensson H, Norbin A, Bengmark S, et al. Embolisation of the liver in the management of metastatic carcinoid tumors. J Surg Oncol 1984; 27:152. Ruszdiewski P, Malka D. Hepatic arterial chemoembolization in the management of advanced digestive endocrine tumors. Digestion 2000; 62(Suppl 1):79. Drougas JG, Anthony LB, Blain TK, et al. Hepatic artery chemoembolization for management of patients with advanced metastatic carcinoid tumors. Am J Surg 1998; 175:408. Moertel CG, Kvols LK, Rubin J. A study of cyproheptadine in the treatment of metastatic carcinoid tumor and the malignant carcinoid syndrome. Cancer 1991; 67:33. Oates JA. The carcinoid syndrome. N Engl J Med 1986; 315:702. Saslow SB, O'Brien MD, Camilleri M, et al. Octreotide inhibition of flushing and colonic motor dysfunction in carcinoid syndrome. Am J Gastroenterol 1997; 92:2250. Codd JE, Prozda J, Merjavy J. Palliation of carcinoid heart disease. Arch Surg 1987; 122:1076. Törnebrandt K, Nobin A, Ericsson M, Thompson D. Circulation, respiration and serotonin levels in carcinoid patients during neuroleptic anaesthesia. Anaesthesia 1983; 38:957. Marsh HM, Martin JK Jr, Kvols LK, Moertel CG. Carcinoid crisis during anesthesia: successful treatment with somatostatin analogue. Anesthesiology 1987; 66:89.

CHAPTER 222 HORMONES AND CARCINOGENESIS: LABORATORY STUDIES Principles and Practice of Endocrinology and Metabolism

CHAPTER 222 HORMONES AND CARCINOGENESIS: LABORATORY STUDIES JONATHAN J. LI AND SARA ANTONIA LI General Considerations Experimental Animal Models Mammary Gland Ovary Uterus Cervix-Vagina Pituitary Gland Testes Kidney Liver Prostate Ductus Deferens and Scent Gland Perinatal Effects Hormones as Cocarcinogens or Promoters In Vitro Cell Culture Models Syrian Hamster Embryo Cell System Balb/C 3T3 Cell System Other Cell Systems Metabolism and Covalent Binding Studies Growth Factor and Oncogene Involvement Chapter References

The resurgence and rapid growth of the field of hormonal carcinogenesis—the role of hormones in the etiology and growth of cancer—are due in large part to growing concerns regarding two of the most common human cancers, breast and prostate.1,2,3 and 4 Although other hormone-associated cancers occur at lower frequencies, they are also of clinical importance; these include endometrial, ovarian, testicular, cervicovaginal, pituitary, thyroid, and sex hormone–related hepatic neoplasms.5,6 and 7 That these cancers cannot be attributable to any specific exogenous physical, environmental, or dietary factor is becoming increasingly clear. Despite the long history of hormonal carcinogenesis research, the precise mechanism whereby hormones affect neoplastic transformation remains elusive. A better understanding of the effect of hormones, both ovarian/testicular and pituitary, on normal cellular processes of growth and differentiation is needed to ascertain more precisely their involvement in neoplastic development. Nevertheless, after intensive study,8 some of the cellular and molecular alterations elicited by hormones during tumorigenesis are beginning to be revealed. That hormones can induce neoplasms in experimental animals has been known for >60 years.9 Moreover, with a few notable exceptions, for nearly every human neoplasm with a hormonal association, a corresponding animal tumor model can be induced by hormones alone. Of the various hormonal agents, sex hormones, particularly estrogens9a and to a lesser extent progesterone and prolactin (PRL) in women, and androgens in men, have been associated with tumor induction.

GENERAL CONSIDERATIONS Hormones can affect neoplastic processes by acting either as the sole etiologic agent or in conjunction with physical agents (i.e., ionizing radiation) or nonhormonal chemical carcinogens.8 A number of general mechanisms exist whereby hormones may modify a target tissue during one or more phases of the events initiated by carcinogenic events, such as viral infection, chemical exposure, or exposure to ionizing radiation. For example, hormones may be involved in (a) promotional or carcinogenic effects, (b) alterations of the host immune system, (c) activation of viruses, and (d) modification of hormone receptors or alteration of metabolic rates affecting carcinogen activation. The primary concern of this chapter, however, is the induction of tumors, benign and malignant, by hormones, either endogenously produced or exogenously administered. The concept that a given hormone acts specifically on individual target organs or tissues is somewhat misleading, because most organs or tissues are differentially sensitive to various hormones acting alone, or in concert with or opposition to, other hormones. The general characteristics of hormonal carcinogenesis are (a) tissue, strain, and species specificity; (b) long induction period; (c) sustained and prolonged hormone exposure; and (d) cellular proliferation. Sex hormones have been implicated in the induction and growth of a wide variety of experimental tumors as summarized in Table 222-1. A common characteristic during endocrine-induced tumorigenesis appears to be a prolonged and severe derangement in normal homeostasis and regulatory relationships as a result of chronic hormone exposure.

TABLE 222-1. Animal Models in Hormonal Carcinogenesis

Although the minimum oncogenic dose for a given sex hormone to elicit a high incidence of tumors at any tissue site is not precisely known, clearly the conditions required to induce a high tumor yield do not require particularly high concentrations of hormones, either at the serum or tissue level (Table 222-2). For example, the serum level of 17b-estradiol (17b-E2) in a female ACI rat in estrus is 75 to 80 pg/mL,10 and the sustained oncogenic dose required to elicit a high incidence of mammary tumors is only twice estrous levels. Such levels approach those found during pregnancy in this species. In the male Syrian hamster, continuous administration of exogenous 17b-E2 induces renal tumors, and the serum estrogen levels are approximately seven-fold higher than the mean estrous levels found in normal untreated females (see Table 222-2). The 17b-E2 levels in the male kidney are equal to or slightly lower than those found in the uterus during estrus, however, because this organ site has only modest ability to concentrate estrogens.11 Similarly, even lower levels of estrogen and androgen are required to induce a high incidence of prostate carcinomas in the male Noble rat.12

TABLE 222-2. Sex Hormone Levels Required for Experimental Hormonal Carcinogenesis

EXPERIMENTAL ANIMAL MODELS Numerous murine and one primate species develop tumors in response to sex hormone exposure alone. Although some of these hormone-induced tumor models either may involve viral mediation or may arise in conjunction with the stimulation of other pertinent hormonal factors, most are considered to be the result of the direct carcinogenic action of the hormonal agents themselves. The induction of tumors by hormones characteristically occurs in hormone-dependent target tissues. The Syrian hamster kidney model is included in this group because it is essentially an estrogen-responsive target organ.7,11 MAMMARY GLAND More than 65 years ago, Lacassagne9 first demonstrated that long-term estrone (E 1) administration induced a high frequency of mammary cancer in male mice. Subsequent studies showed that numerous other mouse strains were also susceptible to the carcinogenic action of estrogens at this tissue site.13 Whereas, in the past, the suggestion was that estrogens were direct carcinogens, the belief now is that, in mice, sex hormones alone cannot effect a high incidence of mammary tumors in the absence of a mouse mammary tumor virus (MMTV) or chemical carcinogen exposure. The lack of an established viral association in rats, however, suggests that mammary tumor induction by female sex hormones in susceptible strains may result from direct hormonal action. High incidences (54–100%) of mammary tumors have been elicited with natural and synthetic estrogens, including E1, 17b-E2, ethynylestradiol (EE), and diethylstilbestrol (DES), in both male and female Noble, Wistar, Long-Evans, and ACI rats after 5.0 to 10.0 months of continuous estrogen treatment (Fig. 222-1).6,10,14 As in the mouse, genetic factors also clearly play a significant role in the induction of mammary tumors by estrogens in the rat. Despite the marked differences in incidence between the sexes in humans, the lack of a sex difference in the ability of estrogens to induce mammary tumors in the rat may actually be analogous to the situation seen in humans. For instance, the strongest risk factor for breast cancer in men is known to be Klinefelter syndrome, a condition resulting from inheritance of an extra X chromosome and characterized by testicular dysfunction and gynecomastia.15 This finding clearly indicates that breast cancer in men develops under conditions favoring excessive endogenous estrogen levels. Pertinent distinctions can be made between mammary tumors in the rat induced by estrogen and by a chemical carcinogen (e.g., dimethylbenzanthracene [DMBA], N-nitrosomethylurea [NMU]). Estrogen-induced primary mammary neoplasms exhibit a modest but distinct frequency of metastases (~15%) to other tissue sites, including lymph nodes, liver, and lung,16,17 whereas rat mammary tumors induced by chemical carcinogens do not exhibit any significant metastatic potential.18 Approximately 85% of the mammary tumors originating in rats are estrogen dependent, similar to breast cancer in postmenopausal women, whereas most (~90%) of the mouse mammary tumors are estrogen independent.18 Also, PRL plays a permissive, if not essential, role in the induction of mammary tumors by estrogen in most rat strains.14 Finally, combined estrogen and progesterone treatment has been shown to induce a higher incidence of mammary tumors in Wistar-WAG rats than estrogen exposure alone.19 Of interest, however, medroxyprogesterone acetate (MPA, Provera) is capable of inducing mammary tumors in mice in the absence of added estrogen.20

FIGURE 222-1. Mammary gland carcinomas (arrowheads) induced after continuous administration of 17b-estradiol for 6 months to a female ACI rat. Hormone pellets (20-mg pellet containing 4 mg of 17b-estradiol) were renewed every 4 months to maintain constant estrogen levels. Serum estradiol concentrations were 165 to 170 pg/mL throughout the treatment period.

The biosynthesis of the estrogens E1 and 17b-E2 from their androgen precursors, androstenedione and testosterone, is catalyzed by aromatase, a microsomal cytochrome P450-dependent enzyme. In postmenopausal women with breast cancer, aromatase activity in the peripheral tissues is a major endogenous source of estrogen for tumor growth.21,22 The dynamics of androgen and estrogen production and metabolism, particularly the percentage of peripheral aromatization in nonhuman primates (cynomolgus, rhesus, and baboons), closely resembles that in humans.23 This is not found in murine species. Primate species have been used as models of human peripheral aromatization to test the therapeutic effects of steroidal and nonsteroidal aromatase inhibitors in vivo.24 OVARY Presently, an animal model for hormonally induced epithelial ovarian tumors does not exist. Nevertheless, granulosa cell tumors of the ovary develop in 25% to 50% of BALB/c mice when they are implanted with progesterone pellets.25 In one study, 19-norprogesterone was more effective than progesterone in inducing these neoplasms. The contraceptive agents norethindrone and norethynodrel elicited a 52% incidence of ovarian tumor. Castrated rats with intrasplenic ovarian transplants that resulted in constant high levels of gonadotropins showed a high frequency of tumors in the transplanted ovaries.26 The resulting ovarian tumors were thecal granulosa cell tumors, however, and were not of epithelial origin. In these rats, the long-term excess of gonadotropins is believed to be largely responsible for promoting ovarian tumor development. UTERUS Despite the well-established association between estrogen and endometrial cancer in women,4 an animal model of similar hormonally induced cancer at this site is lacking. Although endometrial tumors are produced with a high incidence in rabbits after estrogen treatment, these adenocarcinomas are preceded by cystic hyperplasia, which occurs spontaneously with high frequency (75%) in aging animals.27 Endometrial carcinomas in the uterine horns have also been induced with either DES or 17b-E2 in mice.28 Uterine carcinomas were observed in 90% of mice receiving DES for 5 days neonatally.29 The induction of these uterine tumors was age and dose dependent. In addition, these uterine tumors were estrogen dependent because they partially regressed after ovariectomy and, when transplanted into nude mice, required estrogen for continued growth. The involvement of a MMTV in mice uterine tumor development remains a possibility. Similarly, estrogen-dependent uterine tumors can be induced in hamsters when DES is administered to newborn animals.30 Finally, a high incidence of uterine leiomyosarcomas has been induced in hamsters after combined estrogen and androgen treatment.31 Interestingly, the addition of progesterone to this combined treatment inhibited the induction of these uterine smooth muscle cell tumors. Particularly relevant is the induction of uterine mesotheliomas in a non-human primate species (squirrel monkey) after prolonged treatment with either DES or estradiol benzoate.32 CERVIX-VAGINA Cervical or vaginal squamous cell carcinomas occur after prolonged estrogen administration in C3H, C57, and BC mouse strains.33 Moreover, no spontaneous occurrence of such tumors has been reported. Generally, the belief is that these tumorigenic effects are produced by the direct carcinogenic action of estrogens. Also, prolonged testosterone administration causes cervical tumors in female hybrid mice. PITUITARY GLAND Some strains of mice and rats are highly susceptible to the induction of pituitary tumors by estrogens, whereas other strains are largely resistant.34,35 Males appear to be more susceptible to estrogen-induced pituitary tumors than females. Once pituitary tumors develop in mice, they do not regress after estrogen treatment ceases.

Histologically, these tumors are described as chromophobe adenomas. The predominant secretion of these tumors is PRL, and growth-promoting properties as well as adrenocorticotropin-like effects have been reported. These pituitary tumors can be induced either by natural steroidal estrogens or by synthetic steroidal and stilbene estrogens. Intermediate-lobe pituitary adenomas also have been induced in rats and hamsters after prolonged estrogen treatment.36 Present evidence indicates that these pituitary tumors are also induced by direct hormonal action. TESTES The induction of testicular tumors in mice with estrogens has been studied extensively. Initially, malignant tumors of the interstitial cells were reported to develop in the A1 strain of mice receiving 17b-E237 Since then, similar testicular tumors have been induced with high incidence in other mouse strains, including BALB/c, ABi, and ACrg, but not in several other strains.38 As a consequence of estrogen treatment of susceptible mice, alterations in androgen biosynthetic enzyme systems, transient induction of DNA synthesis, and a greater nuclear estrogen content in Leydig cells may contribute to their neoplastic transformation.38 Apparently, the pituitary plays a permissive role in estrogen-induced testicular tumors, because hypophysectomy prevents the appearance of these tumors. KIDNEY The most extensively studied experimental model in hormonal carcinogenesis is the estrogen-induced renal carcinoma of the Syrian hamster. Long-term exposure of either castrated or intact male hamsters (but not female hamsters) to either steroidal or stilbene estrogens results in essentially 100% incidence of multiple bilateral renal neoplasms.39 Because the reproductive and urogenital tracts of the Syrian hamster arise from the same embryonic germinal ridge, the kidney of this species appears to have carried over genes that are expressed and responsive to estrogens. Complete chemoprevention of renal tumorigenesis can be effected by administering the estrogen concomitantly with androgen, progesterone, antiestrogens, or EE.40,41 Evidence strongly indicates that the estrogen-induced renal tumor arises primarily from undifferentiated committed epithelial stem cells in the interstitium.42 Not all estrogens are equally active in inducing these renal tumors.43 With the exception of EE, which elicits only a 10% renal tumor incidence, potent estrogens (17b-E2, DES, hexestrol, and 11b-methoxyethinylestradiol [Moxestrol]) exhibit high incidences of renal neoplasms compared with weak estrogens (estriol, 4-hydroxyestrone). Moreover, estrogens that possess low or negligible estrogenic activity (17a-E2, b-dienestrol, 2-hydroxy-estradiol) do not induce kidney tumors. The lack of strong carcinogenic activity of EE in the hamster kidney, despite its known potent estrogenic activity, may be the result of a differential effect of EE on the proliferation of a subset of renal tubule cells, rather than on the stem cells residing in the in terstitium.41 One of the most unusual features of the Syrian hamster kidney is its ability to behave as an estrogen-responsive and estrogen-dependent organ. Estrogen treatment elevates the level of estrogen receptors and induces progesterone receptors in the kidney. These effects are characteristic of estrogen action in target tissues. A comprehensive model for estrogen carcinogenicity in the hamster kidney is proposed (Fig. 222-2). Briefly, estrogens induce proliferation of preexisting estrogen-sensitive interstitial cells, as well as reparative proliferation secondary to cellular damage. The proliferation of the interstitial cells leads to aneuploidy and chromosomal instability, resulting in gene overexpression, amplification, and suppression (specifically, protooncogene, and suppressor gene expression) and eventually leads to tumor formation via a multistep process.44

FIGURE 222-2. Multistep model for estrogen-induced carcinogenesis in the Syrian hamster kidney. (E, estrogen; ER, estrogen receptor; E2F1, transcription factor E2F1; WT1, suppressor gene Wilm tumor 1.)

LIVER A few hepatic tumors have been induced in mice, rats, and hamsters by various synthetic estrogens and progestins.45 A 20% to 30% incidence of liver tumor has been reported in hamsters after long-term administration of EE. However, in the presence of 0.3% a-naphthoflavone (ANF) in the diet, or in a 20-mg pellet form, EE administration induced an 80% to 100% incidence of hepatocellular carcinomas in castrated male hamsters.46 Because ANF is not known to behave as a carcinogen or to possess substantial mutagenic activity, the belief is that it modifies the metabolism of synthetic estrogens, thus enhancing their carcinogenicity by increasing the amount of the parent hormone. A cocarcinogenic role for ANF cannot be ruled out, however, in the induction of these hamster liver tumors.46 PROSTATE Long-term exposure of either Noble or Lobund Wistar rats to testosterone results in prostatic carcinomas.47 Tumor incidence was 50% when testosterone treatment was applied for 13 months and then E1 was substituted for 6 months. Maximum tumor yields were obtained when testosterone plus 17b-E2 was given for 19 months, with the tumor incidence approaching 90%.48 The resultant tumor nodules attained only microscopic proportions, however, somewhat limiting the usefulness of this model. Similar simultaneous exposure to testosterone plus 17b-E2 for 4 months resulted in consistent dysplastic lesions in the dorsolateral lobe of the prostate in Noble rats.49 When testosterone was replaced by dihydrotestosterone, the active androgen in many species, prostatic tumors were not seen.50 These data suggest that 17b-E2 may be involved in the etiology of these prostatic neoplasms, because testosterone is known to enhance proliferative activity at this organ site. Current evidence suggests that estrogen, acting on the androgen-supported prostate, induces cell proliferation through a receptor-mediated process.51 DUCTUS DEFERENS AND SCENT GLAND As with tumors in the Noble rat prostate model, other hormone-induced tumors also require the presence of two hormones, both estrogen and androgen. Examples are a leiomyosarcoma induced in the hamster ductus deferens31 and an unknown type of epithelial tumor induced in the scent gland after long-term coadministration of these gonadal hormones.52 Although, presently, the relationship between these hormones in inducing these tumors is not well understood, the scent gland tumor is a particularly interesting model system because androgen is required for preneoplasia, and estrogen is required for full tumor development.52 PERINATAL EFFECTS Perinatal effects of estrogens have been studied extensively in the mouse.53 When these animals receive prenatal and neonatal exposure to DES or 17b-E2, cervicovaginal adenosis and adenocarcinomas occur in females, and testicular lesions occur in the rete testis of the males. The mechanism for these transplacental and perinatal effects remains to be elucidated.

HORMONES AS COCARCINOGENS OR PROMOTERS Finally, hormones can act as cocarcinogens or promoters in conjunction with either physical carcinogens (e.g., ionizing radiation) or chemical carcinogens (DMBA, diethylnitrosamine [DEN], N-nitrosobutylurea [NBU]) at different organ sites. For example, either DES or EE plus x-ray treatments yields a high incidence of mammary tumors in female ACI rats, a rat strain that is relatively insensitive to radiation treatment alone.54 These same hormones are capable of promoting mammary tumors in other rat strains exposed to DMBA and DEN/NBU.55

IN VITRO CELL CULTURE MODELS Hormonal effects on in vitro cell transformation and mutagenic assays have important implications regarding the role of these substances in oncogenic processes. Hormonal agents have yielded some negative results in numerous in vitro tests, including lack of gene mutations in the Salmonella typhimurium assay. Positive findings in other in vitro cell assay systems are significant, however, and strongly suggest the possibility that hormones may possess epigenotoxic characteristics that could affect in vivo malignant cell transformation.56 SYRIAN HAMSTER EMBRYO CELL SYSTEM One of the most intensively studied in vitro assays is the Syrian hamster embryo (SHE) cells in culture.57 In this assay system, DES and some of its metabolites induce morphologic and neoplastic transformation of SHE cells. However, no detectable gene mutations at two genetic loci were found. In the presence of a rat postmitochondrial supernatant fraction, DES also induced unscheduled DNA synthesis. BALB/C 3T3 CELL SYSTEM Another in vitro cell system that has been studied in considerable detail is the BALB/c 3T3 cell system.58 In this system, 17b-E2, DES, and E1 induce a statistically significant cell transformation frequency. The natural steroidal estrogens require three- to five-fold higher concentrations than DES to induce an equivalent transformation frequency. OTHER CELL SYSTEMS In other systems studied,59 DES induced gene mutations in mouse lymphoma cells in the presence of a rat liver postmitochondrial supernatant, and unscheduled DNA synthesis in HeLa cells in the presence of the same postmitochondrial supernatant fraction. Sister chromatid exchanges have been induced in human fibroblasts and lymphocytes in culture by DES but not by 17b-E2 The major drawbacks of the cells used in most short-term systems include the fact that many are not primarily of epithelial origin, they are not considered target cells for sex hormones, and in many instances they are neoplastic. METABOLISM AND COVALENT BINDING STUDIES Investigations in both animal models and short-term in vitro cell culture assays provide indirect evidence for the bioactivation of sex hormones as a pertinent aspect of hormone-induced tumor cell transformation. In this regard, estrogens have been more extensively studied. Based on numerous reports, no doubt exists that estrogens can form reactive species capable of covalent binding to cellular macromolecules.60 Whether such reactive intermediates have any involvement in initiating oncogenesis in whole animal systems remains controversial, however, because of the high microsomal protein and hormone concentrations required to demonstrate their formation.61

GROWTH FACTOR AND ONCOGENE INVOLVEMENT An analysis of carcinogenesis, especially hormonal carcinogenesis, must include consideration of the possible role of growth factors and oncogenes in these processes.62 This is especially pertinent because many growth factors produced by normal cells are involved, singly or in combination with other mitogens, in the proliferation of specific target cells, both normal and neoplastic. Growth factors such as insulin-like growth factors (IGF) and transforming growth factors (TGF) can be produced by target cells. Also, the likelihood is that transformed cells may both synthesize and respond to growth factors and, consequently, proliferate independently through autocrine secretion. Thus, growth factors, which are basically peptide hormones, may be involved in the regulation of growth of both normal and neoplastic endocrine tissues. In vitro studies with serum-free media have clearly shown the proliferative effects of epidermal growth factor (EGF) and TGF-a on endocrine target cells. Oncogenes, including cellular protooncogenes, are thought to play an important role in carcinogenesis in animals and humans, perhaps through their proliferative functions. Oncogenes could participate in carcinogenesis in several ways. Some of them possibly may be coding for growth factors or their receptors. The onc gene of simian sarcoma virus (sis) is almost identical to a gene coding for a precursor of one polypeptide chain of platelet-derived growth factor (PDGF).63 The expression of the c-sis protooncogene is known to be under androgenic control in a ductus deferens smooth muscle tumor cell line (DDTMF-2). Moreover, these cells synthesize and secrete a PDGF-like growth factor that is implicated in the autocrine regulation of DDT1MF-2 cell proliferation.64 In addition, studies of the amino-acid sequence of immunoaffinity-purified EGF receptor have shown that the v-erb-B oncogene of avian erythroblastosis virus may encode for a truncated receptor lacking the external ligand-binding domain for EGF. These findings provide direct evidence that oncogenes may contribute to malignant cell transformation by inappropriate production of growth factors or through expression of uncontrolled growth factor receptor functions, causing unregulated cell proliferation. Although little is known about the expression of oncogenes by endocrine target cells, prolonged hormonal stimulation, a prerequisite for hormonal carcinogenesis, may cause inappropriate gene overexpression, amplification, and suppression. For estrogens, immediate estrogen response genes (e.g., c-myc, c-fos, c-jun) may be overexpressed.7 Moreover, cell-cycle genes as well as their regulatory genes (e.g., p16, p21, p27) may be deregulated and frequently overexpressed. In conclusion, many chemical and physical agents are known to be involved in carcinogenesis in animals and humans. These agents have been classified into various categories, such as initiators, promoters, cocarcinogens, and others. Hormones probably possess one or more of these characteristics, depending on the experimental model system in question. A unique and fundamental feature of carcinogenesis resulting from hormonal imbalance is the consistent finding that transformation usually follows a discrete pathway from normal cell hyperplasia to hormone-responsive and hormone-dependent neoplasia to hormone-independent neoplasia (i.e., autonomous tumors). Both hormone-induced and chemical carcinogen–induced tumors require a long latent period. Unlike hormonally induced cancers, however, cancers that are induced by chemical carcinogens in endocrine glands or in their target tissues usually do not depend on hormones for their growth. An exception to this general rule is the chemical carcinogen–induced mammary cancer in rats. What the nature of hormonal involvement is and whether hormones have a direct or indirect influence in one or more parts of the sequence of events leading to carcinogenesis remain to be elucidated. However, hormone-mediated genomic instability may be a key element common to a number of different hormonally induced model systems at various organ sites. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8.

Colditz GA, Stampfer MJ, Willett WC, et al. Prospective study of estrogen replacement therapy and risk of breast cancer in postmenopausal women. JAMA 1990; 264:2648. Toniolo PG. Endogenous estrogens and breast cancer risk: the case for prospective cohort studies. Environ Health Perspect 1997; 105:587. Ross RK, Bernstein L, Lobo RA, et al. 5-Alpha-reductase activity and risk of prostate cancer among Japanese and U.S. white and black males. Lancet 1992; 339:887. Colditz GA, Hankinson SE, Hunter DJ, et al. The use of estrogens and progestins and the risk of breast cancer in postmenopausal women. N Engl J Med 1995; 332:1589. Hertz R. An appraisal of the concepts of endocrine influence on etiology, pathogenesis, and control of abnormal and neoplastic growth. Cancer Res 1957; 17:423. Nandi S. Role of hormones in mammary neoplasia. Cancer Res 1978; 38:4046. Hou X, Li JJ, Chen WB, et al. Estrogen-induced protooncogene and suppressor gene expression in the hamster kidney: significance for estrogen carcinogenesis. Cancer Res 1996; 56:2616. Li JJ, Li SA. The effect of hormones on tumor induction. I. Brief overview of the endocrine system. II. Hormonal carcinogenesis. III. Effect of hormones on carcinogenesis by non-hormonal chemical agents. In: Arcos JC, Argus MF, Woo WT, eds. Chemical induction of cancer. Boston: Birkhauser, 1996:397. 9. Lacassagne A. Apparition de cancers de la mamelle chez la souris male, soumise á des injections de folliculine. Compt Rend Acad Sci 1932; 195:630. 9a. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

Lippert TH, Seeger H, Mueck AO. The impact of endogenous estradiol metabolites on carcinogenesis. Steroids 2000; 65:357. Shull JD, Spady TJ, Snyder M, et al. Ovary-intact, but not ovariectomized female ACI rats treated with 17b-estradiol rapidly develop mammary carcinomas. Carcinogenesis 1997; 18:1595. Li SA, Xue Y, Xie Q, et al. Serum and tissue levels of estradiol during estrogen-induced renal tumorigenesis in the Syrian hamster. J Steroid Biochem Mol Biol 1994; 48:283. Leav I, Ho S-M, Ofner P, et al. Biochemical alterations in sex hormone induced hyperplasia and dysplasia of the dorsolateral prostates of Noble rats. J Natl Cancer Inst 1988; 80:1045. Rudali G, Coezy E, Frederic F, et al. Susceptibility of mice of different strains to the mammary carcinogenic action of natural and synthetic oestrogens. Prev Europ Etudes Clin Biol 1971; 16:425. Blankenstein MA, Broerse JJ, van Zwieten MJ, et al. Prolactin concentration in plasma and susceptibility to mammary tumors in female rats from different strains treated chronically with estradiol 17b. Breast Cancer Res Treat 1984; 4:137. Thomas DB, Jimenez LM, McTieman A, et al. Breast cancer in men: risk factors with hormonal implications. Am J Epidemiol 1992; 135:734. Dunning WF, Curtis MR, Segaloff A. Strain differences in response to diethyl-stilbestrol and the induction of mammary gland and bladder cancer in the rat. Cancer Res 1947; 7:511 Cutts JH, Noble RL. Estrone-induced mammary tumors in the rat. I. Induction and behavior of tumors. Cancer Res 1964; 24:1116. Nandi S, Yang J, Guzman R. Hormones and the cellular origin of mammary cancer: a unifying hypothesis. In: Li JJ, Li SA, Gustafsson JA, et al., eds. Hormonal carcinogenesis, vol II. New York: Springer-Verlag, 1996:11. Hannouche N, Samperez S, Riviere MR, Jouan P. Estrogen and progesterone receptors in mammary tumors induced in rats by simultaneous administration of 17b-estradiol and progesterone. J

20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64.

Steroid Biochem 1982; 17:415. Lanari C, Molinolo AA, Dosne Pasqualini C. Induction of mammary adenocarcinomas by medroxyprogesterone acetate in BALB/c female mice. Cancer Lett 1986; 33:215. Varela, RM, Dao TL. Estrogen synthesis and estradiol binding by human mammary tumors. Cancer Res 1978; 38:2429. Longcope C, Femino A, Johnston ON. Androgen and estrogen dynamics in the female baboon (Papio anubis). J Steroid Biochem 1988; 31:195. Brodie AMH, Hammond JO, Ghosh M, et al. Effect of treatment with aromatase inhibitor 4-hydroxyandrostenedione on the nonhuman primate menstrual cycle. Cancer Res 1989; 49:4780. Dukes M, Edwards PN, Large M, et al. The preclinical pharmacology of “Arimidex” (Anastrozole; ZD1033)—a potent, selective aromatase inhibitor. J Steroid Biochem Mol Biol 1996; 58:439. Lipschutz A, Iglesias R, Pamosevick V, et al. Ovarian tumors and other ovarian changes induced in mice by two 19-nor contraceptives. Br J Cancer 1967; 21:153. Biskind GR, Kordan B, Biskind MS. Ovary transplanted to spleen in rats: the effect of unilateral castration, pregnancy, and subsequent castration. Cancer Res 1950; 10:309. Griffiths CT. Effects of progestins, estrogens, and castration on induced endometrial cancer in rabbits. Surg Forum 1963; 14:399. Highman B, Greeman DL, Norvell MJ, et al. Neoplastic and preneoplastic lesions induced in female C3H mice by diets containing diethylstilbestrol or 17b-estradiol. J Environ Path Toxicol 1980; 4:81. Newbold RR, Bulock BC, McLachlan JA. Uterine adenocarcinomas in mice following developmental treatment with estrogens: a model for hormonal carcinogenesis. Cancer Res 1991; 50:7677. Leavitt WW, Evans RW, Hendry WJ III. Etiology of DES-induced uterine tumors in Syrian hamster. In: Leavitt WW, ed. Hormones and cancer. New York: Plenum Publishing, 1982:63. Kirkman H, Algard FT. Characteristics of an androgen/estrogen-induced uterine smooth muscle cell tumor of the Syrian hamster. Cancer Res 1970; 30:794. McClure HM, Graham CE. Malignant uterine mesotheliomas in squirrel monkeys following diethylstilbestrol administration. Lab Animal Sci 1973; 23:493. Bischoff F. Carcinogenic effects of steroids. Adv Lipid Res 1969; 7:165. Banerjee SK, Zoubine MH, Sarkar DK, et al. 2-Methoxy estradiol blocks estrogen-induced rat pituitary tumor growth and tumor angiogenesis: possible role of vascular endothelial growth factor. Anticancer Res 2000; 20:2641. Nakagawa K, Ohara T, Tashiro K. Pituitary hormones and prolactin-releasing activity in rats with primary estrogen-induced pituitary tumors. Endocrinology 1980; 106:1033. Koneff AA, Simpson ME, Evans HM. Effect of chronic administration of diethylstilbestrol on the pituitary and other endocrine organs of hamsters. Anat Rec 1946; 94:169. Bonser GM, Robison JM. The effects of prolonged estrogen administration upon male mice of various strains: development of testicular tumors in the strong A strain. J Pathol Bacteriol 1940; 51:9. Sato B, Spomer W, Huseby RA, Samuels LT. The testicular estrogen receptor system in two strains of mice differing in susceptibility to estrogen-induced Leydig cell tumors. Endocrinology 1979; 104:822. Kirkman H. Estrogen-induced tumors of the kidney. III. Growth characteristics in the Syrian hamster. Natl Cancer Inst Monogr 1959; 1:1. Li JJ, Cuthbertson TL, Li SA. Inhibition of estrogen carcinogenesis in the Syrian golden hamster kidney by antiestrogens. J Natl Cancer Inst 1980; 64:795. Li JJ, Hou X, Bentel JM, et al. Prevention of estrogen carcinogenesis in the hamster kidney by ethynylestradiol: some unique properties of a synthetic estrogen. Carcinogenesis 1998; 19:471. Oberley TD, Gonzalez A, Lauchner LJ, et al. Characterization of early lesions in estrogen-induced renal tumors in the Syrian hamster. Cancer Res 1991; 51:1922. Li JJ, Li SA, Klicka JK, et al. Relative carcinogenic activity of various synthetic and natural estrogens in the hamster kidney. Cancer Res 1983; 43:5200. Li SA, Hou X, Li JJ. Estrogen carcinogenesis: a sequential, epigenotoxic multistage process. In: Li JJ, Li SA, Gustafsson JA, et al., eds. Hormonal carcinogenesis, vol II. New York: Springer-Verlag, 1996:200. Li JJ, Kirkman H, Li SA. Synthetic estrogens and liver cancer: Risk analysis of animal and human data. In: Li JJ, Nandi S, Li SA, eds. Hormonal carcinogenesis, New York: Springer-Verlag, 1992:217. Li JJ, Li SA. High incidence of hepatocellular carcinoma after synthetic estrogen administration in Syrian hamsters fed a-naphthoflavone: a new tumor model. J Natl Cancer Inst 1984; 73:543. Drago JR. The induction of Nb rat prostatic carcinomas. Anticancer Res 1984; 4:255. Ofner P, Bosland MC, Vena RL. Differential effects of diethylstilbestrol and estradiol-17b in combination with testosterone on rat prostate lobes. Toxicol Appl Pharmacol 1992; 112:300. Bruchovsky N, Lesser B. Control of proliferative growth in androgen responsive organs and neoplasms. Adv Sex Horm Res 1976; 2:1. Pollard M, Snyder DL, Lochert PH. Dihydrotestosterone does not induce prostate adenocarcinoma in L-W rats. Prostate 1987; 10:325. Ho S-M, Yu M, Leav I, Viccione T. The conjoint action of androgens and estrogens in the induction of proliferative lesions in the rat prostate. In: Li JJ, Nandi S, Li SA, eds. Hormonal carcinogenesis. New York: Springer-Verlag, 1992:18. Kirkman, H, Algard FT. Androgen-estrogen-induced tumors I. The flank organ (scent gland) chaetepithelioma of the Syrian hamster. Cancer Res 1964; 24:1569. Bern HA, Talamantes FJ. Neonatal mouse models and their relation to disease in the human female. In: Herbst SL, Bern HA, eds. Developmental effects of diethylstilbestrol (DES) in pregnancy. New York: Thieme-Stratton, 1981:129. Holtzman S, Stone JP, Shellabarger CJ. Synergism of estrogens and x-rays in mammary carcinogenesis in female ACI rats. J Natl Cancer Inst 1981; 67:455. Russo IH, Russo J. Mammary gland neoplasia in long-term rodent studies. Environ Health Perspect 1996; 104:938. Li JJ. Perspectives in hormonal carcinogenesis: animal models to human disease. In: Huff J, Boyd J, Barrett JC, eds. Cellular and molecular mechanisms in hormonal carcinogenesis: environmental influences. Philadelphia: Wiley-Liss 1996:447. Tsutsui T, Degen GH, Schiffmann D, et al. Dependence on exogenous metabolic activation for induction of unscheduled DNA synthesis in Syrian hamster embryo cells by diethylstilbestrol and related compounds. Cancer Res 1984; 44:184. Friedrich U, Thomale J, Nass G. Induction of malignant transformation by various chemicals in BALB/3T3 clone A31-1-1 cells and biological characterization of some transformants. Mutation Res 1985; 152:113. Rao PN, Engelberg J. Structural specificity of estrogens in the induction of mitotic chromatid non-disjunction in HeLa cells. Exp Cell Res 1967; 48:71. Tsibris JCM, McGuire PM. Microsomal activation and binding to nucleic acids and proteins. Biochem Biophys Res Commun 1977; 78:411. Beleh MA, Lin YC, Brueggemeier RW. Estrogen metabolism in microsomal, cell, and tissue preparations from kidney and liver from Syrian hamster. J Steroid Biochem Mol Biol 1995; 52:479. Schavorovsky OG, Rozados VR, Gervasoni SI, Matar P. Inhibition of ras oncogene: a novel approach to antineoplastic therapy. J Biomed Sci 2000; 7:292. Waterfield MD. Oncogenes may encode a growth factor or part of the receptor for a growth factor. Br J Cancer 1984; 50:242. Smith RG, Nag A, Syms AJ, Norris JS. Steroid regulation of receptor concentration and oncogene expression. J Steroid Biochem 1986; 24:51.

CHAPTER 223 SEX HORMONES AND HUMAN CARCINOGENESIS: EPIDEMIOLOGY Principles and Practice of Endocrinology and Metabolism

CHAPTER 223 SEX HORMONES AND HUMAN CARCINOGENESIS: EPIDEMIOLOGY ROBERT N. HOOVER Carcinogenesis and Endogenous Sex Hormone Status Carcinogenesis and Exogenous Sex Hormone Therapy Endometrial Cancer Endogenous Factors in Endometrial Cancer Exogenous Sex Hormones and Endometrial cancer Mechanisms of Action Breast Cancer Endogenous Factors in Breast Cancer Influence of Diet Hypotheses for the Hormonal Causation of Breast Cancer Exogenous Sex Hormones and Breast cancer Future Imperatives Ovarian Cancer Endogenous Factors in Ovarian Cancer Hormone-Replacement Therapy and Ovarian Cancer Oral Contraceptives and Ovarian Cancer Infertility Treatment Mechanisms Cancer of the Uterine Cervix Risk Factors and Cervical Cancer Oral Contraceptives and Cervical Cancer In Utero Diethylstilbestrol Exposure and Cervical Cancer Other Gynecologic Cancers and Exogenous Sex Steroids Male Genital Cancers and Sex Steroids Liver Cancer and Sex Steroids Androgenic-Anabolic Steroids and Liver Cancer Oral Contraceptives and Benign Liver Tumors Oral Contraceptives and Liver Cancer Other Tumors Chapter References

Because of the central role that the hormonal milieu plays in various carcinogenic processes, clinical endocrinologists must be aware of malignancies to which their patients may be predisposed, either because of the nature of their illness or because of the nature of the hormonal therapy being instituted.

CARCINOGENESIS AND ENDOGENOUS SEX HORMONE STATUS Endogenous hormone status has long been thought to be an important factor in the etiology of a number of human malignancies. This belief has been based on animal carcinogenesis studies (see Chap. 222), the responsiveness of a number of tumors to hormonal manipulation (see Chap. 224 and Chap. 225), the relationship of risk of certain tumors to a variety of reproductive and other factors thought to influence hormonal status, and the simple fact that some organs depend on hormonal status for their normal function.1 Speculation about a causal role for hormones has focused on malignancies of the female breast and the reproductive tract. Some evidence for hormonal carcinogenesis has been observed for a variety of other tumors, however, including prostate, liver, testis, thyroid, and gallbladder cancers, and malignant melanoma. Despite these long-standing suspicions, little success has been achieved in identifying the specific hormonal factors that might be responsible for these tumors, with the possible exception of endometrial cancer.

CARCINOGENESIS AND EXOGENOUS SEX HORMONE THERAPY Within the last 50 years, a new element in the area of hormonal influences on cancer risks has been added, that of exogenous sex hormone exposure. Pharmacologic levels of estrogens, progestins, androgens, and pituitary trophic hormones, alone or in combination, have been administered to large segments of the population for various reasons. These large-scale “natural experiments” have provided more specific insights into the relationship between hormonal factors and several different malignancies.2 Moreover, enthusiasm has grown for the widespread treatment of relatively healthy segments of the population (e.g., women receiving oral contraceptive agents or menopausal replacement therapy). Considerable interest has arisen in the use of estrogens for post-menopausal prevention of osteoporosis and osteoporotic fractures3 (see Chap. 64 and Chap. 100). Some evidence supports the long-suspected potential of menopausal estrogens to prevent clinical coronary heart disease.4 In addition, within the general population, a substantial increase has been seen in the use of dietary supplements, many of which have significant hormonal activity (e.g., androstenedione, melatonin). Because of this enthusiasm on the part of physicians and the public, appropriate evaluations of the carcinogenic consequences of these exposures have become important to public health, as well as to understanding the biology of the tumors involved.

ENDOMETRIAL CANCER ENDOGENOUS FACTORS IN ENDOMETRIAL CANCER The cancer for which the evidence for both an endogenous and an exogenous hormonal cause is best established is endometrial cancer. Various factors related to endogenous hormone production have been associated with endometrial cancer.5 Medical conditions related to increased risk include functional (estrogen-secreting) ovarian tumors, the polycystic ovary syndrome, diabetes mellitus, and hypertension. Reproductive factors, including nulliparity and a late natural menopause, also have consistently been found to be related to increased risk. Some dietary factors also seem to influence risk. Obesity is a risk factor and a vegetarian diet is a possible protective factor.6 Age, a determinant of levels of most endogenous hormones, also influences endometrial cancer risk in a unique manner. Endometrial cancer rates are extremely low in women younger than 45 years of age, rise precipitously among women in their late 40s and throughout their 50s (much more dramatically than for other tumors), and then decline in women approximately age 60 and older (Fig. 223-1).

FIGURE 223-1. Age-specific incidence rates for breast and uterine corpus cancers among white women during 1986 through 1990. (Data from the Surveillance, Epidemiology and End Results Program.)

EXOGENOUS SEX HORMONES AND ENDOMETRIAL CANCER

Exposure to exogenous hormones also has been linked to endometrial cancer.5 ESTROGENS AND ENDOMETRIAL CANCER Estrogen-replacement therapy of 2 years or longer for menopausal women is associated with an excess relative risk of endometrial cancer. Table 223-1 shows estimated relative risks (i.e., the risk of the disease among those exposed to estrogen therapy compared with the risk among those not exposed).7,8,9,10,11,12,13 and 14 The relative risk among users compared with nonusers ranges from two-fold to eight-fold. It increases even further with long duration of use and with high average daily doses. Thus far, every type of estrogen that has been investigated has shown this relationship, including conjugated equine estrogens, ethinyl estradiol, and diethylstilbestrol (DES). The highest risk occurs among current users. The risk declines with each year after cessation of use, although apparently some residual excess risk is present even 10 years after cessation. The risk is highest for early-stage malignancies, but a two-fold to three-fold excess risk is seen for the advanced stages of disease as well.

TABLE 223-1. Relative Risks* of Endometrial Cancer Associated with Menopausal Estrogen Use from Selected Case-Control Studies

EFFECT OF ESTROGEN AND PROGESTERONE IN SEQUENCE A profound trend has been seen away from unopposed estrogen treatment of menopausal symptoms and toward treatment with a sequence of an estrogen that is then combined with a progestin. Substantial evidence15 indicates that such cyclic treatment reduces the frequency of hyperplasia and atypical hyperplasia associated with unopposed estrogen treatment. Although the epidemiologic data concerning endometrial cancer risk are still developing, certain patterns are emerging. The risk of endometrial cancer is lower among women using the combined regimen than among women using estrogen alone.10,16 Evidence implies that, at least in the short term, the risk is related to the number of days that a progestin is used with estrogen in a monthly cycle. Those using the progestin for ³10 days per month, including those using the combined regimen continuously, have a risk similar to that of women not using any hormone-replacement therapies. Those using progestins for 317,000 new cases and 41,400 deaths due to prostate cancer.1 Concepts about the endocrine dependence of prostate cancer result largely from the work of Dr. Charles Huggins, who postulated that endocrine-dependent tumors in general, and prostate cancer in particular, contain malignant cells that require hormones for cellular viability and proliferative capacity.2 Thus, treatments designed to induce hormone deprivation would be expected to produce tumoricidal effects and the amelioration of clinical signs and symptoms, if not cure. He demonstrated dramatic regressions of prostate cancer in response to surgical orchiectomy or suppression of androgens with estrogens such as diethylstilbestrol (DES). These observations provided the rationale for endocrine therapy for a number of tumors. Although the hope for cure was not fulfilled, his conceptual approach led to a wide variety of studies, including tumor induction with hormones, receptors as mediators of hormone action, and the use of antihormones to block the mitotic effects of androgens on tumor growth. INCIDENCE Prostate cancer is detected initially in one of three ways: (a) during investigation of suspicious signs and symptoms, (b) incidentally at the time of surgery for other reasons, and (c) at autopsy. The yearly incidence differs for each mode of detection. Prostate cancer is suspected in patients with symptoms of bone pain caused by metastases or with a firm nodule on routine prostate examination. This form of clinically evident prostate cancer is second only to that of the lung in frequency among men in the United States, and ~300,000 new cases are diagnosed yearly. Each year, ~40,000 men die of prostate cancer after its clinical presentation—a number that represents 10% of all male cancer deaths. The second type, occult prostate cancer, is found incidentally at the time of transurethral resection of the prostate, performed as treatment for presumed prostatic hypertrophy. Several thousand new cases are detected yearly in this manner. Only a small fraction of men die from this relatively benign form of prostate cancer. The third mode of detection, which involves the routine examination of the prostate at autopsy, uncovers prostate cancer in >30% of men older than 60 years of age. Autopsy-detected prostate cancer3 is clinically insignificant because symptoms are rarely present before death. These cancers are localized and are never the cause of death of the patient. The three modes of detection of prostate cancer emphasize its wide biologic spectrum of aggressivity. This fundamental concept has important clinical implications for the choice of appropriate therapy. Highly variable biologic behavior is not unique to prostate cancer; it is also a characteristic of the endocrine-dependent tumors of thyroid, breast, and endometrium. Thyroid and endometrial cancers, but not breast cancers, frequently are detected for the first time at autopsy. ETIOLOGY The causes of prostate cancer remain unknown.4 A higher than expected prevalence among relatives suggests genetic factors,4a but histocompatibility antigen typing has shown no confirmatory associations. The higher incidence and mortality for American blacks than for whites also raise the possibility of genetic interactions. Environmental factors or ascertainment bias also could explain this difference, however, and this possibility is suggested by the six-fold lower rate of prostate cancer in Nigerian blacks than in American blacks. Androgens have been considered as possible promoters or initiators of prostate cancer in men. Support for this hypothesis includes the following: eunuchs rarely develop prostate cancer; exogenous androgens or estrogens can induce prostate cancer in an animal model; and, in humans, most prostatic cancers are hormone dependent. However, no consistent abnormalities of androgen production, metabolism, or circulating levels or tissue sensitivity to testosterone have been observed in men with prostate cancer. Exposure to chemicals, dietary factors, or sexual transmission of an infecting agent remains an additional, albeit unproved, etiologic possibility. Protein translation products (i.e., p21 protein) of the c-ras oncogene have been detected in prostate cancer tissue. This is probably the result of secondary oncogene activation during the process of tumor evolution, rather than indication of a prior retrovirus infection.5 A role for inactivation of tumor suppressor genes has also been proposed.6 PATHOLOGY The analysis of tissue sections provides a potential means of predicting the biologic behavior of prostate cancer. Many pathologic classification systems are available. Histologic criteria developed by Gleason7 included the degree of glandular differentiation and structural architecture of a specimen viewed at low-power magnification. The Gleason score correlates with patient prognosis when groups of men are studied8 (Fig. 225-1). The scoring methods developed later added the features of cellular anaplasia, the degree of nuclear roundness, and the appearance of nucleoli as additional criteria. Critiques of each method emphasize problems with the lack of scoring reproducibility among individual observers, the variability of grades within the same tumor, and the lack of predictive power for the individual patient as opposed to groups of patients. Hence, histologic grading scores generally have not been used as the basis for making therapeutic decisions, and no single grading system is universally accepted. Nonetheless, the clinician may wish to use the information gained from histologic assessment when determining the aggressiveness of the therapeutic strategy to be chosen for an individual patient.9,10 Nuclear DNA ploidy has been proposed as an important and independent prognostic variable for patients with stage C and D1 disease.11,12 Patients having tumors with DNA tetraploid or aneuploid patterns experienced tumor progression sooner and died earlier than patients with diploid tumors.11,12

FIGURE 225-1. A, Histologic grading system of Gleason.7 Tumors are graded 1 to 5 based on the degree of glandular differentiation and structural architecture. Tumors commonly contain more than one histologic grade. To take this characteristic into account, the grades are added together and become the Gleason histologic score. This provides the scaling effect of averaging, but without division by 2. In general, histologic grades 1 and 2 become histologic scores 3 and 4 as shown on panel B. B, Correlation of Gleason histologic score with cancer death rate per year. (Modified from Gleason DF. Histologic grading and clinical staging of carcinoma of the prostate. In: Tannenbaum M, ed. Urological pathology: the prostate. Philadelphia: Lea & Febiger, 1977:171.)

The detection of an early premalignant lesion has been difficult, but one study has suggested that pS2 immunoreactivity could be useful for the diagnostic evaluation of early premalignant lesions in the setting of a negative tumor biopsy.13 Prostate tissue obtained from patients without malignant disease consistently lacked pS2 protein expression. In contrast, nonneoplastic prostatic tissue from patients with locally advanced prostate cancer exhibited a variable degree of pS2 expression in the normal or hyperplastic gland and in intraepithelial hyperplasia adjacent to the cancerous lesion. Also, the expression of this pS2 was closely associated with neuroendocrine differentiation. From many ongoing studies, several other gene markers are expected to be found to facilitate the determination of the degree of tumor progression. CLINICAL STAGING Clinical staging of prostate cancer provides the major means of determining prognosis and is widely used (Fig. 225-2). Treatment decisions depend heavily on this parameter.

FIGURE 225-2. Clinical staging systems in general use. Stage A represents occult disease found at the time of transurethral resection of the prostate for the suspected clinical diagnosis of benign prostatic hypertrophy. Subsets A1 and A2 depend on the number of tissue chips found to contain cancer. Stage B disease is palpable clinically and is divided into subsets representing various degrees of involvement of the prostate. Stage C represents local extraprostatic metastasis contiguous to the gland, and Stage D, distant lymph node, bone, or other metastases. Approximate percentage of patients dead at 5 years is indicated by the heavy squares containing percentages drawn for each category. Prognostic data represent a compilation of information from varying time intervals. (Summarized from Catalona WJ, ed. Endocrine therapy. In: Prostate cancer. New York: Grune & Stratton, 1984:40.)

Patients thought to have local disease after initial clinical assessment often have occult spread to lymph nodes or to distant sites at the time of surgical exploration for staging. Therefore, surgical staging is required when the results would influence treatment decisions. Initially, noninvasive studies, including skeletal survey, intravenous pyelogram, measurement of serum acid phosphatase levels and levels of more sensitive markers such as serum prostate-specific antigen14 values, bone scans, and, when indicated, magnetic resonance imaging and sonograms14a are ordered. The finding of metastases or extensive local spread by these methods obviates the need for surgical staging. If these test results are negative, however, then surgical exploration of the lymph nodes usually is required. The yield of this procedure in detecting occult spread is 2% in stage A1, 23% in stage A2, 18% in stage B1, 35% in stage B2, and 46% in stage C.15 Surgical staging is important particularly for patients with disease of clinical stages A2 and B who are being considered for curative surgical therapy and who would not be candidates for such treatment if positive nodes were detected at the time of surgical exploration.

TUMOR BIOLOGY A stem cell model has been proposed for the organization of the prostatic epithelium. This model explains the growth and transformation of normal epithelium into cancer cells.16 According to this model, the prostate has three basic layers: secretory luminal, basal, and endocrine paracrine cells. The proliferative compartment, which is located in the basal cell layer, usually is androgen independent but contains androgen-responsive target cells. During the malignant transformation, the proliferative zone is shifted to luminal cell types with formation of neoplastic basement membranes. These proliferative changes in prostate cancer are exclusively restricted to exocrine cell types, and the majority of exocrine cells are androgen dependent, whereas endocrine-differentiated cells lack the nuclear androgen receptor. A small stem cell population located in the basal cell layer is the source for all epithelial layers in the normal, hyperplastic, and neoplastic cells. This differentiating process (from basal cells to secretory luminal cells) via intermediate phenotypes is induced by circulating androgen and largely depends on the androgen-responsive target cells in the basal cell layers. The biologic behavior of neuroendocrine lineage cells is quite different from that of exocrine lineage cells and usually is more aggressive. A study of a transgenic mouse model of metastatic prostate cancer has shown that the neuroendocrine cell lineage (from several cell types) is exquisitely sensitive to transformation.17 The simian virus 40 T antigen was expressed in a subset of neuroendocrine cells in all lobes of mouse prostate. After 7 weeks, prostatic intraepithelial neoplasias developed and rapidly progressed to local invasion. After 6 months, these tumors metastasized to lymph nodes, liver, lung, and bone; this tumor was not androgen dependent. Neuroendocrine differentiation was strongly associated with the progression of the disease in the presence of endocrine therapy.18 At the time of clinical diagnosis of this tumor, heterogeneous subpopulations of malignant cells with varying characteristics are present.19,20 Some of the cells are absolutely hormone dependent and die when deprived of their androgen support. Others, the hormone-sensitive population, grow faster when stimulated with androgen but become quiescent and enter the G0 (resting) stage of the cell cycle on androgen deprivation. Still other clones of cells (hormone independent) grow in the complete absence of androgenic stimulation. The relative numbers of each of these three cell types determine whether the initial response of the patient to androgen deprivation consists of objective tumor regression, stabilization of disease, or continued progression. In general, when highly stringent criteria are applied, 40% of men with prostate cancer exhibit objectively measurable tumor regression and a further 40% experience disease stabilization when deprived of androgens.21,22 At most, only 10% to 20% of tumors continue to grow after androgen deprivation and, thus, consist exclusively of hormone-independent cells. The proportion of each cell type changes progressively in response to hormonal treatment. After initial hormonal deprivation, the subsets of cells that are absolutely hormone independent at the outset continue to grow and gradually repopulate the tumor. Hence, hormonal treatment does not cure patients, and therapy is nearly always followed, within months to years, by relapse. In all patients, the hormone-independent cells represent almost all of those within the tumor at the time of the patient's death.

The natural history of prostatic cancer cells entering G0 during hormonal deprivation (i.e., the hormone-sensitive population) is less well known. The readministration of androgens after initial deprivation elicits marked increases in bone pain and objective tumor progression in nearly 80% of patients with prostate cancer.23 This observation provides evidence for the persistence of hormone-sensitive cells in most patients. In some patients, these cells appear to regrow in the presence of much lower amounts of androgen. Objective responses to secondary hormonal therapy, such as hypophysectomy, adrenalectomy, antiandrogens, or medical adrenalectomy, probably occurs from a response of the substantial fraction of such cells that had begun to regrow. GENETICS Several studies have provided insights regarding the relationships between tumor suppressor or transcriptional factors and the development, progression, or endocrine characteristics of prostate cancer. The study of genetic changes with comparative hybridization techniques has shown that losses of several chromosomal sites (8p, 13q, 6q, 16q, 18q, 9q) are common changes in primary tumors, a finding which suggests that deletion-inactivation of putative tumor-suppressor genes at these sites is likely to contribute to the development of prostatic cancer.24 Likewise, through the identification of homozygous deletions in malignant tissue, chromosome band 12p12-13 has been suggested as a potential location of a novel tumor-suppressor gene for advanced prostate cancer.25 Although the gene has not yet been isolated, this study should facilitate the identification of genes involved in the progression of prostate adenocarcinoma. Also, p53 mutations are frequently associated with increased metastases. In one study, these mutations were found at approximately twice the frequency in metastatic prostate cancer as in unselected samples of primary prostatic cancer.26 Immunohistochemical staining for the expression of the KAI1 gene (a metastatic suppressor gene for prostatic cancer) correlates with tumor characteristics.27 In benign prostatic hyperplasia tissues, KAI1 protein is uniformly expressed in the glandular cell membrane at cell-to-cell borders. This protein is similarly expressed in untreated prostate cancer, but the percentage of protein-positive cells correlates inversely with the Gleason pattern and clinical stage. Also, this protein was not expressed in either primary or metastatic cancer tissue from patients who died after relapse from endocrine therapy. Presently, the molecular mechanism for the growth-promoting effects of androgens is not clearly understood, but one study has shown a close correlation between pRB gene expression (one of the transcriptional factors involved in the suppression of carcinogenesis) and androgen-receptor activity.28 In this study, over-expression of the pRB gene increased androgen-receptor transcriptional activity; loss of pRB protein inhibited androgen-receptor functions. Also, the suggestion has been made that the loss of pRB activity during the progression of cancer may directly decrease the response to androgens. Likewise, a transcriptional factor, ETS2, is critical to the maintenance of the transformed state of human prostate cancer cell lines. When the action of this protein was blocked, the transformed properties were reduced.29 HORMONE DEPENDENCE Androgenic effects mediated through the action of dihydrotes-tosterone are responsible for stimulation of prostatic cancer growth. The exact nature of the stimulatory process is unknown, and it may reflect a secondary enhancement of androgen-dependent growth factors or direct proliferative effects of the androgens themselves. Nonetheless, dihydrotestosterone must first bind to nuclear receptors in prostate carcinoma tissue to initiate the events leading to tumor growth. Thus, therapies for prostate cancer are designed to lower the tissue levels of dihydrotes-tosterone or to antagonize its action at the receptor level. Human prostate cancer also may be estrogen dependent. Estrogen receptors can be demonstrated in some human prostate cancers.30 Objective responses, albeit in 500 fmol/mg DNA), responses to endocrine therapy often persisted for >1 year, whereas relapse occurred within 1 year in all patients with receptor-poor tumors (i.e., levels of 24 Gy place patients at high risk of developing GH deficiency.5,9 The central nervous system is more sensitive to radiation at an early age; for any given dose of irradiation, the incidence of GH deficiency is lower in adults than in children, and the very young child who receives craniospinal irradiation is most at risk of extreme short stature. In adults, hyperprolactinemia or deficiencies of gonadotropins, TSH, or adrenocorticotropic hormone (ACTH) are very uncommon at doses under 40 to 50 Gy.3 Slowed growth velocity that is inappropriate for a child's age and stage of puberty is a very common, but not universal, outcome of GH deficiency in this setting. Age at irradiation is correlated not only with final height, with the youngest at irradiation having the worst growth prognosis, but also with the age at onset of puberty.10 The majority of subjects who experience premature sexual maturation also have GH deficiency, and early puberty contributes to their poor growth.3,10 An inverse relationship exists between time since therapy and stimulated peak GH responsiveness11 (Fig. 226-1). In contrast to findings in patients with isolated idiopathic GH

deficiency, young adults with radiation-induced GH dysfunction rarely revert to normal GH status.12

FIGURE 226-1. Relationship between the time since cranial irradiation (XRT) and the peak growth hormone (GH) response to an insulin-tolerance test in 32 adults who had received XRT in childhood as part of their treatment for acute lymphoblastic leukemia. The fitted lines are from the regression of log 10 of the peak GH response on time since therapy and dose group. Solid circles depict the subjects who received 18 Gy XRT; solid squares depict the subjects who received 24–25 Gy XRT. (Reprinted from Brennan BMD, Rahim A, Mackie EM, et al. Growth hormone status in adults treated for acute lymphoblastic leukaemia in childhood. Clin Endocrinol 1998; 48:777.)

Little direct evidence exists that cytotoxic drugs impair anterior pituitary function (see Table 226-1). Growth deceleration in children has been seen to occur during antileukemic multimodality therapy, however, with some degree of “catchup” growth occurring after completion of chemotherapy.13 Moreover, final height as well as growth velocity after craniospinal irradiation for brain tumors is more profoundly affected in children who have received adjuvant chemotherapy than in those receiving craniospinal irradiation alone, suggesting potentiation of radiation-induced growth failure by the chemotherapy.10 Cancer patients receiving chemotherapy regimens that include short-term, highdose courses of corticosteroids are at risk for the development of adrenal suppression, which does not necessarily correlate with either the corticosteroid dosage or the duration of therapy. Impaired release of other pituitary hormones in response to provocative stimulation also can occur in such patients and is not inconsistent with the well-known multiple effects of corticosteroids on hypothalamic-pituitary function. PREVENTION AND TREATMENT Routine hypothalamic-pituitary shielding during cranial radiation, dose targets of £18 Gy, and, in patients undergoing BMT, fractionation of TBI are measures that can provide some degree of endocrine protection. Nevertheless, all patients who have received incidental hypothalamic-pituitary irradiation should undergo periodic evaluation for the integrity of the hypothalamic-pituitary target organ axes. In survivors of childhood leukemia, growth deceleration may occur after an interval of improved growth; therefore, all such patients should be followed until they have attained their final heights. As false-negative results may be common in children who have demonstrated slowed growth velocities after such irradiation, the wise course is to conduct pharmacologic testing for GH reserve with at least two agents, preferably including a carefully supervised insulin-induced hypoglycemic stimulus, which may provide a discriminatory advantage. A single plasma determination of insulin-like growth factor–binding protein-3 (IGFBP-3) is not a useful screening test for GH deficiency in this population. Pituitary size visualization with magnetic resonance imaging correlates with nocturnal GH secretion, but its utility in evaluating GH secretory capacity is unknown.14 In children with documented GH deficiency and growth retardation caused by HPA irradiation, GH replacement therapy should be initiated as promptly as is prudent—generally, in most institutions, after a 1-year disease-free interval. Although GH replacement can cause a gratifying increase in growth velocity, the final height, especially in children who have received craniospinal irradiation, is often significantly less than the midparental height. Concomitant therapy to suppress puberty should be considered in GH-deficient patients with early puberty who have a poor predicted final height. Traditionally, GH therapy in children has been discontinued when final height has been achieved. However, awareness is growing of adverse effects of adult GH deficiency. Altered body composition, physical performance, psychological well-being, and substrate metabolism, as well as an increased risk of cardiovascular disease and osteopenia, have been reported.15 GH replacement corrects or improves many of these alterations, although longer-term effects have not been fully addressed. Although data from long-term studies of children with both solid tumors and hematologic malignancies suggest that no increased risk of cancer recurrence is associated with GH therapy,15 the safety of long-term GH replacement in an adult cancer population remains to be established. Replacement hormonal therapy should be instituted in patients shown to be hypothyroid, hypoadrenal, or hypogonadal. Infusion therapy with GnRH can induce ovulatory cycles in a small number of patients after cranial irradiation.6 Overt clinical adrenal insufficiency after short-term, high-dose corticosteroid administration is rare, but patients so treated should undergo testing of adrenal reserve and are candidates for corticosteroid coverage if exposed to acute stress. Although the highest incidence of endocrine complications after hypothalamic-pituitary irradiation occurs within the first 5 years after therapy, life-long surveillance in this population appears to be warranted.

THERAPY-INDUCED THYROID GLAND DISEASE THYROID TUMORS Abnormalities of thyroid morphology, including focal hyperplasia, single or multiple adenomas, chronic lymphocytic thyroiditis, colloid nodules, and fibrosis, are frequently found in patients exposed to radiation for nonmalignant conditions. Palpable thyroid lesions occur in 20% to 30% of an irradiated population, whereas a 1% to 5% prevalence of palpable nodular thyroid disease is found in the general population.16 In patients who received head and neck irradiation for childhood cancer (thyroid dose, 22.5–40 Gy), thyroid sonography performed a decade later detects widespread abnormalities.17 Nearly all patients have diffuse atrophy; half have discrete nodularity, with 39% developing new focal abnormalities at followup 6 to 18 months later. The thyroid is one of the organs most sensitive to the neoplastic effects of radiation. Direct or incidental thyroid irradiation increases the risk of well-differentiated thyroid cancer, usually papillary, with an excess relative risk of 7.7 per Gy reported in a pooled analysis of seven studies in nearly 120,000 subjects.18 In patients treated for childhood malignancies, the risk of secondary thyroid cancer is increased 15- to 53-fold.19,20 Thyroid cancer risk increases with duration of follow-up. Although it appears to peak at 15 to 19 years after treatment, an excess risk is still apparent at 40 or more years after treatment.18 At lower-dose exposures, the dose response is linear, even down to 0.10 Gy, with some indication of a leveling off at the high end of the radiation dose curve (³60 Gy).18,20 The risk is highest after radiation at a young age (Fig. 226-2). A pattern of oncogene involvement that may be characteristic of radiation-induced thyroid cancers is suggested by detection of p53 mutations (usually confined to anaplastic thyroid cancers) and a distinct pattern of ret oncogene rearrangements, a highly prevalent molecular alteration in thyroid tumors that developed in children after the Chernobyl reactor accident.21,22 Data23 suggest that a DNA repair defect is likely to be present in patients with radiation-associated thyroid tumor. The mortality does not seem to differ from that seen in nonirradiated thyroid cancer patients.19

FIGURE 226-2. Relative risk of thyroid cancer after exposure to external radiation: a pooled analysis of seven studies.18 (Reprinted from Ron E, Saftlas AF. Head and neck radiation carcinogenesis: epidemiologic evidence. Otolaryngol Head Neck Surg 1996; 115:403.)

Iodine-131 (131I), used in the diagnosis and treatment of Graves disease and thyroid cancer for nearly 50 years, remains the most frequently used radioisotope for radioimmunotherapy in cancer patients. Previous epidemiologic studies of patients who had been treated with 131I failed to reveal an increased risk of thyroid cancer.24 Although most of these patients were adults, children whose thyroid glands had received as much as 2 Gy of131I irradiation also showed no increased risk of thyroid cancer. These findings suggested that 131I was less carcinogenic than acute exposure to x-rays. However, public interest in the late health effects of 131I was rekindled by the 1986 Chernobyl nuclear reactor accident, which released very large amounts of131I and shortlived radioiodines into the atmosphere. In those exposed to thyroid doses of ³0.1 to >10 Gy, after a short latency period, a markedly increased rate of childhood papillary thyroid cancer has now been reported, with a calculated excess relative risk of 22 to 90 per Gy. 24,25 A small but statistically significant increased risk in thyroid cancer mortality after 131I treatment for adult hyperthyroidism has been reported after long-term follow-up (mean, 21 years).26 No excess risk of thyroid cancer has been described after exposure to alkylating agents or vinca alkaloid treatment, when radiation exposure is controlled for; however, a suggestion is seen of increased risk (relative risk, 39; 95% confidence interval, 1.6–947) in those receiving higher dose radiation (³10 Gy) and dactinomycin than in those receiving radiation alone.20 THYROID DYSFUNCTION Primary hypothyroidism is the most common clinical consequence of irradiation of the thyroid in patients who have received therapeutic doses to the cervical area (30–70 Gy).19 In patients with Hodgkin disease, onehalf to two-thirds of those receiving incidental thyroidal irradiation can be expected to develop elevated TSH levels, with half frankly hypothyroid at presentation.27 Most reports cite a 30% to 40% incidence of hypothyroidism in head and neck cancer patients. Almost twice as many have overt hypothyroidism as have subclinical hypothyroidism (an isolated TSH elevation with normal serum thyroxine and triiodothyronine), a pattern opposite to that seen in Hodgkin patients.28 The prevalence of thyroid dysfunction in patients with Hodgkin disease and other lymphomas appears somewhat increased compared with that in patients with carcinoma of the head and neck treated by radiotherapy alone, despite the fact that the latter group of patients generally receive higher doses of neck irradiation and frequently undergo hemithyroidectomy. Information has been conflicting regarding the role of the iodine load associated with the use of radiographic contrast agents (especially the ethiodized oil used in lymphangiography) in predisposing such patients to radiation injury of the thyroid. More recently, a time-adjusted multivariate analysis of data for patients with Hodgkin disease, which accounted for other potentially important variables, found lymphangiography to be the only variable that significantly influenced the development of hypothyroidism.29 Other cancer populations at risk for primary thyroid dysfunction after cancer therapy are being identified. In children and adolescents treated for acute lymphoblastic leukemia with cranial or craniospinal irradiation, subtle primary hypothyroidism is relatively common, with significantly elevated mean nadir diurnal TSH and mean peak nocturnal TSH levels reported.30 Among patients receiving single-fraction TBI for BMT in childhood, 73% develop overt (15%) or subclinical (58%) hypothyroidism within a mean follow-up period of 3.2 years; fractionating the irradiation results only in transiently elevated TSH levels in 25%.31 Spinal axis irradiation for central nervous system malignancies results in hypothyroidism in 20% to 68% of children.19 Irradiation-induced hypothyroidism shows a dose dependency, with the prevalence and severity of thyroid dysfunction lower in patients receiving 30 Gy. The differences between curves 1 and 2 (p = .0001), curves 2 and 3 (p = .0083), and curves 1 and 3 (p 20% above ideal body weight), the adjusted body weight should be used in the calculation of energy and protein needs: Adjusted body weight = (current weight – ideal body weight [from standard tables or equations] × 0.25) + ideal body weight The estimated dry weight should be used to determine energy and protein needs (see later) in settings of peripheral or central edema due to fluid overload or the

capillary leak syndrome. Energy expenditure varies considerably from day to day in critically ill ICU patients. In light of the complications related to overfeeding outlined earlier, energy provision in the range of 25 to 30 kcal/kg per day is generally safe for most ICU patients and for stable patients without severe malnutrition.23,24 In clinically stable non-ICU patients with severe malnutrition who require nutritional repletion, 35 to 40 kcal/kg per day may be provided, with careful monitoring of serum chemistries, as outlined later. Carbon dioxide overproduction, as evidenced by an indirect calorimetry-derived respiratory quotient of >1.0 (the ratio of carbon dioxide production to oxygen consumption) was not uncommon in the past, when ICU patients routinely received excessive energy doses. This complication is unusual, however, with current standards of nutritional care in ICU settings. Dextrose in PN or tube feeds should be given at a dosage not to exceed 5 mg/kg per minute (~500 g per day for a 70-kg person).23 Catabolic patients are unable to efficiently oxidize larger carbohydrate loads, which may induce hyperglycemia, hepatic steatosis, and/or excessive carbon dioxide production over time. Dextrose should provide 70% to 80% of nonprotein energy, unless the patient is hyperglycemic. In this case, the dextrose load should be reduced and/or regular insulin should be provided in parenteral feeding or as a separate insulin drip to maintain blood glucose between 100 and 150 mg/dL.25 Intravenous lipid emulsions are used to provide essential linoleic and linolenic fatty acids or as an energy source, and are generally infused over a 24-hour period in patients requiring PN. The maximal recommended rate of fat emulsion infusion is ~1.0 g/kg per day.23 ICU patients generally clear intravenous fat emulsions well from plasma; however, large doses of fat emulsion have been associated with impaired reticuloendothelial function and possibly immune suppression. Serum triglycerides should be monitored serially to assess the clearance of the intravenous fat emulsion. The triglyceride levels should be maintained at 2.0 g/kg per day are not efficiently used for protein synthesis, and the excess may be oxidized, contributing to azotemia. In most catabolic patients requiring specialized feeding, a generally recommended protein dose is 1.5 g/kg per day in individuals with normal renal function.9,23 The administered protein dosage should be adjusted downward as a function of the degree of azotemia and hyperbilirubinemia. This strategy takes into account the relative inability of ICU patients to efficiently use exogenous nutrients and the knowledge that most protein and lean tissue repletion occurs over a period of several weeks to months during convalescence. Adequate nonprotein energy is essential to allow amino acids to be effectively used for protein synthesis. The nonprotein calorie/nitrogen ratio used in most centers ranges from 100:1 to 150:1 (nitrogen = protein/6.25). Highly catabolic patients are given protein loads at the lower end of this range, assuming near-normal renal and hepatic function.9,23

TABLE 231-5. Guidelines for Protein Administration in Hospitalized Patients

To diminish the risk of phlebitis, peripheral vein PN solutions provide low concentrations of dextrose (10 cigarettes per day. Cigarette smoking may increase dopamine excretion centrally, reducing serum prolactin levels.4 If injected subcutaneously, nicotine reduces both serum prolactin and luteinizing hormone (LH) levels in animals. Male long-term smokers have increased serum follicle-stimulating hormone concentrations. THYROID GLAND In Sweden, the frequency of goiter is higher among smokers (30%) than among nonsmokers (16%). A higher incidence of goiter in smokers (15%) than in nonsmokers (9%) also was found in a Danish study; in this study, serum thyroglobulin levels were higher and thyrotropin (thyroid-stimulating hormone) levels were lower in long-term smokers, but no difference was found in levels of serum thyroxine (T4), triiodothyronine (T3), and reverse T3.5 Another study, which used plasma cotinine, thiocyanide, and blood carboxyhemoglobin levels to measure the degree of short-term cigarette smoking, found lower serum T3 and T4 levels but the same levels of thyrotropin in long-term smokers. These studies were done 30 to 60 minutes after cigarette smoking, which may reflect both immediate and long-term effects. 6 Women with hypothyroidism who smoke have a greater degree of hypothyroidism and higher serum low-density lipoprotein than in nonsmoking hypothyroid women. These findings suggest an alteration of thyroid function as well as blunted thyroid hormone action.7 Graves ophthalmopathy is more common and severe among tobacco smokers. Interestingly, several constituents of tobacco smoke have known antithyroid properties. Smoking increases levels of plasma thiocyanate, which enters the thyroid by the same active transport system as iodide and competitively inhibits the uptake, activation, and organification of iodide by the follicular cell. Pyridine compounds in tobacco can inhibit the peripheral conversion of T4 to T3, as well as the activation, organification, and coupling pathways (i.e., the peroxidase-dependent reactions) in the follicular cell. These factors may promote goiter formation, a protective mechanism against development of hypothyroidism. Infants of smoking mothers have relative thyroidal hyperfunction at birth.8 Serum calcitonin levels increase acutely after cigarette smoking.9 This appears to emanate from the lung, however, rather than the thyroid gland (see Chap. 177). CALCIUM AND BONE METABOLISM Cigarette smoking is a risk factor for osteoporosis.10 Also, smoking hinders bone healing after injuries or surgery.11 ADRENAL GLANDS Studies of the effect of cigarette smoking on the adrenal glands of humans have had discordant results. In one study, urinary levels of 17- and 11-hydroxycorticoids and catecholamines in smokers did not significantly differ from those in nonsmokers. Conversely, an investigation of the relationship between changes in plasma nicotine, adrenocorticotropic hormone (ACTH), and cortisol levels after smoking by men found significant rises of serum cortisol levels in 73% and of ACTH levels in 45% after a high-nicotine cigarette (2 mg) was smoked, but no changes after a low-nicotine cigarette (0.48 mg) was smoked. The serum ACTH increase occurred if nausea occurred.12 Long-term smokers are particularly prone to the development of the acute hypercortisolemia of smoking.13 In postmenopausal women, long-term smoking is associated with increased plasma levels of the adrenal androgens dehydroepi-androsterone sulfate and androstenedione.14 After a 12-hour abstinence from smoking, men with hypertension who were receiving no medication had a significant, immediate rise in plasma aldosterone, cortisol, ACTH, and catecholamine levels after smoking. A later rise in plasma renin activity occurred, which was associated with a transient, but significant, increase in blood pressure and pulse rate.15 Plasma norepinephrine and epinephrine levels have been determined in smokers before and after adrenergic blockade. Before blockade, they were increased, with an associated increase in pulse rate, blood pressure, and blood lactate/pyruvate ratio. These effects were blocked by the prior administration of phentolamine and propranolol.16 These circulatory changes preceded plasma increases in catecholamine levels and, therefore, probably were mediated by a neurocrine effect of norepinephrine from adrenergic axon terminals. The effect of the release of catecholamines on coronary artery disease or on acute coronary events must be determined in humans; in animals, smoking and acute myocardial ischemia synergistically produce ventricular fibrillation. Furthermore, smoking decreases the level of high-density lipoprotein.

GONADAL AND REPRODUCTIVE FUNCTION Evidence suggests that women who smoke have an increased frequency of menstrual disorders and decreased fertility.17 Women who smoke have earlier menopause.18,18a The percentage of postmenopausal women, standardized according to age, rises with the amount of smoking. For example, of those women aged 44 to 53 years who never smoked, only 35% were postmenopausal, compared with 49% of those who smoked more than one pack per day. Ex-smokers (who stopped at least 1 year earlier) and smokers consuming less than one-half pack per day had intermediate percentages of 36% and 43%, respectively. In another study, current smokers reached menopause an average of 1.7 years earlier than did nonsmokers. Two mechanisms have been proposed for the earlier menopause in smokers: nicotine may directly affect the interplay of hormones involved in the menopausal process, or the cigarette smoke may induce secretion of hepatic enzymes that affect the metabolism of estrogen.19 Partly associated with the earlier menopause and lower body weight seen in female smokers, postmenopausal osteoporosis shows a higher incidence in smokers. Male smokers have a higher incidence of osteoporosis as well.20 The Apgar scores of the offspring of women who smoke are lower at 1 and 5 minutes after birth and are inversely proportional to the amount of cigarettes smoked by the mother during pregnancy. Furthermore, the hemoglobin and hematocrit of the newborn are increased in proportion to the extent of maternal smoking. Maternal tobacco smoking disturbs the endocrine states of the human fetus, resulting in increased serum levels of prolactin, growth hormone, and insulin-like growth factor-I.21 In one study of almost 7000 women, smokers had a stillbirth rate three times as high as that of nonsmokers. The neonatal death rate was equal in smokers and nonsmokers.22 In most studies, little correlation is seen between maternal smoking and congenital malformations. Mothers who smoke produce smaller infants.23 Although all investigations are not in accord, one study found that serum levels of human placental lactogen were significantly lower in pregnant smokers than in nonsmokers. This relationship persisted throughout pregnancy.24 A direct correlation is seen between the level of human placental lactogen and birth weight. Maternal serum chorionic gonadotropin and estradiol levels also are decreased in smokers.25 Estriol excretion, an index of intrauterine fetal growth, does not differ in smoking and nonsmoking pregnant women. More data are required to determine the extent to which serum hormone levels during pregnancy may be correlated with smoking and lower-birth-weight infants. Reduced placental blood flow may be responsible for the decreased serum human placental lactogen and chorionic gonadotropin levels, as well as the low birth weight. In sheep, however, short-term smoking or nicotine administration has little effect on uterine blood flow. Nevertheless, babies born to mothers at higher altitudes, presumably with less available oxygen, have lower birth weights than do the babies of mothers at sea level. Some investigators, however, believe the low birth weights of these infants result from low maternal weight gain during pregnancy rather than from hormonal mediation. In studies of infertile men, semen analyses have shown a significantly greater percentage of abnormal spermatozoa in the tobacco smokers.26,27 Cigarette smoking is associated with lowered sperm density.28 In one study, men who were smokers had significantly higher serum follicle-stimulating hormone levels and lower cortisol and testosterone levels, along with an increased percentage of abnormal spermatozoa. In dogs, long-term cigarette smoking increases the hepatic metabolism of testosterone.29 Current cigarette smokers have higher endogenous serum estrogen levels than do nonsmokers.30 Erectile dysfunction may occur.31 DIABETES MELLITUS Generally, persons with diabetes smoke fewer cigarettes than do persons without diabetes. Smoking is less common in older patients with diabetes than in younger ones. The largest group of ex-smokers among patients with diabetes are middle-aged females. Nevertheless, heavy users of cigarettes have an increased risk of type 2 diabetes.31a The suggestion has been made that chronic hypoxia caused by carbon monoxide in smoke and an increased tendency for platelet aggregation might play a role in diabetic microangiopathy. A study was made of 181 patients with diabetic retinopathy to determine whether smoking influenced the incidence or severity of retinopathy.32 Although the degree of control of diabetes was not ascertained, the incidence of retinopathy rose in smokers with increasing duration of diabetes, and the incidence of proliferative retinopathy rose with increase in tobacco consumption. No such association between proliferative retinopathy and duration of diabetes was found among nonsmokers. Other researchers have reported a relationship between diabetic retinopathy in men and smoking.33 A large study of Oklahoma Native Americans with type 2 diabetes, however, found no relationship between smoking and retinopathy, and another study of a different ethnic group also found no association.34,35 In patients with type 1 diabetes, the frequency of diabetic nephropathy is significantly higher among cigarette smokers, and the condition progresses more rapidly.36 Furthermore, a correlation between smoking and both nephropathy and retinopathy in men, but not in women, has been reported. A consensus exists that smoking exacerbates the coronary and peripheral vascular disease of persons with and without diabetes. The influence of smoking on diabetic neuropathy has been insufficiently studied, although many reports of smoking-related neuropathic syndromes have appeared. Cigarette smoking raises blood levels of endothelin-1. This hormone may play an important role in the development of atherosclerosis in smokers with or without diabetes.37 In patients with diabetes, cigarette smoking may affect insulin requirements. In one study of long-term smokers, subcutaneous insulin absorption was not influenced by the short-term effect of cigarettes. On the other hand, smokers with diabetes have been reported to have 15% to 20% higher insulin requirements than nonsmokers with diabetes, as well as 30% higher serum triglyceride concentrations. In another study, however, patients with type 1 diabetes had no change in insulin sensitivity.38 In smokers without diabetes, chain smoking increased the 40- and 60-minute serum glucose levels and decreased the 120-minute level of the oral glucose-tolerance test. Compared with nonsmokers without diabetes and with ex-smokers, chain smokers without diabetes had a lower glucose disappearance constant in the intravenous glucose-tolerance test.39 Thus, evidence suggests that cigarette smoking in patients with diabetes increases the incidence of retinopathy and nephropathy. Moreover, the possibility of increased insulin requirements and the higher incidence of coronary and peripheral vascular disease in smokers are ample reasons to insist that patients with diabetes not smoke and to provide programs and encouragement to help them to quit.40 BODY WEIGHT AND CALORIC REGULATION Many investigations have found that smokers tend to weigh less than nonsmokers and that weight gain occurs after cessation of smoking.41 Oral satisfaction and keeping one's hands occupied are two often-quoted explanations. Cigarette smoking has been linked to a decreased sense of taste, which may be relieved by cessation of smoking. Moreover, one postulation is that some ex-smokers who gain weight have lost an acquired signal to terminate eating, which had been provided by the cigarette smoked after the meal. Smokers eat less sugar and fewer fatty foods than do nonsmokers. Evidence is found, however, that smokers actually consume more calories than their heavier counterparts but may be more “energy efficient,” storing energy in protein rather than fat. Pregnant women who smoke consume more food than do control patients who do not smoke, but gain less weight during pregnancy. Nicotine may modulate gastrointestinal peptides, such as cholecystokinin, involved in regulation of feeding.42 An unexplored subject is the effect, if any, of “passive” smoking (environmental cigarette smoke) on the endocrine system.43,44

ENDOCRINE EFFECTS OF MARIJUANA Cannabis (hashish, marijuana) is obtained from the flowering tops of hemp plants. It is a psychoactive substance that usually is smoked. Its principal psychoactive ingredient is L -D-9-tetrahydro-cannabinol (THC). Despite its illegal status in the United States, it has become a widely used social drug. Its long-term use alters the activities of the endogenous cerebral opiate system and the cate-cholaminergic and histamine systems. Many of the endocrine and metabolic effects of marijuana are mediated by hypothalamic responses. In long-term users of this drug, a tolerance to some effects develops. Many of the changes appear to be reversible. Marijuana affects gonadal function. Sporadic cases of gynecomastia have occurred, that are probably related to this substance.45 Men who otherwise were healthy but

who smoked marijuana on a long-term basis were reported to have lower plasma testosterone levels and sperm counts than did nonsmokers. Some studies in humans have shown a depression of plasma testosterone levels after marijuana smoking, which sometimes was associated with lower levels of plasma LH.46,47 Studies using male rats, mice, monkeys, pigeons, and toads have shown that cannabis extract (THC), usually given parenterally or inhaled, reduces the weight of the male reproductive glands (i.e., seminal vesicles, prostate, epididymis, and seminiferous tubules) and decreases sperm production.48 Plasma levels of growth hormone are decreased after THC treatment. Some of these effects are mediated by the hypothalamic-pituitary axis. In the female animal, marijuana inhibits the secretion of LH and follicle-stimulating hormone, and decreases plasma levels of prolactin. In rats, cannabis resin depresses radioactive iodine uptake in the thyroid gland. Body temperature was decreased in another study. In animals, levels of adrenal cortical hormones increase after short-term cannabinoid administration; this effect is thought to be mediated through the hypothalamic-pituitary axis.48 In men, marijuana smoking caused an acute increase of serum cortisol levels.49 Another study of the acute effects of marijuana smoking has shown no difference in serum cortisol levels after ACTH administration between normal men and those who smoked marijuana on a long-term basis.50 Plasma catecholamine levels appear to decrease with administration of THC.51 THC crosses the placenta. Although teratogenic effects in humans have not been confirmed, growth retardation and delayed development have been observed in infants of mothers who smoked marijuana. 52,53

ENDOCRINE EFFECTS OF COCAINE Cocaine is an alkaloid obtained from the leaves of coca plants (Erythroxyloncoca) an evergreen shrub that grows at relatively high altitudes. Cocaine exerts many endocrine effects. Some of the data appear contradictory because of the paucity of human studies and the differences between short- and long-term administration. Magnetic resonance imaging studies of men who abuse opiates and cocaine reveal an increased pituitary volume.54 More than one-third of cocaine-dependent patients have hyperprolactinemia.55 In cocaine-dependent men, intravenous cocaine acutely increases serum ACTH levels. This is mediated via corticotropin-releasing hormone (CRH).56 Human cocaine addicts, however, fail to exhibit the normal increase in prolactin, ACTH, b-endorphin, or cortisol after a hyperthermic stress.57 In female monkeys in the early follicular stage, cocaine acutely increases LH levels and also enhances gonadotropin-releasing hormone–induced stimulation of LH secretion. It disrupts menstrual and ovarian cyclicity.58,58a Clinically, humans who use cocaine extensively do not appear to have altered thyroid function. However, there is a blunted response of thyroid stimulating hormone to thyroid releasing hormone administration.59 In laboratory animals, cocaine induces a central stimulation of the sympathoadrenal neural axis and may play a role in the pressor and tachycardic effects of the drug.60 Gonadal and reproductive function may be affected by cocaine. In rats, cocaine significantly reduces ovulation rates.61 Long-term administration of cocaine to rats causes testicular lesions; this may be due to ischemia secondary to local vasocon-striction.62 Human cocaine addicts have decreased sperm concentrations. In addition, human sperm may act as a vector to transport cocaine into the ovum, resulting in the abnormal development that occasionally occurs in the offspring of cocaine-exposed men.63 Cocaine passes the placenta. Cocaine usage during pregnancy impairs transplacental amino-acid transport.64 Maternal cocaine use may affect the endocrine homeostasis of the fetus and newborn.65 In humans, the cortisol response of the newborn to stress is blunted. The short-term administration of cocaine to pregnant ewes increases both maternal and fetal glucose and lactate levels.66 Patients with diabetes who are undergoing dialysis and use cocaine have an increased rate and greater severity of infections (e.g., cellulitis, abscess, and sepsis).67 In addition to cocaine and marijuana, other street drugs (e.g., heroin, amphetamines, hallucinogens, volatile substances, phencyclidine [PCP]) merit ongoing study concerning possible endocrine and metabolic effects. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

Day MD. Autonomic pharmacology. Experimental and clinical aspects. New York: Churchill Livingstone, 1979. Seckl JR, Johnson M, Shakespear C, Lightman SL. Endogenous opioids inhibit oxytocin release during nicotine-stimulated secretion of vasopressin in man. Clin Endocrinol (Oxf) 1988; 28:509. Waeber B, Schaller M, Nussberg J, et al. Skin blood flow reduction induced by cigarette smoking: role of vasopressin. Am J Physiol 1984; 247:H895. Andersen AN, Semczuk M, Tabor A. Prolactin and pituitary-gonadal function in cigarette smoking infertile patients. Andrologia 1984; 16:391. Christensen SB, Ericsson UB, Janzon L, et al. Influence of cigarette smoking on goiter formation, thyroglobin, and thyroid hormone levels in women. J Clin Endocrinol Metab 1984; 58:615. Sepkovic DW, Haley NJ, Wynder EL. Thyroid activity in cigarette smokers. Arch Intern Med 1984; 144:501. Müller B, Zulewski H, Huber P, et al. Impaired action of thyroid hormone associated with smoking in women with hypothyroidism. N Engl J Med 1995; 333:964. Meberg A, Marstein S. Smoking during pregnancy—effects on the fetal thyroid function. Acta Paediatr Scand 1986; 75:762. Tabassian AR, Nylén ES, Giran AE, et al. Evidence for cigarette smoke-induced calcitonin secretion from lungs of man and hamster. Life Sci 1988; 42:2323. Thomas TN. Lifestyle risk factors for osteoporosis. Medsurg Nurs 1997; 6:275. Haverstock BD, Mandracchia VJ. Cigarette smoking and bone healing: implications in foot and ankle surgery. J Foot Ankle Surg 1998; 37:69. Cherek DR, Smith JE, Lanc JD, Branchi JT. Effect of cigarettes on saliva cortisol levels. Clin Pharmacol Ther 1982; 32:765. Gossain VV, Sherma NK, Srivastava L, et al. Hormonal effects of smoking. II. Effects on plasma cortisol, growth hormone, and prolactin. Am J Med Sci 1986; 291:325. Khaw K-T, Tazuke S, Barrett-Connor E. Cigarette smoking and levels of adrenal androgens in postmenopausal women. N Engl J Med 1988; 318:1705. Baer L, Radichevich I. Cigarette smoking in hypertensive patients: blood pressure and endocrine responses. Am J Med 1985; 78:564. Cryer PE, Haymond MW, Santiago JV, Shah SD. Norepinephrine and epinephrine release and adrenergic mediation of smoking associated hemodynamic and metabolic events. N Engl J Med 1976; 295:753. 17. Weisberg E. Smoking and reproductive health. Clin Reprod Fertil 1985; 3:175. 18. McKinlay SM, Bifano NL, McKinlay JB. Smoking and age at menopause in women. Ann Intern Med 1985; 103:350. 18a.Cooper GS, Sandler DP, Bohlig M. Active and passive smoking and the occurrence of natural menopause. Epidemiology 1999; 10:771. 19. Michnovicz JJ, Hershcopf RJ, Naganuma H, et al. Increased 2-hydroxylation of estradiol as a possible mechanism for the anti-estrogenic effect of cigarette smoking. N Engl J Med 1986; 315:1305. 20. Seeman E, Melton LJ III, O'Fallon WM, Riggs BL. Risk factors for spinal osteoporosis in men. Am J Med 1983; 75:977. 21. Beratis NG, Varvarigou A, Makri M, Vagenakis AG. Prolactin, growth hormone and insulin-like growth factor-I in newborn children of smoking mothers. Clin Endocrinol (Oxf) 1994; 40:179. 22. Schwartz D, Goujard J, Kaminski A, et al. Smoking and pregnancy. Results of a prospective study of 6989 women. Rev Eur Etud Clin Biol 1972; 18:867. 23. Kramer MS. Socioeconomic determinants of intrauterine growth retardation. Eur J Clin Nutr 1998; 52(Suppl 1):S29. 24. Boyce A, Schwartz D, Hubert C, et al. Smoking, human placental lactogen and birth weight. Br J Obstet Gynaecol 1975; 82:964. 25. Bernstein L, Pike MC, Lobo RA, et al. Cigarette smoking in pregnancy results in marked decrease in maternal hCG and oestradiol levels. Br J Obstet Gynaecol 1989; 96:92. 26. Evans HJ, Fletcher J, Torrance M, Hargreave TB. Sperm abnormalities and cigarette smoking. Lancet 1981; 1:627. 27. Shaarawy M, Mahmoud KZ. Endocrine profile and semen characteristics in male smokers. Fertil Steril 1982; 38:255. 28. Vine MF, Margolin BH, Morrison HI, Huka BS. Cigarette smoking and sperm density: a meta-analysis. Fertil Steril 1994; 61:35. 29. Mittler JC, Pogach L, Ertel NH. Effects of chronic smoking on testosterone metabolism in dogs. J Steroid Biochem 1983; 18:759. 30. Barrett-Connor E, Khaw KT. Cigarette smoking and increased endogenous estrogen levels in men. Am J Epidemiol 1987; 126:187. 31. Jeremy JY, Mikhailidis DP. Cigarette smoking and erectile dysfunction. J R Soc Health 1998; 118:151. 31a.Perrson PG, Carlsson S, Svanstrom L, et al. Cigarette smoking, oral moist snuff use and glucose intolerance. J Intern Med 2000; 248:103. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46.

Paetkau ME. Diabetic retinopathy and smoking. Lancet 1978; 2:1098. Walker JM, Cove DH, Beevers DG, et al. Cigarette smoking, blood pressure and the control of blood glucose in the development of diabetic retinopathy. Diabetes Res 1985; 2:183. West KM, Stober JA. Smoking and diabetic retinopathy. Lancet 1978; 2:49. Klein R, Klein BE, Davis MD. Is cigarette smoking associated with diabetic retinopathy? Am J Epidemiol 1983; 118:228. Sauiski PT, Didurget U, Mühlauser I, et al. Smoking is associated with progression of diabetic nephropathy. Diabetes Care 1994; 17:126. Haak T, Dungmann E, Raab C, Usadel KH. Preliminary report: elevated endothelin-1 levels after cigarette smoking. Metabolism 1994; 43:267. Helve E, Yki-Jarvinen H, Koivisto VA. Smoking and insulin sensitivity in type I diabetic patients. Metabolism 1986; 35:874. Janzon L, Berntorp K, Hanson M, et al. Glucose tolerance and smoking: a population study of oral and intravenous glucose tolerance tests in middle-aged men. Diabetologia 1983; 25:86. Willett WC, Green A, Stampfer MF, et al. Relative and absolute excess risks of coronary heart disease among women who smoke cigarettes. N Engl J Med 1987; 317:1303. Wack JT, Rodin J. Smoking and its effects on body weight and the systems of caloric regulation. Am J Clin Nutr 1982; 35:366. Chowdhury P, Inoue K, Rayford PL. Effect of nicotine on basal and bombesin stimulated canine plasma levels of gastrin, cholecystokinin and pancreatic polypeptide. Peptides 1985; 6:127. Byrd JC, Shapiro RS, Schiedermayer DL. Passive smoking: a review of medical and legal issues. Am J Public Health 1989; 79:209. Cigarette smoking. Focus on the workplace. N Y State J Med 1989; 89:1. Cates W Jr, Pope JN. Gynecomastia and cannabis smoking. A nonassociation among US Army soldiers. Am J Surg 1977; 134:613. Barnett G, Chiang CW, Licko V. Effects of marijuana on testosterone in male subjects. J Theor Biol 1983; 104:685.

47. Kolodny RC, Lessin RC, Toro G, et al. Depression of plasma testosterone with acute marijuana administration. In: Brande MC, Szara J, eds. The pharmacology of marijuana. New York: Raven Press, 1976:217. 48. Mendelson JH, Mello NJ. Effects of marijuana on neuroendocrine hormones in human males and females. Natl Inst Drug Abuse Res Monogr Ser 1984; 44:97. 49. Cone EJ, Johnson RE, Moore JD, Roache JD. Acute effects of smoking marijuana on hormones, subjective effects and performance in male human subjects. Pharmacol Biochem Behav 1986; 24:1749. 50. Perez-Reyes M, Brine D, Wall ME. Clinical study of frequent marijuana use: adrenal cortisol reserve, metabolism of a contraceptive agent and development of tolerance. In: Dornbush RL, Freedman AM, Fink M, eds. Chronic cannabis use. New York: New York Academy of Sciences, 1976:173. 51. Harclerode J. Endocrine effects of marijuana in the male: preclinical studies. Natl Inst Drug Abuse Res Monogr Ser 1984; 44:46. 52. Smith CG, Asch RH. Acute, short-term, and chronic effects of marijuana on the female primate reproductive function. Natl Inst Drug Abuse Res Monogr Ser 1984; 44:82. 53. Hatch EE, Bracken MB. Effect of marijuana use in pregnancy on fetal growth. Am J Epidemiol 1986; 124:986. 54. Teoh SK, Mendelson JH, Woods BT, et al. Pituitary volume in men with concurrent heroin and cocaine dependence. J Clin Endocrinol Metab 1993; 76:1529. 55. Kranzler HR, Wallington DJ. Serum prolactin level, craving, and early discharge from treatment in cocaine-dependent patients. Am J Drug Alcohol Abuse 1992; 18:187. 56. Sarnyai Z. Neurobiology of stress and cocaine addiction. Studies on corticotropin-releasing factor in rats, monkeys, and humans. Ann N Y Acad Sci 1998; 851:371. 57. Vescovi PP, Coiro V, Volpi R, et al. Hyperthermia in sauna is unable to increase the plasma levels of ACTH/cortisol, beta-endorphin and prolactin in cocaine addicts. J Endocrinol Invest 1992; 15:671. 58. Mello NK, Mendelson JH, Kelly M, Bowen CA. The effects of cocaine on basal and human chorionic gonadotropin-stimulated ovarian steroid hormones in female rhesus monkeys. J Pharmacol Exp Ther 2000; 294:1137. 59. Vescovi PP, Pezzarossa A. Thyrotropin-releasing hormone-induced GH release after cocaine withdrawal in cocaine addicts. Neuropeptides 1999; 33:522. 60. Tella SR, Schindler CW, Goldberg SR. Cocaine: cardiovascular effects in relation to inhibition of peripheral neuronal monoamine uptake and central stimulation of the sympathoadrenal system. J Pharmacol Exp Ther 1993; 267:153. 61. Barroso-Moguel R, Mendez-Armenta M, Villeda-Hernandez J. Testicular lesions by chronic administration of cocaine in rats. J Appl Toxicol 1994; 14:37. 62. Li H, George VK, Crawford SC, Dhabuwala CB. Effect of cocaine on testicular blood flow in rats: evaluation by percutaneous injection of xenon-133. J Environ Pathol Toxicol Oncol 1999; 18:73. 63. Yazigi RA, Odem RR, Polakoski KL. Demonstration of specific binding of cocaine to human spermatozoa. JAMA 1991; 266:1956. 64. Pastrakuljic A, Derewlany LO, Koren G. Maternal cocaine use and cigarette smoking in pregnancy in relation to amino acid transport and fetal growth. Placenta 1999; 20:499. 65. Chiriboga CA. Neurological correlates of fetal cocaine exposure. Ann N Y Acad Sci 1998; 846:109. 66. Owiny JR, Sadowsky D, Jones MT, et al. Effect of maternal cocaine administration on maternal and fetal glucose, lactate, and insulin in sheep. Obstet Gynecol 1991; 77:901. 67. D'Elia JA, Weinrauch LA, Paine DF, et al. Increased infection rate in diabetic dialysis patients exposed to cocaine. Am J Kidney Dis 1991; 18:349.

CHAPTER 235 ENVIRONMENTAL FACTORS AND TOXINS AND ENDOCRINE FUNCTION Principles and Practice of Endocrinology and Metabolism

CHAPTER 235 ENVIRONMENTAL FACTORS AND TOXINS AND ENDOCRINE FUNCTION LAURA S. WELCH Hypothalamus Pituitary Gland Thyroid Gland Calcium and Bone Metabolism Adrenal Glands Sexual and Endocrine Function in Women Sexual and Endocrine Function in Men Fuel Metabolism Diabetes Mellitus Hypoglycemia Lipids Occupational Factors Chapter References

In the 1960s, the populations of various bird species in the United States began to decline. Thinning eggshells and decreased survival of the young were caused by the effects of the pesticide dichlorodiphenyltrichloroethane (DDT) on the endocrine and other body systems of the birds. This important discovery sparked increasing research on the effects of environmental toxins on both animal and human health. Now, years later, although these observations have caused public and governmental concern, knowledge of the effects of environmental and industrial chemicals on the endocrine system of humans still is incomplete. A ban on, or restriction on use of, many persistent organic pollutants has resulted in a definite decrease in the body burden of these chemicals in humans and other animals. What level is “safe” and will not result in an endocrine-like effect is an area of active debate.1,1a Research is providing important information, but many of the effects of environmental substances on people are uncovered because astute clinicians remember to take occupational histories and to look for the causes of clusters of specific diseases. Only by the routine incorporation of occupational and environmental histories into clinical practice can physicians identify causal relationships.2 This chapter reviews the known and probable effects of environmental and occupational exposures on the endocrine system. Endocrine responses to physical or chemical agents and the role of the endocrine system in modulating toxic effects are described. Data from animal experiments are presented to elucidate the mechanism of human toxicity.3,4 The importance of taking occupation into account in the treatment of endocrine disorders is discussed. The environmental agents alcohol, smoking, and drugs are discussed in Chapter 233, Chapter 234 and Chapter 239, respectively.

HYPOTHALAMUS Beryllium, a metal widely used in specialty alloys (Table 235-1), causes a syndrome with many of the manifestations of sarcoidosis. Berylliosis can present as an acute or subacute disease, which usually includes chest disease as a prominent manifestation. The differentiation of berylliosis from sarcoidosis depends on a history of beryllium exposure and either the documentation of a body burden of beryllium or the demonstration of activation of lymphocytes, obtained from blood or bronchial lavage, on exposure of the cells to the metal. Berylliosis may affect the hypothalamus in the same way as sarcoidosis (see Chap. 9): the granulomatous involvement of the hypothalamus can cause posterior or anterior pituitary dysfunction.

TABLE 235-1. Toxins Associated with Endocrine Dysfunction

PITUITARY GLAND Lead exposure in a battery plant has caused decreased levels of serum thyroid-stimulating hormone (TSH), thus reducing serum thyroxine (T4) concentrations. Response to thyrotropin-releasing hormone stimulation was flat or delayed, suggesting anterior pituitary involvement.5,6 Moreover, subtle effects on the pituitary have been found in monkeys exposed to lead.7 Removal of the patients from the lead exposure was followed by improvement after months to years, but the natural history of this dysfunction still is not clear. The patients studied had non-specific symptoms, such as fatigue, headache, muscle pains, and impotence, that may have resulted from the known central nervous system effects of lead. Styrene causes a rise in serum prolactin and growth hormone levels in exposed workers; levels of prolactin increase as the exposure increases.8 Styrene is a solvent used extensively in the manufacture of reinforced plastics, such as boat hulls or molded chairs. In humans, styrene and similar solvents cause dose-related effects on visual-motor, visual-perceptive, and memory functions. In rabbits, styrene depletes striatal and tuberoin-fundibular levels of catecholamines, and a similar phenomenon probably occurs in humans. Prolactin is the hormone most likely to be affected by such involvement because it is regulated directly by dopaminergic feedback. The effects of toxins on sexual function are discussed later; some of these effects may be due to effects on the pituitary as well.

THYROID GLAND Many chemicals affect human and animal thyroid function.8a Table 235-1 lists these chemicals and the jobs and industries in which workers may be exposed. Generally, most of the chemicals cause goiter or hypothyroidism. For some of the compounds, the mechanism of action is well understood. In addition, many naturally occurring substances act as goiter-producing agents, and sporadic outbreaks of endemic goiter have been traced to them.9 These compounds include thiouracil, thiocyanates, and thiourea. Plants in the cabbage family cause goiter in animals and have caused an outbreak in people drinking milk from cows feeding on these plants. The active ingredient in these plants is a thiocyanate that inhibits organic iodination in the thyroid (see Chap. 30 and Chap. 38). Many compounds affecting the thyroid are used in industry. Ethylene thiourea is used as an accelerator in the manufacture of rubber and as a chemical intermediate in the manufacture of pesticides, fungicides, dyes, and chemicals. It is in the same chemical class as propylthiouracil and has been used previously in the treatment of thyrotoxicosis. One study showed that men working as rubber mixers with significant exposure to ethylene thiourea dust had serum T4 levels that were significantly lower than those of a control group, although all levels were within the normal range. One of the men had an increased serum TSH level.10 Halogenated aromatic hydrocarbon compounds affect thyroid function in humans and animals. Table 235-1 describes where these compounds are found in the environment. Figure 235-1 displays some of the most common chemicals in this group. Polybrominated biphenyls (PBBs) caused primary hypothyroidism in 11% of men working in a PBB production facility. The serum free T4 level was decreased, and the serum TSH level was elevated. Thyroid antimicrosomal antibodies were

present.11 Although PBBs are no longer produced, they still are found in the environment. Polychlorinated biphenyls (PCBs) appear to mimic T4 structurally.12,13,13a

FIGURE 235-1. The chemical structures of several halogenated aromatic hydrocarbon compounds. Note the structural similarities among the group and the similarity to estradiol and diethylstilbestrol (see Fig. 235-2).

FIGURE 235-2. Chemical structures of equol, the phytoestrogen formed in the gastrointestinal tract of humans and animals; estradiol; the potent synthetic estrogen, diethylstilbestrol; and several phytoestrogens of plant origin. (From Sectell KDR, Doriello SD, Hulme P, et al. Nonsteroidal estrogens of dietary origins: possible roles in hormone dependent disease. Am J Clin Nutr 1984; 40:569.)

The pesticide hexachlorobenzene (HCB) causes thyroid enlargement in humans. In the 1950s, an epidemic of HCB poisoning occurred in Turkey; 4000 persons were affected, and 400 died. Seed grain had been treated with HCB and, during a period of famine, local farmers used the seed for baking bread. A subset of these persons were studied again between 1977 and 1981. The major effect was the development of a syndrome of porphyria cutanea tarda and joint disease. Moreover, 40% had thyroid enlargement, which affected 60% of the women and 27% of the men in the subset. Two of these persons underwent thyroidectomy, and no malignant changes were noted.14 Experimentally, feeding HCB to rats has been found to decrease T4 levels.15 This effect results from interference with plasma thyroid transport protein by the hydroxylated metabolites of HCB and from increased metabolism of T4 induced by HCB.16 Several polychlorinated compounds, such as 2,3,7,8-tetra-chlorodibenzo-p-dioxin (TCDD, or dioxin), PCBs, and PBBs, have similar biologic effects. This observation has led to the discovery of a specific receptor, called the Ah receptor, for these compounds.17,18 The Ah receptor is similar in structure to steroid receptors, but the endogenous ligand remains unknown. Future work should determine the role of this receptor in the endocrine toxicity of these compounds. The administration of TCDD and PCBs to rats decreases serum T4 concentrations.18a Histologically, hyperplasia and hypertrophy of thyroid follicular cells are seen. These chemicals seem to impair the release of thyroid hormone. Studies of the effect of organochlorine compounds on birds demonstrate definite effects on thyroid function.19 At low levels of exposure, DDT caused hyperthyroidism in pigeons; at higher levels, it caused hypothyroidism. Herring gulls from polluted areas of the Great Lakes had goiters and decreased serum triiodothyronine (T3) and T4 levels. Examination of the thyroid glands demonstrated a decreased colloid content, epithelial hyperplasia, and decreased follicular size. In addition, several bird species had abnormal incubating behavior, which was linked to abnormalities of both sex and thyroid hormones. No studies of the effects of DDT and TCDD on human thyroid function are available, but a reasonable suspicion is that exposure to a chlorinated aromatic compound may contribute to thyroid disease. An occupational history should concentrate on intense or prolonged exposure to any substances in this category. A series of pesticides were administered to ewes, and chlorpyrifos, lindane, 2,4-D, and pentachlorophenol produced a marked decrease in T4 levels.20 Physical agents also cause thyroid disease. Radiation causes thyroid cancer, both from gamma irradiation and from thyroid uptake of the radioactive iodine released in nuclear tests (see Chap. 40). The inhabitants of the Marshall Islands were exposed to radioactive fallout from test detonation of an atomic bomb in 1954. Subsequent studies showed a high incidence of thyroid neoplasms and hypothyroidism.21,22 A risk may exist from lower levels of radioactive iodine exposure from above-ground testing and nuclear accidents, from radioiodine, and from occupational, therapeutic, or diagnostic exposure to radiation.23,24

CALCIUM AND BONE METABOLISM Beryllium and lead interact with calcium metabolism. Other metals, such as radium, aluminum, and fluoride, are stored in bone and can cause adverse health effects. Berylliosis can mimic sarcoidosis and can be associated with hypercalcemia. Removal from beryllium exposure is an essential aspect of the treatment of this disease.25 Lead is widespread in the environment. It is used in gasoline additives, in metal alloys, and in paints and pigments (see Table 235-1). Lead accumulates in the body, and significant stores are present in bone. Although the mechanism is poorly understood, lead is handled in the body like the calcium ion, and some endocrine disorders that affect calcium balance may affect lead balance in lead-exposed patients. In addition, lead may be directly toxic to osteoblasts.26 Hyperparathyroidism, immobilization, and treatment with corticosteroids may cause an elevation in the blood lead level and may precipitate symptoms of lead toxicity. Men working in lead industries develop an acute mononeuropathy from bed rest, perhaps from bone resorption.27 Table 235-1 lists jobs with potential lead exposure. Fluoride is stored in bone, and significant doses cause osteo-sclerosis as part of the syndrome of fluorosis. Table 235-1 lists potential sources of exposure in the occupational setting (see Chap. 131). Ingested radium is deposited in bone. The use of radium salts in the 1920s and 1930s in luminescent paint for clocks and watches caused jaw necrosis, osteosarcoma, and aplastic anemia in the watch painters, who pointed their brushes by touching them to the tips of their tongues.28 The use of radium for fluorescence ceased in the 1960s. Aluminum also is deposited in bone if ingested and predominantly in the lungs if inhaled. Aluminum may cause a vitamin D–resistant osteomalacia (see Chap. 61 and Chap. 131).

ADRENAL GLANDS Adrenocortical suppression has occurred in men working in the manufacture of potent synthetic glucocorticoids, even though they adhered to recommended safety precautions in the workplace. In one man, adrenal insufficiency resulted after the exposure suddenly stopped.29 Paraquat is a commonly used herbicide that has caused death after accidental or intentional ingestion. It causes pulmonary toxicity. A dozen fatal cases of diffuse

bilateral adrenal necrosis have been reported in patients with paraquat poisoning; this may occur in only the most severe cases of poisoning.30 These cases point to the importance of considering corticosteroid replacement therapy in patients with paraquat poisoning. Many toxic substances are not harmful until metabolized. This metabolism occurs in the microsomal P450 enzyme system of the adrenals; these enzymes catalyze hydroxylation reactions for the production of adrenal glucocorticoids and mineralocorticoids. The system also acts rapidly on foreign compounds. For example, dichlorodiphenyldichloroethane (DDD) is an insecticide that induces a selective necrosis of the zonae fasciculata and reticularis of the adrenal cortex. Adrenal metabolism of DDD causes this toxicity. This discovery led to the development of the active isomer o,p'-DDD (mitotane), now used in the treatment of inoperable adrenal carcinoma. The insecticide is no longer used in the United States. One review suggests that the adrenal cortex may be more vulnerable to chemically induced injury than are other endocrine tissues.31 Local activation probably is involved in the adrenocortical necrosis seen with carbon tetrachloride. This agent is metabolized by microsomal monooxygenases in the liver to a highly reactive trichloromethyl radical.32

SEXUAL AND ENDOCRINE FUNCTION IN WOMEN Toxic exposures can cause two types of female sexual dysfunction. Exposure may interfere with endocrine functioning and, if the exposure occurs during pregnancy, the toxins may affect fetal development. Much of the study of the effects of environmental exposures on female reproduction has focused on abnormal fetal growth and development or fetal loss. The direct toxic effects on the fetus are not discussed in this chapter, except for those affecting the developing fetal endocrine system.33,34 Exposures to toxins can affect the development of the fetal endocrine system, including the development of reproductive organs and their subsequent functioning. The administration of estrogens has a feminizing effect on male embryos and causes abnormal genitourinary development. Exposure of animals to polyaromatic hydrocarbons in utero causes the loss of ovarian follicles in the newborn; no studies of this effect in humans have been performed.34 Many environmental compounds (e.g., chlorinated aromatic hydrocarbons) have estrogen-like activity (see Table 235-1). For example, TCDD (dioxin) exposure has been reported to reduce serum levels of follicle-stimulating hormone, luteinizing hormone, and testosterone in men.35 Assessment of the effect of these compounds is complicated, for a specific compound may be an estrogen agonist in one tissue and an antagonist in another. Exposure to these compounds in the environment usually is to a mixture of endocrine-active compounds, so predicting the effect is difficult.36 Estrogenic activity of many of these compounds is known to impair reproductive function in adults of either sex. For example, occupational exposure to estrogens in pharmaceutical manufacture caused irregular menses and infertility.37 High-level exposures have been shown to affect wildlife populations. At this time, little direct evidence is found that exposures to ambient levels of estrogenic compounds affects reproductive health in humans. Some studies suggest that exposure to estrogenic compounds increases the risk of breast cancer, 38,39 but others do not.40,41 and 42 Milk production is mediated by hormones. Although toxic exposures have not been shown to interfere with the endocrine aspects of breast-feeding, significant concentrations of toxic chemicals may be excreted in breast milk.43,44 DDT, methoxychlor, and other DDT derivatives cause an increase in uterine weight and endometrial hyperplasia. DDT binds to estrogen receptors and induces estrogen-sensitive protein synthesis. Low-level exposure to TCDD caused a high incidence of endometriosis in rhesus monkeys45; other studies are investigating this effect in humans. Moreover, the injection of DDT into gull eggs caused feminization of the male embryos, similar to the effect of estradiol administration. The cyclodiene insecticides have similar estrogenic activity.44 Plants, especially legumes such as soya and clover, contain phytoestrogens. Structurally, these compounds are phenols rather than steroids; they have both weak estrogenic and antiestrogenic activity (Fig. 235-2). Resveratrol, a phytoestrogen in grapes, which is present in red wine, is a mixed agonist/antagonist for estrogen receptors alpha and beta.45a When herbivorous animals eat large amounts of plants containing these compounds, they may have prolonged periods of estrus. Diets high in clover have caused reproductive disorders, including infertility, in other animals.46,47 and 48 Daidzein, a phytoestrogen precursor in soya, is metabolized to equol, a compound with weak estrogenic activity. Feed containing soya has caused uterotropic effects in rats, and a high-soya diet in men caused a 1000-fold increase in the urinary excretion of equol.49,50 Whether these phytoestrogens can cause similar effects in humans and animals is unknown. Potentially, high-soya diets could cause reproductive abnormalities in men and women, and the amount of phytoestrogens in the diet could affect rates of breast cancer or other estrogen-responsive diseases. The toxic effects of HCB have included hirsutism.

SEXUAL AND ENDOCRINE FUNCTION IN MEN Gynecomastia may occur as a result of hepatic injury from carbon tetrachloride or other industrial toxins. In these conditions, hepatic inactivation of endogenous estrogens is diminished.51 Gynecomastia also has been reported in workers engaged in the manufacture of stilbestrol.52 Organochlorine insecticides, such as lindane, have estrogenic activity, and extensive exposure could cause gynecomastia.53 Presumably, the ingestion of large amounts of phytoestrogens could cause gynecomastia. The effects of environmental exposures on male fertility have been studied extensively in humans. At least half a dozen agents cause semen abnormalities, and others cause testicular toxicity in animals.54 Much of the research followed the discovery that the pesticide dibromochloropropane caused an epidemic of infertility in workers involved in its manufacture; 13% of the exposed group had azoospermia, and another 30% had oli-gospermia (see Table 235-1). For the substances that have been investigated in humans, the mechanism in most cases is a direct toxic effect on the testes, rather than suppression of the hypothalamic-pituitary axis. In these studies, serum follicle-stimulating hormone levels are normal unless azoospermia results, at which time they rise. The mechanism of action of the ethylene glycol ethers has been studied extensively in animals; they directly affect the early pachytene spermatocyte. Some animal experiments suggest that a toxic effect on testicular function is mediated by hypothalamic dysfunction. For example, methyl chloride suppresses the release of luteinizing hormone–releasing hormone, causing a decreased sperm count.55 Evidence exists that lead may suppress pituitary function, causing testicular atrophy. However, testicular biopsies of azoospermic men exposed to lead have revealed fibrosis, suggesting a direct testicular effect as well.56 Significant exposure to lead (blood levels of 23–98 µg/dL) were associated with abnormal sperm morphology and, to some degree, with sperm immotility.57 Toxins may act at many stages of sperm development, causing abnormal sperm count, motility, or morphology. The initial studies in men primarily looked at sperm count per cubic centimeter (density) and sperm count per ejaculate as end points. Motility and morphology traditionally have been assessed in a relatively subjective manner, and these measures were fairly insensitive for detection of subtle differences between groups in an epidemiologic study. In recent years, the use of videomicrography for assessing sperm motility and computer-generated scoring systems for evaluating sperm morphology have allowed objective measurement of semen production (see Chap. 114 and Chap. 118).

FUEL METABOLISM DIABETES MELLITUS Environmental factors may play a role in the development of type 1 diabetes mellitus. An acute diabetes-like syndrome was reported from poisoning with an insecticide, N-3-pyridyl-methyl-N-p-nitrophenylurea (Vacor).58 This chemical is structurally related to alloxan and streptozocin, which induce islet cell failure. Whether low doses of similar chemicals could contribute to the development of diabetes is unknown. A study in Taiwan suggests that long-term arsenic exposure may induce diabetes mellitus in humans.59 HYPOGLYCEMIA Severe hepatic deficiency caused by environmental toxins can produce hypoglycemia. Parathion, an insecticide, can lower blood glucose levels in animals and has caused hypoglycemia in at least one case of human exposure.60

The ingestion of the unripe fruit of the ackee tree (primarily eaten in Jamaica) causes a poisoning that includes severe hypoglycemia. The fruit contains two toxins, L -a-aminomethyl-enecyclopropylproprionic acid and its a-glutamyl conjugate, which inhibit transport of long-chain fatty acids into mitochondria, suppress oxidation, and inhibit gluconeogenesis.61 LIPIDS TCDD and related aromatic halogenated hydrocarbons induce changes in animal lipid metabolism, including fatty liver. Serum triglyceride and free fatty acid levels may be increased at low doses, and both serum total cholesterol and high-density lipoprotein cholesterol levels may be increased at higher doses.62 TCDD exposure leads to depletion of brown adipose tissues of rats. This causes a significant energy imbalance, with less efficient energy use and subsequent wasting. Wasting has not been reported in humans, although lipid metabolism may be impaired. In humans exposed to PCBs, elevated serum triglyceride levels have been reported. The clinical significance is unknown, but this elevation may serve as a useful marker of exposure. Carbon disulfide, a solvent, causes elevation of serum cholesterol levels in exposed workers.63 Some phytoestrogens increase resistance of low-density lipoprotein to oxidation.64 Some have lipid-lowering effects.65

OCCUPATIONAL FACTORS The natural history of many of the chemically induced endocrine disorders discussed is obscure. As a general principle, removal from exposure to a toxin may cause regression of the disease, unless irreversible anatomic alterations have occurred. Removal from work must be prescribed with an understanding of the financial and legal implications.25,33 Some occupations entail risks to persons with coexistent endocrine disorders; these need to be considered in the treatment of the disease. The occupational factor most likely to have an effect is heat. Patients with diabetes insipidus, Addison disease, hypoaldosteronism, and diabetes mellitus require special education about fluid balance in job environments with significant risk for heat stress. These include bakeries; brick-firing and ceramics kilns; electrical utilities (especially boiler maintenance); food canneries; iron, nonferrous, and steel foundries; laundries; chemical, glass, and rubber plants; mines; outdoor locations in hot weather; restaurant kitchens; smelting facilities; and steam tunnels. Occupational exposure to cold also has endocrine effects. For example, one study has demonstrated an adaptive decrease in thyroid hormone–binding capacity with increases of free T4 and T3.66 That environmental and occupational agents can influence the endocrine system is now apparent. As our awareness of the omnipresent and often unsuspected toxins in our environment increases, and as our ability to detect subtle endocrine disorders improves, many additional and previously unsuspected relationships undoubtedly will appear.25,67 CHAPTER REFERENCES 1. Crisp TM, Clegg ED, Cooper RI, et al. Environmental endocrine disruption: an effects assessment and analysis. Environ Health Perspect 1997; 1:11. 1a. 2. 3. 4. 5. 6. 7. 8. 8a. 9. 10. 11. 12. 13.

Cooper RL, Kavlock RJ. Endocrine disruptors and reproductive development: a weight-of-evidence review. J Endocrinol 1997; 152:159. Butterworth KR, Mangham BA. The application of clinical toxicology. Crit Rev Toxicol 1987; 18:81. Clayton GD, Clayton FE. Patty's industrial hygiene. New York: John Wiley and Sons, 1981. Doull J, Klaossen CD, Amdur MO, eds. Casarett and Doull's toxicology. New York: Macmillan, 1980. Robins JM, Cullen MR, Conners BB. Depressed thyroid indexes associated with occupational exposure to inorganic lead. Arch Intern Med 1983; 143:220. Gustafson A, Heder P, Schutz A, Skerfving S. Occupational lead exposure and pituitary function. Int Arch Occup Environ Health 1989; 61:277. Foster WG, McMahon A, YoungLai EV, et al. Reproductive endocrine effects of chronic lead exposure in the male cynomolgus monkey. Reprod Toxicol 1993; 7:203. Mutti A, Vescovi PP, Falzoi M, et al. Neuroendocrine effects of styrene on occupationally exposed workers. Scand J Work Environ Health 1984; 10:225. Brucker-Davis F. Effects of environmental synthetic chemicals on thyroid function. Thyroid 1998; 8:827. Van Etten CH. Goitrogens. In: Liener IE, ed. Toxic constituents of plant foodstuffs. New York: Academic Press, 1969:103. Smith DM. Ethylene thiourea: thyroid function in two groups of exposed workers. Br J Ind Med 1984; 41:362. Bahn A, Mills J, Snyder P, et al. Hypothyroidism in workers exposed to polybrominated biphenyls. N Engl J Med 1980; 302:31. McKinney JD, Waller CL. Polychlorinated biphenyls as hormonally active structural analogues. Environ Health Perspect 1994; 102:290. Safe S. Toxicology, structure-function relationship, and human and environmental health impacts of polychlorinated biphenyls: progress and problems. Environ Health Perspect 1993; 100:259.

13a. Brouwer A, Longnecker MP, Birnbaum LS, et al. Characterization of potential endocrine-related health effects at low-dose levels of exposure to PCBs. Environ Health Perspect 1999; 107:639. 14. Peters HA, Gocmen A, Cripps DJ, et al. Epidemiology of hexachloroben-zene-induced porphyria in Turkey. Arch Neurol 1980; 39:744. 15. Van Raaij JAGM, Frijters CMG, Van den Berg KJ. Hexachlorobenzene induced hypothyroidism. Biochem Pharmacol 1993; 46:1385. 16. Den Besten C, Bennik MHJ, Bruggleman I, et al. The role of oxidative metabolism in hexachlorobenzene-induced porphyria and thyroid hormone homeostasis: a comparison with petachlorobenzene in a 13-week feeding study. Toxicol Appl Pharmacol 1993; 119:181. 17. Vanden Heuvel JP, Lucier G. Environmental toxicity of polychlorinated dibenzo-p-dioxins and polychlorinated dibenzofurans. Environ Health Per-spect 1993; 100:189. 18. Whitlock JP. Genetic and molecular toxicity of 2,3,7,8-tetrachlorobenzo-p-dioxin action. Annu Rev Pharmacol Toxicol 1990; 30:251. 18a. Kohn MC. Effect of TCDD on thyroid hormone homeostasis in the rat. Drug Chem Toxicol 2000; 23:259. 19. Rattner BA. Endocrine responses to avian environmental pollutants. J Exp Zool 1984; 232:683. 20. Rawlings NC, Cook SJ, Waldbillig D. Effects of the pesticides carbofuran, chlorpyrifos, dimethoate, lindane, triallate, trifluralin, 2,4-D, and pentachlo-rophenol on the metabolic, endocrine and reproductive endocrine system in ewes. J Toxicol Environ Health 1998; 54:21. 21. Conrad RA, Paglia DE, Larsen PR. Review of medical findings in a Mar-shallese population 26 years after accidental exposure to radioactive fallout. Upton, NY: Brookhaven National Laboratory, 1980. Brookhaven National Laboratory report 51261. 22. Larsen PR, Conrad RA, Knudsen KD. Thyroid hypofunction after exposure to fallout from a hydrogen bomb explosion. JAMA 1982; 247:1571. 23. Reizenstein P. Carcinogenicity of radiation doses caused by the Chernobyl fallout in Sweden and prevention of possible tumors. Med Oncol Tumor Pharmacother 1987; 4:1. 24. Zeighami EA, Morris MD. Thyroid cancer risk in the population around the Nevada test site. Health Phys 1986; 50:19. 25. Rosenstock L, Cullen MR. Clinical occupational medicine. Philadelphia: WB Saunders, 1994. 26. Klein RF, Wiren KM. Regulation of osteoblastic gene expression by lead. Endocrinology 1993; 132:2531. 27. Cagin CR, Diloy-Puray M, Westerman MP. Bullets, lead poisoning and thy-rotoxicosis. Ann Intern Med 1978; 89:509. 28. Hunter D. The diseases of occupations. London: English Universities Press, 1975:879. 29. Newton RW, Browning MC, Iqbal J, et al. Adrenocortical suppression in workers manufacturing synthetic glucocorticoids. BMJ 1978; 1:73. 30. Reif RM, Lewinsohn G. Paraquat myocarditis and adrenal cortical necrosis. J Forensic Sci 1983; 28:505. 31. Ribelin WE. Effects of drugs and chemicals upon the structure of the adrenal gland. Fundam Appl Toxicol 1984; 4:105. 32. Colby HD, Eacho PI. Chemically induced adrenal injury: role of metabolic activation. In: Thomas JA, ed. Endocrine toxicology. New York: Raven Press, 1985:35. 33. Welch LS. Decision making about reproductive hazards. Semin Occup Med 1986; 1:97. 34. Paul M. Clinical evaluation and management. In: Occupational and environmental reproductive hazards: a guide for clinicians. Baltimore: Williams & Wilkins, 1993:113. 35. Egeland GM, Sweeney MH, Fingerhut MA, et al. Total serum testosterone and gonadotropins in workers exposed to dioxin. Am J Epidemiol 1994; 139:272. 36. Safe S, Commor K, Rananoorthy K, et al. Human exposure to endocrine-active compounds: hazard assessment problems. Regul Toxicol Pharmacol 1997; 26:52. 37. Bulger WH, Kupfer D. Estrogenic activity of pesticides and other xenobiotics. In: Thomas JA, ed. Endocrine toxicology. New York: Raven Press, 1985. 38. Wolff MS, Toniolo PG, Lee EW, et al. Blood level of organochlorine residues and risk of breast cancer. J Natl Cancer Inst 1993; 85:648. 39. Dewailly E, Dodin S, Verreault R, et al. High organochlorine body burden in women with estrogen receptor positive breast cancer. J Natl Cancer Inst 1994; 86:232. 40. Van't Veer P, Lobbezoo IE, Martin-Moreno JM, et al. DDT (dicophane) and postmenopausal breast cancer in Europe: case-control study. BMJ 1997; 31:81. 41. Lopez-Carrillo L, Blair A, Lopez-Cervantes M, et al. Dichlorodiphenyl-trichloroethane serum levels and breast cancer risk: a case-control study from Mexico. Cancer Res 1997; 57:3728. 42. Hunter DJ, Hankinson SE, Laden F, et al. Plasma organochlorine levels and the risk of breast cancer. N Engl J Med 1997; 337:1253. 43. Wolff MS. Occupationally derived chemicals in breast milk. In: Mattison DR, ed. Reproductive toxicology, vol 117. New York: Alan Liss, 1983:259. 44. Kacew S. Adverse effects of drugs and chemicals in breast milk on the nursing infant. J Clin Pharmacol 1993; 33:213. 45. Rier HE, Martin DC, Bowman RE, et al. Endometriosis in Rhesus monkeys (Macaca mulatta) following chronic exposure to 2,3,7,8-tetrachlorbenzo-p-dioxin. Fundam Appl Toxicol 1993; 21:433. 45a. Bowers JL, Tyulmenkov VV, Jernigan SC, Klinge CM. Resveratrol acts as a mixed agonist/antagonist for estrogen receptors alpha and beta. Endocrinology 2000; 141:3657. 46. Bennetts HW, Underwood EJ, Shier FL. A specific breeding problem of sheep on subterranean clover pastures in Western Australia. Aust Vet J 1946; 22:2. 47. Kallela K, Heinonen H, Saloniemi H. Plant oestrogens; the cause of decreased fertility in cows. A case report. Nord Vet Med 1984; 36:124. 48. Lightfoot RJ, Croker KP, Neil HG. Failure of sperm transport in relation to ewe infertility following prolonged grazing on oestrogenic pasture. Aust J Agric Res 1967; 18:755.

49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67.

Drane HM, Patterson DSP, Roberts BA, Saba N. The chance discovery of oestrogenic activity in laboratory rat cake. Food Chem Toxicol 1975; 13:491. Sectell KDR, Doriello SD, Hulme P, et al. Nonsteroidal estrogens of dietary origin: possible roles in hormone dependent disease. Am J Clin Nutr 1984; 40:569. Morrione T. Effects of estrogens on the testes in hepatic insufficiency. Arch Pathol 1944; 37:38. Fitzsimmons M. Gynaecomastia in stilboestrol workers. Br J Ind Med 1944; 1:235. Raizada RB, Pushpa M, Saxena I, et al. Weak estrogenic activity of lindane in rats. J Toxicol Environ Health 1980; 6:483. Wyrobek AJ, Gordon LA, Bulchart JG. An evaluation of human sperm as indicators of chemically induced alterations of spermatogenic function: a report of the US Environmental Protection Agency Gene-Tox Program. Mutat Res 1983; 114:77. Chapin RE, White RD, Morgan KT, Bus JS. Studies of lesions induced in the testis and epididymis of F-344 rats by inhaled methyl chloride. Toxicol Appl Pharmacol 1984; 76:328. Cullen MR, Robins JM, Eskenazi B. Adult inorganic lead intoxication—presentation of 31 new cases and a review of recent advances in the literature. Medicine (Baltimore) 1983; 62:221. Robins TG, Bornman MS, Ehrlich RI, et al. Semen quality and fertility of men employed in a South African lead acid battery plant. Am J Ind Med 1997; 32:369. Pesticidal diabetes. (Editorial). BMJ 1979; 2:292. Lai M-S, Hsueh Y-M, Chen C-J, et al. Ingested inorganic arsenic and prevalence of diabetes mellitus. Am J Epidemiol 1994; 139:484. Hruban Z, Schulman S, Warner NE, et al. Hypoglycemia resulting from insecticide poisoning. JAMA 1969; 184:509. Tanaka K, Kean EA, Johnson B. Jamaica vomiting sickness. Biochemical investigation of two cases. N Engl J Med 1976; 295:461. Poland A, Knutson JC. 2,3,7,8-Tetrachlorodibenzo-p-dioxin and related halogenated aromatic hydrocarbons: examination of the mechanism of toxicity. Annu Rev Pharmacol Toxicol 1982; 22:517. Lillis R. Carbon disulfide. In: Rom WN, ed. Environmental and occupational medicine. Boston: Little, Brown and Company, 1983:627. Wiseman H, O'Reilly JD, Adlercreutz H, et al. Isoflavone phytoestrogens consumed in soy decrease F(2)-isoprostane concentrations and increase resistance of low-density lipoprotein to oxidation in humans. Am J Clin Nutr 2000; 72:395. Lissin LW, Cooke JP. Phytoestrogens and cardiovascular health. J Am Coll Cardiol 2000; 35:1403. Solter M, Brkic K, Petek M, et al. Thyroid hormone economy in response to extreme cold exposure in healthy factory workers. J Clin Endocrinol Metab 1989; 68:168. Rom WN, ed. Environmental and occupational medicine. Boston: Little, Brown and Company, 1992:3.

CHAPTER 236 COMPENDIUM OF ENDOCRINE-RELATED DRUGS Principles and Practice of Endocrinology and Metabolism

CHAPTER 236 COMPENDIUM OF ENDOCRINE-RELATED DRUGS DOLLY MISRA, MICHELLE FISCHMANN MAGEE, AND ERIC S. NYLÉN

TABLE 236-1.

Index of General Categories of Drugs Found in Table 236-2

There has been a great proliferation of preparations used in the clinical practice of endocrinology, and this will undoubtedly increase further with time. Any drug that is to be entered into the market in the United States goes through a rigorous certification procedure that may last for years, until it is “approved” for use in humans by the U.S. Food and Drug Administration (FDA). The manufacturer must first obtain an approved New Drug Application (NDA) from the FDA. A Notice of Claimed Investigational Exemption for a New Drug (IND) must then be accepted by the FDA. Subsequently, the drug is entered into phase I trials in normal subjects to identify “tolerable” dose ranges. In phase II trials, limited numbers of patients are treated with the drug to establish efficacy and doses. Finally, phase III involves more extensive clinical trials in patients. After a drug has been approved and is in clinical use, there is an ongoing process of postmarketing surveillance, whereby all reports of adverse effects, additional clinical experience, and other relevant data must be submitted by the manufacturer to the FDA. Drugs are approved for use for specific indications. The FDA does not attempt to compel prescribers to adhere to officially labeled uses. Unlabeled uses are considered to be proper if based on “rational scientific theory, reliable medical opinion, or controlled clinical studies.” For all prescription drugs, FDA-approved labeling must be contained in the package insert or official label. This provides a summary of information about the “chemical and physical nature of the product, its pharmacology, the indications and contraindications, the means of administration, the appropriate dosage, side effects and adverse reactions, how the drug is supplied, and any other information pertinent to its safe and effective use.” The Physicians' Desk Reference (PDR) is mainly a compilation of these official labels. The use of any drug during pregnancy should be considered only if clearly needed. Before using a drug in a pregnant woman, one should consider its potential to cause harm to the fetus. The FDA has established five pregnancy categories for drugs. On the basis of the degree of documentation required concerning the validity of the relationship, these categories indicate the potential for a drug to cause birth defects and also the drug risk/benefit ratio in pregnancy. This chapter is a practical listing of many drugs in current use in the clinical practice of endocrinology. The information in Table 236-2, which follows, provides specifics for each drug, including preparations available, indications for use, dose ranges and rates of administration, and other comments that the authors feel are germane to the agents' use. Where such specifics are discussed in detail in other parts of this textbook, the reader is referred to the relevant chapter. This table is intended as a useful lexicon of the available drugs and many of their proprietary (trade) names. Their use is not necessarily being recommended. The indications, dose, side effects, precautions, and usage in pregnancy are presented in an abbreviated form, and the following references should be checked for more complete details. The United States Pharmacopeia (USP DI), Drug Information for the Health Care Professional 2000. Published by an alliance between the United States Pharmacopeial Convention and Micromedex, Inc. Contains specific pharmacological information on drugs, including formulae, formulations of preparations, and assays for identification. On-line access: www.micromedex.com Drug Facts and Comparisons, 2000. Facts and Comparisons Division of JB Lippincott Co, Philadelphia. A comprehensive drug information compendium whose core of information is derived from the most current FDA-approved package literature, updated monthly. On-line access: www.drugfacts.com Physicians' Desk Reference, 2001. Medical Economics Inc, Oradell, New Jersey. Product overviews of selected drugs, usually based on official package circulars. On-line access: www.pdr.net Physicians' Desk Reference for Nonprescription Drugs, 2001. Medical Economics Inc, Oradell, New Jersey. A guide to the contents of many nonprescription drugs. Drug Information, 2000. Compiled by American Hospital Formulary Service and published by the American Society of Health-System Pharmacists, Inc, Bethesda. A detailed guide of drugs and their use. The Medical Letter. Bi-weekly publication of evidence-based drug data. On-line access: www.medletter.com TABLE 236-1.

Index of General Categories of Drugs Found in Table 236-2

CHAPTER 237 REFERENCE VALUES IN ENDOCRINOLOGY Principles and Practice of Endocrinology and Metabolism

CHAPTER 237 REFERENCE VALUES IN ENDOCRINOLOGY D. ROBERT DUFOUR Reference Ranges: Do They Mean “Normal”? Definitions of Normal Reference Values Medical Decision Levels Individual Reference Ranges Common Causes of “Abnormal” Endocrine Test Results Physiologic Influences on Endocrine Test Results Effects of Drugs on Endocrine Test Results Hemoconcentration Changes After Collection of Specimens Clerical and Analytic Errors Use of Laboratory Tests in Diagnosis “Abnormal” Test Results—Bayes' Theorem Chapter References Reference Values in Endocrinology

Laboratory tests are widely used by endocrinologists in the diagnosis and monitoring of disease. On receiving the results of such tests, the first thing physicians do is check to see whether the results are “normal” by comparing them with a list of values prepared by the laboratory, which commonly are referred to as “normal ranges.” Textbooks often provide such lists. Although comparing a test result to a published standard may seem simple, many considerations may influence the interpretation. This chapter, instead of simply providing a list of normal values, discusses the factors that may affect the interpretation of test results as well as common situations that may change test results. Table 237-1, at the end of the chapter, lists common factors that may influence the results of each of these tests.

REFERENCE RANGES: DO THEY MEAN “NORMAL”? DEFINITIONS OF NORMAL Although one review documented seven possible meanings of the word normal, to most endocrinologists and their patients, normal is synonymous with absence of disease.1 By corollary, abnormal indicates the presence of disease, which has caused a change from the normal, healthy state. A second common meaning of normal, which is derived from statistics, is often used by scientists: a bell-shaped frequency distribution in which 95% of all results are within 2 standard deviations (SD) of the average value. Although these two definitions do not necessarily describe the same thing, they frequently are used interchangeably in medicine, particularly in the interpretation of laboratory results.2,3 REFERENCE VALUES All clinical pathology laboratories publish a list of reference values. Until recently, these were called tables of normal values, with the laboratory using the term normal in the sense of the second definition given earlier. Reference value now is widely used to indicate that the value provides a reference, something to be used for comparison with a test result, rather than a declaration of values expected in normal, healthy persons.4 This subtle but important difference is the source of much confusion among physicians seeking to interpret the results of laboratory tests. Less commonly, reference values may be established for disease, such as expected values of glycated hemoglobin in persons with varying degrees of glycemic control. Because the principles and assumptions used in establishing such reference values are similar, this latter type of reference range is not discussed further here. ESTABLISHMENT OF A REFERENCE RANGE Reference ranges can be established in several ways. A sample of suitable size is needed to guarantee the statistical validity of results. Usually, 100 to 200 persons constitute a sample large enough to be valid but small enough to be workable. For many difficult or expensive procedures, published reference ranges may be based on results from fewer than 20 persons. Representative Sampling. The first assumption is that the sample is representative of the population that ultimately will be tested. Most volunteers are hospital employees, blood donors, or medical students. None of these groups is necessarily representative of the population that eventually will be examined.5 This is especially a problem in inner city and charity hospitals, in which the usual volunteers are not likely to be from the same population as the patients. If the sample is not representative, the reference range may or may not be valid, but its validity cannot be tested. Absence of Disease. A second assumption is that no persons in the sample group have the diseases for which the testing is being done. This is important for disorders with a high prevalence of asymptomatic disease. An example is the current widespread effort to screen the U.S. population for risk factors for coronary artery disease. Autopsy studies show a high incidence of atherosclerosis in persons without symptoms.6 Serum cholesterol is thought to be related to the development of atherosclerotic plaques.7,8 To develop a reference range for serum cholesterol levels to separate persons without atherosclerosis from those with atherosclerosis, a sample of persons without atherosclerosis would have to be selected. This is not commonly done. Until the late 1980s, most laboratories used reference ranges derived from “normal” volunteers, many of whom actually had atherosclerosis. This led to extremely wide reference ranges in which virtually no one except those with familial hypercholesterolemia had abnormal results.9 Currently, “reference” values for cholesterol are based on the risk of development of coronary artery disease, rather than the selection of “typical” values.10 (Other common disorders that may affect laboratory tests are diabetes mellitus, hypertension, and alcohol abuse.) Conditions of Testing. Although the method used for obtaining samples is not commonly considered, differences in patient preparation and the technique used for venipuncture can affect test results.11 Fasting increases bilirubin and uric acid levels, whereas glucose, triglycerides, and insulin levels increase after meals. A number of constituents show diurnal variation, including cortisol, osteocalcin, and urinary excretion of collagen fragments (N-telopeptide, deoxypyridinoline). If all volunteers are requested to come in after an overnight fast (a common instruction given by laboratories), then reference ranges will be valid only for comparing samples obtained under those conditions.5 In most clinical settings, patients are seen at other times of the day for blood drawing, and they may have results outside the reference range solely for that reason. In an elegant series of experiments, the effect of various forms of prevenipuncture position on test results was evaluated.12 Drawing blood after the person stood and walked for 30 minutes before sitting down increased the apparent concentrations of proteins and protein-bound molecules (which affect many of the tests related to endocrine evaluation) by 3% to 5% over those obtained after the same person had been sitting for 30 minutes. If the person had been supine for 30 minutes before venipuncture, a 5% to 10% decrease was seen in the same constituents. Because the first condition describes the common procedure for obtaining blood from outpatients, and the second describes the usual method of obtaining blood from inpatients, different interpretations of results could easily be made. In virtually all laboratories, reference ranges are derived using samples obtained from ambulatory persons, so that some results are 8% to 15% higher than expected for supine inpatients. STATISTIC INTERPRETATION Mean and Standard Deviation. The typical approach in evaluating the data is to consider the central 95% of all values as the reference range. In most laboratories, this is determined by calculating the mean and SD and defining the reference limits as the mean ± 2 SD. Several problems are associated with this seemingly simple statistical procedure. The assumption is that the data are graphically distributed in a bell-shaped fashion, that is, as the Gaussian or normal distribution. For many parameters, distribution of results is not symmetric but is skewed to one side or the other.13 Often, the degree of skewing of the data is not enough to invalidate the assumption that values within 2 SD of the mean include 95% of all data points. Some statistical tests demonstrate whether the degree of skewing is too great for this assumption to be used.14 Many laboratories neither inspect the data for skewing nor use these statistical tests to determine whether skewing is too great; they simply define the reference range as the mean ± 2 SD. Figure 237-1 illustrates the consequences of this approach. If values are skewed toward higher concentrations, >5% of the sample have results above the upper reference limit, and the converse is true for values skewed to lower concentrations. Even if valid sampling techniques are used, such statistical errors lead to incorrect reference limits. Evaluation of the sensitivity of several statistical techniques to differences in data distributions has shown

that some are virtually unaffected by skewing of results.15

FIGURE 237-1. This skewed distribution of test results is typical of the pattern observed for many serum constituents. Using the mean ± 2 standard deviations (SD) as reference limits would falsely classify results between 0 and 5 as normal and results between 40 and 45 as abnormal.

Limitations of Gaussian Statistics. An even more important problem is associated with this method of defining limits of a reference range: Gaussian statistics describe the distribution of repeated measurements of the same parameter.16 For example, if 100 different scales were used to measure one person's weight, the chance would be 95% that the actual weight would lie within 2 SD of the average value for all weight measurements. When results obtained from populations are described, no valid statistical reason exists for selecting these limits as defining an expected range of normality. Once limits are set to include only the central 95% of all values as normal, then 5% of results from the sample are considered to be outside the reference limits, even though all persons in the sample originally were considered normal. Thus, if reference limits are used as a guide to what is normal in the medical sense, then 5% of results from the sample must indicate disease. This same condition holds true for each test performed. If each test assays an independent variable, and results for each test are distributed randomly in the population, then the likelihood that n tests are within the reference range is (0.95)n. The likelihood (p) that all independent test values are within the reference ranges (assuming that reference limits include 95% of all values) is as follows for different values of n: n = 1, p = .95; n = 2, p = .90; n = 6, p = .74; n = 10, p = .60; n = 12, p = .54; n = 15, p = .46; n = 20, p = .36; n = 24, p = .29. As n becomes larger and approaches the number of tests commonly included in screening profiles by many laboratories, most patients should have at least one result outside the reference limits solely because of the chance distribution of results. Because of this statistical fact, results outside reference values are frequently produced, even for apparently healthy persons. The author has been using results of medical students' laboratory tests to illustrate this phenomenon to the students for several years. When samples were taken from 597 healthy medical students, and reference values were defined as within 2 SD of the average, only 65% of students had all “normal” test results; 27% had one “abnormal” test result; and 8% had more than one “abnormal” test result. When the central 95% of the distribution was used as the reference values (because several tests showed statistically significant skewness), the results were similar (64% had all “normal” results, 27% had one “abnormal” result, and 9% had two or more “abnormal” results), although in ~15% of students, the two methods disagreed on the number of “abnormal” results. Several publications have highlighted this statistical phenomenon and have used it to suggest that widespread screening programs are unlikely to be beneficial, because most “abnormal” results are probably the result of random statistical occurrences.17,18 The statistical assumptions of independence become more complicated, however, because the results of many commonly measured tests change in parallel, rather than independently (i.e., sets of compounds such as serum electrolytes; calcium, phosphate, and magnesium; blood urea nitrogen and creatinine; and triiodothyronine [T3], thyroxine [T4], and thyroid-stimulating hormone [TSH]). Another assumption limiting the use of reference values is that results are randomly distributed in the population, and that an individual is no more likely to have any one result than another. Many studies have shown that each individual maintains concentrations of many critical parameters within an extremely narrow range and, thus, results would not be randomly distributed in the population.19,20 and 21 The ratio of average variation within an individual to the variation within a population has been termed the index of individuality. An index below 1.4 suggests that standard reference ranges will be insensitive markers of significant changes in the condition of a person. As an example, the range of cholesterol values seen in the medical students (central 95% of values) is from 126 to 252 mg/dL. In contrast, the average day-to-day variation in cholesterol level in a person in relation to his or her average blood cholesterol level is ~6% (1 SD). Thus, significant changes in serum cholesterol can occur without producing an “abnormal” result. Other common tests with small individual variation in results include calcium, alkaline phosphatase, and thyroid function tests. IMPLICATIONS OF A REFERENCE RANGE The usefulness of reference ranges depends on the techniques used to establish them. Even if all of the influences have been considered, it does not follow that a result within the reference range indicates health or that a result outside the reference range indicates disease. Because the reference range is a descriptive statistic of the sample (and, if valid sampling techniques were used, of the population), then results outside the reference range indicate only that an individual is not from the same population. This could be due to disease, to population differences, or to physiologic variations in the individual. Similarly, a result within the reference range does not guarantee health, because some diseases may have a high enough prevalence in the population that the test cannot separate healthy and diseased individuals. These observations have led to several alternatives to reference ranges. MEDICAL DECISION LEVELS Medical decision levels are cutoff points that reliably indicate disease or the need for additional action.22 The limits for these values usually are wider than those for reference ranges, so that finding a result beyond these limits as a result of chance alone would be uncommon. Because they indicate the need for medical intervention, less reason exists to consider qualifying factors in interpreting results. Tables of medical decision levels are found in several articles, and one book gives medical decision levels for 137 tests.23 Decision levels may prove useful for analytes that are commonly subject to minor, physiologic variations. SENSITIVITY AND SPECIFICITY Sensitivity refers to the ability to detect disease, expressed as percentage of patients with disease having an abnormal test result. Specificity refers to the ability to exclude disease, expressed as percentage of persons without disease having a normal test result. One limitation of medical decision levels is that, although they are highly specific for disease, they tend to be less sensitive and are of less use in screening persons for early stages of a disease process that produces gradual changes in test results. Figure 237-2 shows the effect of changing the reference limits of a test to improve either sensitivity or specificity: Improving the sensitivity and specificity of a given test simultaneously is impossible. For most disease-test combinations, sensitivity and specificity are unknown.24 In addition, the use of sensitivity and specificity has been criticized, because values usually are derived by applying the test to patients with known disease, which was confirmed by some other technique of equal or greater specificity and sensitivity. The original results for sensitivity and specificity are unlikely to remain valid when the same test is used for patients with fewer manifestations of disease.25

FIGURE 237-2. The distribution of fasting glucose levels in patients with normal glucose tolerance (solid line) and those with “diabetic” glucose tolerance (dotted line) illustrates the impossibility of improving both sensitivity and specificity simultaneously. Use of a fasting glucose level of 140 mg/dL (solid bar) to predict abnormal glucose tolerance excludes all persons with normal glucose tolerance (specificity 100%), but it misses 40% of those with abnormal glucose tolerance (sensitivity 60%).

With use of a fasting glucose level of 115 mg/dL (open bar), sensitivity is improved to 100%, but specificity falls to 75%. With any value between these two points, specificity improves and sensitivity falls. Moving below 115 or above 140 mg/dL lowers specificity or sensitivity, respectively, without improving the other component.

RELATIVE OPERATING CHARACTERISTIC CURVES Relative operating characteristic (ROC) curves are a graphic depiction of sensitivity and specificity over a continuous range of cutoff points or decision levels.26 Typically, sensitivity is plotted in an ascending fashion on the Y-axis and specificity is plotted in a descending fashion on the X-axis. Such graphs are useful for selecting optimal decision points to use in diagnosing disease and for comparing the performance of diagnostic tests. The point closest to the upper left corner of the graph (perfect performance) correctly classifies the greatest percentage of patients. For comparison of two or more tests, the curve closer to the upper left corner performs better in clinical usage. Decision points can be selected to give high sensitivity or high specificity, depending on the use of the test. When medical decision levels are used, the specificity is increased to virtually 100%, meaning that all abnormal results indicate disease. Figure 237-3 illustrates a typical ROC curve for a diagnostic test. Several possible cutoff points are indicated, showing why medical decision levels have a lower sensitivity than other possible decision levels.

FIGURE 237-3. This relative operating characteristic curve illustrates the results obtained in several screening programs for primary hyper-parathyroidism. Point A represents a serum calcium level of 10.0 mg/dL; all patients with primary hyperparathyroidism continuously have serum calcium levels above this value, but ~30% of normal persons have one or more values above this. Point B represents an upper reference limit of 10.5 mg/dL. Approximately 90% of serum calcium values from patients with primary hyperparathyroidism fall above this level; 5% to 8% of values from normal persons are above it, usually due to hemoconcentration. Point C is the medical decision level of 11.0 mg/dL recommended by Statland.26 Only 70% of values from patients with primary hyperparathyroidism are above this level, but it reliably indicates an elevated serum calcium value of clinical significance.

INTERLABORATORY DIFFERENCES Because laboratories use many different techniques for measuring most analytes, different cutoff values may be needed for different laboratories. Thus, some time will be required before enough universally applicable formulas are available for diagnosis of disease by laboratory tests alone. INDIVIDUAL REFERENCE RANGES Another approach is the use of individual reference ranges.4,27 The concentrations of many analytes are closely regulated in most individuals, so that the variation within a person is much smaller than the variation between persons.19,20 and 21 With the feasibility of performing widespread population screening for different disease states and the common practice of doing “profiles” of laboratory tests as part of periodic physical examinations, a database exists for determining a person's typical concentrations of several analytes. If the results of laboratory tests are available on computers, this analysis is easier to perform. One way in which these ranges can be useful is in diagnosing myocardial infarction. In one common approach, standard reference ranges are used and serum isoenzyme determinations of creatine kinase are obtained only if the enzyme levels exceed the reference limit. In several studies, many patients with apparent myocardial infarction had extremely low baseline values for serum creatine kinase and showed acute rises in the level of this enzyme that did not exceed the upper reference limit.28,29 Thus, comparing a patient's test results to his or her own values improves the sensitivity of creatine kinase measurements for diagnosing myocardial infarction.30 The use of individual reference ranges should better define the expected limits of normal physiologic changes and provide the ultimate in sensitivity for early detection of disease. As discussed earlier, intraindividual variation is generally much less than the variation seen in the population.31 The major impediments are the extremely high cost of screening the population using every conceivable test and the enormous amount of data storage capacity required. Another possible problem is the lack of acceptable reference ranges for allowable changes in an individual. Although the studies cited provide approximate reference ranges for intraindividual variation, they are not all inclusive, and a wide difference is seen in the values given. Moreover, data suggest that intraindividual variation is greater in persons with certain disease processes.32 A range that may detect early changes in healthy persons may be a false-positive indicator of impending complications in persons with disorders such as diabetes. Despite these problems, the use of individual reference ranges probably will increase.

COMMON CAUSES OF “ABNORMAL” ENDOCRINE TEST RESULTS Although reference ranges often are used to define “normal,” results outside reference limits are commonly seen when no evidence of an underlying disease process is present. In this section, some frequent causes of variation in laboratory test results are described. A more complete listing is available in Table 237-1 at the end of this chapter. PHYSIOLOGIC INFLUENCES ON ENDOCRINE TEST RESULTS The effects of normal physiologic changes in a patient are not often considered in interpreting laboratory findings, yet they are commonly the explanation for an unexpected test result. DIET Occasionally, dietary factors are responsible for unexpected changes, either due to the ingestion of food (e.g., transient alkalosis, intracellular shifts in phosphate, release of intestinal alkaline phosphatase into the circulation, elevated levels of hormones) or due to the chemical substances in food that cause physiologic changes (e.g., caffeine-induced catecholamine release).33,34 DIURNAL AND PULSATILE RHYTHMS Diurnal variation is a well-known phenomenon for cortisol and adrenocorticotropic hormone, but smaller degrees of diurnal variation also are found for other hormones, including prolactin, TSH, and testosterone.35,36 and 37 Marked diurnal variation in serum iron concentrations is a common finding, and many hormones, such as gonadotropins, growth hormone, and prolactin, are released in a pulsatile fashion, such that interpreting a single test result is difficult.38 Even for substances measured by common hematology and chemistry tests, a significant diurnal variation is seen.38a For measurement of most pituitary hormones, pooling several serum samples and assaying the pooled specimen provides a more accurate indication of average hormone concentration and pituitary function. For most other tests, consistently sampling in the early morning yields more reproducible results. PHYSICAL AND EMOTIONAL STRESS Exercise leads to the release of catecholamines, prolactin, and muscle-specific enzymes into the circulation.39 Even mild exercise can cause a marked increase in the frequency of menstrual irregularities,39a probably due to hormonal changes. Stress leads to catecholamine release; prolonged stress can cause marked changes in the

concentrations of various blood lipids and hormones.40,41 and 42 An entire section of this book (Part XVI) is devoted to the endocrine changes seen in acute illness. AGE The values of many serum constituents change markedly in concentration during a person's lifetime. Although it is accepted that values may differ in children (see Chap. 7, Chap. 18, Chap. 47, Chap. 70, Chap. 83, Chap. 87, Chap. 90, Chap. 91, Chap. 92, Chap. 157, Chap. 161 and Chap. 198), changes also occur at other times in life. For example, testosterone and renin decrease with increasing age in adults, and alkaline phosphatase and parathyroid hormone (PTH) levels continue to increase in persons older than 50 years. (Also see Chap. 199.) RACE For many commonly measured parameters, values in persons of African ancestry are different from those in persons of European ancestry; differences in other races have not been evaluated as carefully. For example, blacks tend to have higher values for high-density lipoprotein cholesterol and PTH, but lower values for vitamin D metabolites and renin. SEX Obvious differences are found between the sexes in levels of sex hormones and prolactin, but women and men often have different concentrations of many of the analytes commonly measured. Serum levels of free T4 and copper are higher in women, but lower values for renin, aldosterone, and most blood lipids also are the rule. MENSTRUATION AND PREGNANCY During the normal menstrual cycle, in addition to the obvious changes in levels of estrogens and progesterone, levels of vasopressin, prolactin, and PTH increase. The onset of menstruation causes a fall in serum sodium and phosphate levels, and a rise in renin and aldosterone concentrations. Pregnancy induces some unexpected changes: Levels of PTH, calcitonin, cortisol, and aldosterone increase, and fasting levels of glucose and glycohemoglobin (hemoglobin A1c) fall. The effects of pregnancy may last long after the baby is delivered. Studies have shown lower levels of prolactin and dehydroepiandrosterone in women who have borne children than in those who have never been pregnant. HEIGHT AND WEIGHT A person's height and weight are related to the concentrations of several substances. In children, a positive correlation is seen between height and serum alkaline phosphatase levels. Weight is much more closely associated with the concentrations of several parameters, especially in obese persons, who have higher serum concentrations of cortisol, insulin, and glucagon, but lower than normal levels of gonadotropins and sex hormone–binding globulin. Weight loss in obese persons causes changes, such as a fall in renin and aldosterone levels (see Chap. 126). ALTITUDE Changes in the concentrations of many analytes are found in persons living at altitudes above 5000 ft. Although some of these changes are transient, occurring only during the process of acclimatization, others appear to persist. Persons living at these heights tend to have higher levels of erythropoietin and lower levels of renin, angiotensin II, and aldosterone. EFFECTS OF DRUGS ON ENDOCRINE TEST RESULTS Drugs affect laboratory tests by two mechanisms.43 The first is a direct pharmacologic effect of the drug, such as hypokalemia due to diuretic therapy. Although most endocrinologists are familiar with this particular effect, many unfamiliar drug effects occur. For example, anticonvulsants increase levels of alkaline phosphatase (and related enzymes such as g-glutamyltransferase), prolactin, and vasopressin, but decrease total and free T4 and 25-hydroxyvitamin D levels, as well as urinary excretion of corticosteroid metabolites.44 A second type of drug interference results from cross-reaction in the assay. Two common examples are the cross-reaction of phenothiazines with some assays for urinary metanephrine, and the cross-reaction of certain cephalosporins with many assays for creatinine.45,46 Usually, information on the medications being taken by each patient tested is not given to the laboratory; even if this information were available, screening all test requests manually to detect possible drug-test interferences would be virtually impossible. With the increasing use of computers, however, it should be possible to construct a program to review the pharmacy record for each patient and search for drug interferences when abnormal results are encountered. Several comprehensive lists of drug effects on laboratory tests are available,47,48 and a compendium of drug effects on endocrine laboratory tests is provided in Chapter 239. HEMOCONCENTRATION Hemoconcentration causes an increase in proteins (e.g., albumin) and, consequently, in the level of protein-bound substances in the blood. The common causes of hemoconcentration are dehydration, postural differences, and the use of tourniquets during blood collection, but evaporation of serum after collection may produce the same effect. Ambulatory patients have a mild degree of relative hemoconcentration as a result of the shift of extracellular fluid from intravascular to extravascular locations, caused by the increased hydrostatic pressure in the upright position.12 This causes an increase of 3% to 5% in the concentration of proteins and protein-bound substances (e.g., some hormones, lipids, and ions such as Mg, Ca, and Fe). A similar hemoconcentrating effect is produced by leaving on a tourniquet for as little as 40 seconds while drawing blood.12 The suggested approach is to use a tourniquet only after a vein has been located and the skin has been prepared; longer use can cause 5% to 10% hemoconcentration. Fist clenching during blood collection, with a tourniquet applied, leads to leakage of muscle contents, especially potassium, which can increase by 1 to 1.5 mmol/L in as little as 1 minute. The combined effects of posture and tourniquet use may cause a 10% to 20% increase in the apparent concentration of proteins. In one study, 15% of all persons evaluated for hypercalcemia eventually were found to have normal serum calcium levels; hypercalcemia was artifactual as a result of concentration of serum proteins.49 In the author's laboratory, 5% to 10% of persons with “hypercalcemia” have serum albumin levels of >4.5 g/dL on initial study; usually, their calcium levels are within the reference range if their blood was collected (without prolonged use of a tourniquet) after they sat for 30 minutes. In one year, two patients were admitted for the evaluation of “hypercalcemia” that was found to be caused by hemoconcentration. CHANGES AFTER COLLECTION OF SPECIMENS Difficulties occasionally arise from changes that occur in the test tube during or after specimen collection. Some of the more common specimen-related problems include hemolysis, which increases the apparent concentration of all abundant substances within cells, such as potassium, phosphate, magnesium, and enzymes; and delay in transport of the specimen to the laboratory, which allows utilization and exhaustion of glucose, after which the cells leak their intracellular contents.50,51 Less commonly, extremely high white blood cell or platelet counts cause difficulties; the former increases the rate of glucose use, and the latter leads to the release of potassium during coagulation.52,53 In rare instances, an interaction occurs between a substance in the collection tube and the patient's blood that causes artifactual abnormalities, such as an increase in potassium induced by heparin in patients with extremely high lymphocyte counts.54 Renin levels increase with storage in ice water, as a result of the conversion of prorenin to renin.55 CLERICAL AND ANALYTIC ERRORS A final source of unexpected test results may be the laboratory itself if the results reported are not matched to the correct patient. Although all laboratories strive to minimize errors, such mistakes do occur. Clerical errors in specimen identification, which can occur at any time from collection to final reporting, are the single most common cause (25%) of erroneous results. 56 Less commonly, actual errors in performing the test cause incorrect results. Explainable causes of abnormal laboratory test results always should be considered. These factors are likely if the result is only minimally outside the reference limits; if evidence exists of hemoconcentration, such as a serum albumin level above the reference limit; if patients are receiving multiple medications; or if results from the current specimen differ markedly from previous results for a patient who is clinically stable. In any of these circumstances, the best approach is to contact the laboratory and submit a new specimen. Many reference laboratories perform repeated testing at no charge or at a reduced rate to maintain good customer

relationships and to document possible causes of unexpected results.

USE OF LABORATORY TESTS IN DIAGNOSIS Although the major purpose of this chapter is to familiarize the endocrinologist with the techniques used in laboratories that can influence the interpretation of laboratory tests, it would not be complete without a brief discussion of the uses of laboratory tests in endocrine diagnosis. “ABNORMAL” TEST RESULTS—BAYES' THEOREM Not all endocrine results that fall outside the reference range indicate disease. When such results occur, the physician must answer this question: What is the likelihood that this result indicates a particular disease in this patient? Several statistical models could be used to answer the question. The most widely publicized is Bayes' theorem. Bayes' theorem answers the question directly, giving a numeric probability that the test result indicates a given disease. The principles of Bayes' theorem are given in Figure 237-4. The information needed to answer the question includes the frequency with which abnormal results are found in persons with the disease (sensitivity), the frequency with which normal results are found in persons without the disease (specificity), and the frequency of the disease in the population. The answer is called the predictive value of a positive result. Figure 237-5 illustrates the importance of disease prevalence on the predictive value. The percentage of abnormal results that indicate disease decreases as the frequency of the disease decreases.

FIGURE 237-4. Bayes' theorem. The predictive value of a positive result answers the question: What is the likelihood that a positive result indicates the presence of disease? For any disease, increasing sensitivity or specificity of the tests used increases the predictive value. The major determinant of predictive value is disease prevalence. If the prevalence is low, then the true-positive results (TP) become small compared with false-positive results (FP), because TP = sensitivity × prevalence. The equation then reduces to TP/FP, or TP/[(1 – specificity) × (1 – prevalence)]. Because (1 – prevalence) is ~1, the predictive value is [sensitivity/([1 – specificity] × prevalence)] × prevalence.

FIGURE 237-5. Examples of the use of Bayes' theorem. A, In obese patients with abdominal striae and centripetal obesity, the frequency of Cushing syndrome is ~25%. In testing 100 patients using urinary free cortisol levels, the sensitivity and specificity are each ~95%. The predictive value of elevated urinary free cortisol levels in this setting is 86%; thus, the certainty of diagnosis is increased from 25% to the higher figure by using this one test. B, In screening for primary hyperparathyroidism, ~90% of all serum calcium values in affected persons are above 10.5 mg/dL. Because this is the upper reference limit, 2.5% of normal persons have a serum calcium value above this figure by chance alone. In screening studies, the prevalence of hyperparathyroidism is ~1:1000. The likelihood that a serum calcium value above 10.5 mg/dL represents primary hyperparathyroidism, given only the result of the calcium test, is 3.5%. C, Neonatal hypothyroidism screening programs sometimes use measurements of thyroid-stimulating hormone (TSH) in cord blood for the initial test. Because ~10% of cases are due to hypothalamic-pituitary insufficiency, the sensitivity of such testing is 90%. The cutoff value is >3 standard deviations above the mean value for normal neonates; this excludes 99.75% of normal infants. Because the prevalence of hypothyroidism is ~1:4000 newborns, the predictive value of an elevated TSH level is ~8%, but most affected infants have an elevated TSH level. Thus, when 4000 infants are screened, only 11 infants need further studies to find the one infant with hypothyroidism.

Since the introduction of Bayes' theorem into medical use, it has found increasing application as a model for the diagnostic process, and several reviews highlighting it have appeared in the internal medicine literature.1,57,58 The weaknesses of Bayes' theorem for medical decision making have not been emphasized. The first limitation is its use of a single decision level for predicting the presence of disease. For example, in the evaluation of hypercalcemia, a serum calcium level of 10.6 mg/dL would be treated the same as a calcium level of 12.1 mg/dL or a calcium level of 16.5 mg/dL. Few practicing endocrinologists would view such levels this way. A second obstacle is the difficulty of using Bayes' theorem for more than one variable. Although applications for multiple variables do exist, they are not commonly used, and they further compound the first limitation by looking at only a single decision level for each variable studied. Furthermore, equal emphasis is placed on each result. Physicians are well aware, however, that certain findings are virtually pathognomonic, whereas others are extremely nonspecific. A further drawback lies not in the theorem itself, but in the way it has been used. In screening for rare diseases (e.g., neonatal hypothyroidism) or even in screening for relatively common endocrine disorders (e.g., primary hyperparathyroidism), most abnormal results from screening tests do not indicate the presence of disease. This has led many to criticize the screening of asymptomatic persons as worthless, because most abnormal results are false-positives.17,18 The purpose of screening tests is not to make diagnoses, however, but to identify persons who are at high risk for having a disease. For example, as shown in Figure 237-5, only ~8% of infants with an elevated level of TSH have congenital hypothyroidism. What the screening program does, however, is virtually rule out hypothyroidism in 99.75% of the screened newborns. The task of finding affected infants is now much simpler because further tests can be performed only on the few infants with elevated TSH levels. Without screening, discovering infants with hypothyroidism would be like finding a needle in a haystack; screening puts the needle into a pincushion. Bayes' theorem is an inadequate model for the diagnostic process. Although its popularization has focused attention on the inherent uncertainty in the interpretation of any data, other mathematical models and artificial intelligence systems (e.g., Internist) provide much closer approximations to the probability of diagnoses based on data obtained.59

REFERENCE VALUES IN ENDOCRINOLOGY Reference values should be used as a relative means of comparison, not as an absolute declaration of health or disease. Table 237-1 provides a context for the endocrinologist to interpret results for an individual patient. Reference values are given first in the conventional units used in the United States. The conversion factor to transform these results to SI units and the reference range in SI units also are given. SI refers to the Système International, which is an attempt to create a universally accepted scientific nomenclature. This system is used widely in Europe and in other parts of the world. The SI recommends expressing concentration in moles (or fractions thereof) of a substance per liter, rather than in weight per volume. Young outlined the rationale for making this conversion. In summary, the major advantage of SI units is that they make interrelations between

substances much easier to understand.651 A good example is in the formula for calculated osmolality, in which glucose is divided by 18 and blood urea nitrogen by 2.8 to convert from weight units to moles (osmolality is related to moles of a substance in solution). In this example, expressing concentration in moles makes the calculation easy and eliminates the need for remembering the appropriate factors for conversion. Other advantages of using SI units include a better understanding of relationships between a compound and its target or receptor, and universality in the reporting system, which should improve the international usefulness of scientific studies. The recommendation has been made that peptide hormone concentrations still be expressed in standard mass nomenclature. This decision is difficult to understand, because hormones with active forms of different molecular weights cannot be expressed accurately using such mass nomenclature. Hormone concentrations are reported in mass units; however, because all immunoassays detect molar concentrations of a substance, hormone immunoassays make use of some conversion factor. For the many hormones with multiple circulating forms, the conversion factor may be either an average molecular weight of all forms present, a weighted average, or the molecular weight of the most abundant or most physiologically important form (i.e., the immunoassay standard). Because the exact conversion factors may vary, Table 237-1 gives the conversion factor only for hormones with a single predominant circulating form. The major disadvantage of conversion to SI units is that physicians in the United States would be required to learn a new set of reference ranges for all commonly ordered tests. Canada adopted SI units without major difficulty. In the late 1980s, several journals announced their intention to convert to SI units for all published scientific articles; however, most have returned to the use of conventional units. The reference ranges given in Table 237-1 are those used in the author's laboratory or in the reference laboratories the author uses, employing the method listed first under “Methodologic Considerations.” Reference ranges are for serum or plasma unless otherwise specified. For some tests, reference ranges published in the literature are included, with the reference number in superscript. Some of these reference ranges are based on results from fewer than 20 persons. These ranges may not be applicable to other methods or laboratories, and readers are urged to utilize reference ranges published by the laboratory they most commonly use. For tests in which reference ranges often differ by an order of magnitude (because of the lack of an acceptable standard or differences in reactivity of antibodies), the reference range carries the notation “method dependent.” For each test, the attempt has been made to define the common intraindividual, physiologic, drug-related, and methodologic factors influencing the results for a particular test. For each test, these comments apply to the specimen type (e.g., urine or serum) listed first, for the method given in the “Methodologic Considerations” column unless otherwise specified. Intraindividual variation includes day-to-day variation, within-day variation, and changes occurring during the menstrual cycle. Physiologic changes include commonly encountered alterations in expected physiology, such as exercise, obesity, pregnancy, and the effects of age, sex, and race. The effects of renal insufficiency and alcohol or tobacco use on test results also are given when indicated. This is not a comprehensive list of reference ranges for children; instead, the magnitude and direction of such changes is given. For more information on pediatric reference ranges, the reader should consult pertinent chapters of this book. In addition, two textbooks devoted to pediatric chemistry include endocrine values.652,653 The column for drug effects includes only pharmacologic effects of drugs; cross-reactions are listed under “Methodologic Considerations,” as are any special collection or handling requirements. Several general references and a publication on drug effects are the sources of much of the data in this table.47,48,652,653,654,655,656,657,658 and 659 In addition, specific references for each individual test are listed under the test name. CHAPTER REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38.

Galen RS, et al. Beyond normality—the predictive value and efficiency of medical diagnoses. New York: John Wiley and Sons, 1975. Murphy EA, et al. The normal range—a common misuse. J Chronic Dis 1967; 20:79. Feinstein AR. Clinical biostatistics. XII. On exorcising the ghost of Gauss and the curse of Kelvin. Clin Pharmacol Ther 1971; 12:1003. Grasbeck R, et al. Establishment and use of normal values. Scand J Clin Lab Invest 1969; 110(Suppl):62. Statland BE, et al. Reference values: are they useful? Lab Med 1984; 4:61. Enos WF, et al. Coronary disease among United States soldiers killed in action in Korea. Preliminary report. JAMA 1953; 152:1090 (reprinted in JAMA 1986; 256:2859). Ross R, et al. The pathogenesis of atherosclerosis (first of two parts). N Engl J Med 1976; 295:369. Rifai N. Lipoproteins and apolipoproteins. Composition, metabolism, and association with coronary heart disease. Arch Pathol Lab Med 1986; 110:694. McManus BM. Reference ranges and ideal patient values for blood cholesterol. Can there be reconciliation? Arch Pathol Lab Med 1986; 110:469. Summary of the second report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel II). JAMA 1993; 269:3015. Statland BE, et al. Response of clinical chemistry quantity values to selected physical, dietary, and smoking activities. Prog Clin Pathol 1981; 8:25. Statland BE, et al. Factors contributing to intra-individual variation of serum constituents: 4. Effects of posture and tourniquet application on variation of serum constituents in healthy subjects. Clin Chem 1974; 20:1513. Elveback LR, et al. Health, normality, and the ghost of Gauss. JAMA 1970; 211:69. Reed AH, et al. Influence of statistical method used on the resulting estimate of normal range. Clin Chem 1971; 17:275. Shultz EK, et al. Improved reference-interval estimation. Clin Chem 1985; 31:1974. Feinstein AR. Clinical biostatistics. XXVII. The derangements of the “range of normal.” Clin Pharmacol Ther 1974; 15:528. Cebul RD, et al. Biochemical profiles. Applications in ambulatory screening and preadmission testing of adults. Ann Intern Med 1987; 106:403. Berwick DM. Screening in health fairs. A critical review of benefits, risks, and costs. JAMA 1985; 254:1492. Pickup JF, et al. Intra-individual variation of some serum constituents and its relevance to population-based reference ranges. Clin Chem 1977; 23:842. Van Steirteghem AC, et al. Variance components of serum constituents in healthy individuals. Clin Chem 1978; 24:212. Costongs GMPJ, et al. Short-term and long-term intra-individual variations and critical differences of clinical chemical laboratory parameters. J Clin Chem Clin Biochem 1985; 23:7. Statland BE. Turning lab values into action. Diagn Med 1980; 3:56. Statland BE. Clinical decision levels for lab tests, 2nd ed. Oradell, NJ: Medical Economics Company, 1987. Sheps SB, et al. The assessment of diagnostic tests. A survey of current medical research. JAMA 1984; 252:2418. Feinstein AR. Clinical biostatistics XXXI. On the sensitivity, specificity, and discrimination of diagnostic tests. Clin Pharmacol Ther 1975; 17:104. Beck JR, et al. The use of relative operating characteristic (ROC) curves in test performance evaluation. Arch Pathol Lab Med 1986; 110:13. Winkel P, et al. Using the subject as his own referent in assessing day-to-day changes of laboratory test results. Contemp Top Anal Clin Chem 1977; 1:287. Dillon MC, et al. Diagnostic problem in acute myocardial infarction: CK-MB in the absence of abnormally elevated total creatine kinase levels. Arch Intern Med 1982; 142:33. Hong RA, et al. Elevated CK-MB with normal total creatine kinase in suspected myocardial infarction: associated clinical findings and early prognosis. Am Heart J 1986; 111:1041. Dufour DR, et al. Rapid serial enzyme measurements in evaluation of patients with suspected myocardial infarction. Am J Cardiol 1989; 63:652. Fraser CG. Biological variation in clinical chemistry. An update: collated data, 1988–1991. Arch Pathol Lab Med 1992; 116:916. Holzel WGE. Intra-individual variation of some analytes in serum of patients with insulin-dependent diabetes mellitus. Clin Chem 1987; 33:57. Statland BE, et al. Factors contributing to intra-individual variation of serum constituents. 2. Effects of exercise and diet on variation of serum constituents in healthy subjects. Clin Chem 1973; 19:1380. Levi L. The effect of coffee on the function of the sympatho-adrenomedullary system in man. Acta Med Scand 1967; 181:431. Arendt J, et al., eds. Biological rhythms in clinical practice. Boston: Wright, 1989. Touitou Y, Haus E, eds. Biologic rhythms in clinical and laboratory medicine. Berlin: Springer-Verlag, 1992. Weitzman ED. Circadian rhythms and episodic hormone secretion in man. Annu Rev Med 1976; 27:225. Statland BE, et al. Factors contributing to intra-individual variation of serum constituents. 1. Within-day variation of serum constituents in healthy subjects. Clin Chem 1973; 19:1374.

38a. Leppanen E, et al. When to collect blood specimens: midmorning vs fasting samples. Clin Chem 1998; 44:2537. 39. Nuttall FQ, Jones B. Creatine kinase and glutamic oxalacetic transaminase activity in serum: kinetics of change with exercise and effect of physical conditioning. J Lab Clin Med 1968; 71:847. 39a. De Souza MJ, et al. High frequency of luteal phase deficiency and anovulation in recreational women runners: blunted elevation in follicle-stimulating hormone observed during luteal-follicular transition. J Clin Endocrinol Metab 1998; 83:4220. 40. Tamir I, et al. Measurement of lipids and evaluation of lipid disorders. In: Henry JB, ed. Clinical diagnosis and management by laboratory methods, 16th ed. Philadelphia: WB Saunders, 1979:201. 41. Van der Poll T, et al. Tumor necrosis factor: a putative mediator of the sick euthyroid syndrome in man. J Clin Endocrinol Metab 1990; 71:1567. 42. Bhakri HL, et al. Longitudinal study of thyroid function in acutely ill elderly patients using a sensitive TSH assay-defer testing until recovery. Gerontology 1990; 36:140. 43. Weindling H, et al. Drug interactions and clinical laboratory data. Lab Med 1975; 6:24. 44. Fortman CS, et al. Serum 5'-nucleotidase in patients receiving antiepileptic drugs. Am J Clin Pathol 1985; 84:197. 45. Blumberg AG, et al. The interference of chlorpromazine metabolites in the analysis of urinary methoxy-catecholamines. Clin Chem 1966; 12:803. 46. Kroll MH, et al. Reaction of picrate with creatinine and cepha antibiotics. Clin Chem 1984; 30:1664. 47. Young DS, et al. Effects of drugs on clinical laboratory tests, 4th ed. Washington: AACC Press, 1995. 48. Siest G, et al., eds. Drug effects on laboratory test results: analytical interferences and pharmacological effects. Littleton, MA: PSG Publishing, 1988. 49. Fisken RA, et al. Hypercalcaemia—a hospital survey. Q J Med 1980; 49:405. 50. Laessig RH, et al. The effects of 0.1 and 1.0 per cent erythrocytes and hemolysis on serum chemistry values. Am J Clin Pathol 1976; 66:639. 51. Laessig RH, et al. Changes in serum chemical values as a result of prolonged contact with the clot. Am J Clin Pathol 1976; 66:598. 52. Sazama K, et al. Is antiglycolysis required for routine glucose analysis? Clin Chem 1979; 25:2038. 53. Ingram RH, et al. Pseudohyperkalemia with thrombocytosis. N Engl J Med 1962; 267:895. 54. Dufour DR, et al. Artifactual hyperkalemia due to heparin in patients with extreme leukocytosis. Clin Chem 1987; 33:914. 55. Nicar MJ. Specimen processing and renin activity in plasma. Clin Chem 1992; 38:598. 56. Grannis GF, et al. Proficiency evaluation of clinical chemistry laboratories. Clin Chem 1972; 18:222. 57. Griner PF, et al. Selection and interpretation of diagnostic tests and procedures. Principles and applications. Ann Intern Med 1981; 94:557. 58. Sox HA. Probability theory in the use of diagnostic tests. An introduction to critical study of the literature. Ann Intern Med 1986; 104:60.

59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150. 151. 152. 153. 154. 155. 156. 157. 158. 159. 160. 161.

Myers JD. The computer as a diagnostic consultant, with emphasis on use of laboratory data. Clin Chem 1986; 32:1714. Hansen JL, et al. Direct assays of lactate, pyruvate, beta-hydroxybutyrate, and acetoacetate with a centrifugal analyzer. Clin Chem 1978; 24:475. Hall SEH, et al. Ketone body kinetics in humans: the effects of insulin-dependent diabetes, obesity, and starvation. J Lipid Res 1984; 25:1184. Brenckman WD, et al. Unpredictable fluctuations in serum acid phosphatase activity in prostatic cancer. JAMA 1981; 245:2501. Carson JL, et al. Diagnostic accuracy of four assays of prostatic acid phosphatase. Comparison using receiver operating characteristic curve analysis. JAMA 1985; 253:665. Benvenuti M, et al. Circadian rhythm in prostatic acid phosphatase (PAP): a potential tumor marker rhythm in prostatic cancer (Pca). Chronobiologia 1983; 10:383. Ohno Y, et al. Change of peripheral levels of pituitary hormones and cytokines after injection of interferon (IFN)-beta in patients with chronic hepatitis C. J Clin Endocrinol Metab 1998; 83:3681. Yalow RS, et al. Change of peripheral levels of pituitary hormones and cytokines after injection of interferon (IFN)-beta in patients with chronic hepatitis C. J Clin Endocrinol Metab 1973; 36:415. Schoneshofer M, et al. Corticotropin in human plasma. General considerations. Surv Immunol Res 1984; 3:55. Farrell PA, et al. Plasma adrenocorticotropin and cortisol responses to submaximal and exhaustive exercise. J Appl Physiol 1983; 55:1441. Laakman G, et al. Effects of psychotropic drugs (desimipramine, chlorimipramine, sulpiride and diazepam) on the human HPA axis. Psychopharmacology (Berl) 1984; 84:66. Laue L, et al. Effect of chronic treatment with the glucocorticoid antagonist RU 486 in man: toxicity, immunological, and hormonal aspects. J Clin Endocrinol Metab 1990; 71:1474. Venturoli S, et al. Ketoconazole therapy for women with acne and/or hirsutism. J Clin Endocrinol Metab 1990; 71:335. Cugini P, et al. Toward a chronophysiology of circulating aldosterone. Biochem Med 1984; 32:270. Saruta T, et al. Mechanism of age-related changes in renin and adrenocortical steroids. J Am Geriatr Soc 1980; 28:210. Sancho JM, et al. Interference by ranitidine with aldosterone secretion in vivo. Eur J Clin Pharmacol 1984; 27:495. Zipser RD, et al. Hyperreninemic hypoaldosteronism in the critically ill: a new entity. J Clin Endocrinol Metab 1981; 53:867. Fiselier TJW, et al. Levels of renin, angiotensin I and II, angiotensinconverting enzyme and aldosterone in infancy childhood. Eur J Pediatr 1983; 141:3. Saito I, et al. Effect of a calcium entry blocker on blood pressure, plasma renin activity, aldosterone and catecholamines in normotensive subjects. Clin Endocrinol (Oxf) 1986; 24:565. Crofton PM. Biochemistry of alkaline phosphatase isoenzymes. Crit Rev Clin Lab Sci 1982; 16:161. Broulik PD, et al. Alterations in human serum alkaline phosphatase and its bone isoenzyme by chronic administration of lithium. Clin Chim Acta 1984; 140:151. Schoenau E, et al. “Fragmented” isoenzymes of alkaline phosphatase in the diagnosis of transient hyperphosphatasemia. Clin Chem 1986; 32:2211. Stein P, et al. Transient hyperphosphatasemia of infancy and early childhood: clinical and biochemical features of 21 cases and literature review. Clin Chem 1987; 33:313. Gordon T. Factors associated with serum alkaline phosphatase level. Arch Pathol Lab Med 1993; 117:187. Schiele F, et al. Total bone and liver alkaline phosphatases in plasma: biological variations and reference limits. Clin Chem 1983; 29:634. Rosalki SB, et al. Two new methods for separating and quantifying bone and liver alkaline phosphatase isoenzymes in plasma. Clin Chem 1984; 30:1182. Valenzuela GJ, et al. Time-dependent changes in bone, placental, intestinal, and hepatic alkaline phosphatase activities in serum during human pregnancy. Clin Chem 1987; 33:1801. Hannon R, et al. Response of biochemical markers of bone turnover to hormone replacement therapy: impact of biological variability. J Bone Miner Res 1998; 13:1124. Withold W. Monitoring of bone turnover biological, preanalytical and technical criteria in the assessment of biochemical markers. Eur J Clin Chem Clin Biochem 1996; 34:785. Kyd PA, et al. Clinical usefulness of bone alkaline phosphatase in osteoporosis. Ann Clin Biochem 1998; 35:717. Romagnoli E, et al. Assessment of serum total and bone alkaline phosphatase measurement in clinical practice. Clin Chem Lab Med 1998; 36:163. Douglas AS, et al. Seasonal differences in biochemical parameters of bone remodelling. J Clin Pathol 1996; 49:284. Lau KH, et al. Phenytoin increases markers of osteogenesis for the human species in vitro and in vivo. J Clin Endocrinol Metab 1995; 80:2347. Karlsson MK, et al. Indicators of bone formation in weight lifters. Calcif Tissue Int 1995; 56:177. Panteghini M, et al. Biological variation in bone-derived biochemical markers in serum. Scand J Clin Lab Invest 1995; 55:609. Medri G, et al. Pituitary glycoprotein hormones in chronic renal failure: evidence for an uncontrolled alpha-subunit release. J Endocrinol Invest 1993; 16:169. Oppenheim DS, et al. Effects of chronic GnRH analogue administration on gonadotrophin and alpha-subunit secretion in post-menopausal women. Clin Endocrinol (Oxf) 1992; 36:559. Madersbacher S, et al. Serum glycoprotein hormones and their free alpha-subunit in a healthy elderly population selected according to the SENIEUR protocol. Analyses with ultrasensitive time resolved fluoroimmunoassays. Mech Ageing Dev 1993; 71:223. Taylor A, et al. Determination of aluminium in serum: findings of an external quality assessment scheme. Ann Clin Biochem 1985; 22:351. Kostyniak PJ. An electrothermal atomic absorption method for aluminum analysis in plasma: identification of sources of contamination in blood sampling procedures. J Anal Toxicol 1983; 7:20. De Broe ME, et al. Correlation of serum aluminum values with tissue aluminum concentration. Contrib Nephrol 1984; 38:37. Cannata JB, et al. Effect of acute aluminium overload on calcium and parathyroid-hormone metabolism. Lancet 1983; 1:501. Morimoto I, et al. Studies on the origin of androstanediol and androstanediol glucuronide in young and elderly men. J Clin Endocrinol Metab 1981; 52:772. Greep N, et al. Androstanediol glucuronide plasma clearance and production rates in normal and hirsute women. J Clin Endocrinol Metab 1986; 62:22. Lookingbill DP, et al. Clinical and biochemical parameters of androgen action in normal healthy Caucasian versus Chinese subjects. J Clin Endocrinol Metab 1991; 72:1242. Gray A, et al. Age, disease, and changing sex hormone levels in middle-aged men: results of the Massachusetts Male Aging Study. J Clin Endocrinol Metab 1991; 73:1016. Andre M, et al. Reference values for androstanediol glucuronide. Clin Chem 1994; 40:162. Brody S, et al. Serum levels of 4-androstene-3,17-dione in menstruating and postmenopausal women. Evaluation of a radioimmunoassay and correlation with bone mineral content and endometrial pathology. Acta Obstet Gynecol Scand 1983; 62:531. Hummer L, et al. An easy and reliable radioimmunoassay of serum androstenedione: age-related normal values in 252 females aged 2 to 70 years. Scand J Clin Lab Invest 1983; 43:301. Siklosi G, et al. Episodic secretion of hormones and the diagnostic value of single blood estimates. III. Testosterone, androstenedione, dehydroepiandrosterone, dehydroepiandrosterone sulphate, cortisol. Acta Med Hung 1984; 41:213. Long JP, et al. Prolactin and the hypothalamic-pituitary-testicular axis in cimetidine-treated men. Ir Med J 1985; 78:48. Goldman J, et al. Contrast analysis for the evaluation of the circadian rhythms of plasma cortisol, androstenedione, and testosterone in normal men and the possible influence of meals. J Clin Endocrinol Metab 1985; 60:164. Luppa P, et al. Serum androgens in intensivecare patients: correlations with clinical findings. Clin Endocrinol (Oxf) 1991; 34:305. Workman RJ, et al. Circulating levels of angiotensin I measured by radioimmunoassay in hypertensive subjects. J Lab Clin Med 1979; 93:847. Matthews PG, et al. Hormonal changes with long-term converting-enzyme inhibition by captopril in essential hypertension. Clin Sci 1979; 57(Suppl 5):135s. Nussberger J, et al. True versus immunoreactive angiotensin II in human plasma. Hypertension 1985; 7(3 Pt 2):I1. Crum R, et al. Neuroendocrinology of chronic renal failure and renal transplantation. Transplantation 1991; 52:818. Studdy PR, et al. Angiotensin-converting enzyme and its clinical significance—a review. J Clin Pathol 1983; 36:938. Smallridge RC, et al. Serum angiotensin-converting enzyme. Alterations in hyperthyroidism, hypothyroidism, and subacute thyroiditis. JAMA 1983; 250:2489. Rasmussen AB, et al. The influence of normotensive pregnancy and pre-eclampsia on angiotensin-converting enzyme. Acta Obstet Gynecol Scand 1983; 62:341. Pietila K, et al. Increase of serum angiotensin-converting enzyme activity after freezing. Scand J Clin Lab Invest 1984; 44:453. Thompson PJ, et al. Angiotensin-converting enzyme. Investigation of diurnal variation, the effect of a large dose of prednisolone, and prednisolone pharmacokinetics in patients with sarcoidosis. Am Rev Respir Dis 1986; 134:1075. Inukai T, et al. Clinical characteristics of patients with the initial diagnosis of NIDDM with positivity for antibodies to glutamic acid decarboxylase. Exp Clin Endocrinol Diabetes 1997; 105:327. Zimmet PZ, et al. The ethnic distribution of antibodies to glutamic acid decarboxylase: presence and levels of insulin-dependent diabetes mellitus in Europid and Asian subjects. J Diabetes Complications 1993; 7:1. Tuomi T, et al. Antibodies to glutamic acid decarboxylase reveal latent autoimmune diabetes mellitus in adults with a non-insulindependent onset of disease. Diabetes 1993; 42:359. Schmidli RS, et al. Disease sensitivity and specificity of 52 assays for glutamic acid decarboxylase antibodies. The Second International GADAB Workshop. Diabetes 1995; 44:636. Hermitte L, et al. Diverging evolution of anti-GAD and anti-IA-2 antibodies in long-standing diabetes mellitus as a function of age at onset: no association with complications. Diabet Med 1998; 15:586. Imagawa A, et al. High prevalence of antibodies to glutamic acid decarboxylase in comparison to islet cell antibodies in patients with long-standing insulindependent diabetes mellitus. Res Commun Mol Pathol Pharmacol 1996; 92:43. Betterle C, et al. II. Adrenal cortex and steroid 21-hydroxylase autoantibodies in children with organ-specific autoimmune diseases: markers of high progression to clinical Addison's disease. J Clin Endocrinol Metab 1997; 82:939. Falorni A, et al. 21-hydroxylase autoantibodies in adult patients with endocrine autoimmune diseases are highly specific for Addison's disease. Belgian Diabetes Registry. Clin Exp Immunol 1997; 107:341. Laureti S, et al. Levels of adrenocortical autoantibodies correlate with the degree of adrenal dysfunction in subjects with preclinical Addison's disease. J Clin Endocrinol Metab 1998; 83:3507. Betterle C, et al. I. Adrenal cortex and steroid 21-hydroxylase autoantibodies in adult patients with organ-specific autoimmune diseases: markers of low progression to clinical Addison's disease. J Clin Endocrinol Metab 1997; 82:932. Hawa M, et al. Value of antibodies to islet protein tyrosine phosphataselike molecule in predicting type 1 diabetes. Diabetes 1997; 46:1270. Savola K, et al. IA-2 antibodies—a sensitive marker of IDDM with clinical onset in childhood and adolescence. Childhood Diabetes in Finland Study Group. Diabetologia 1998; 41:424. Bingley PJ. Interactions of age, islet cell antibodies, insulin autoantibodies, and first-phase insulin response in predicting risk of progression to IDDM in ICA+ relatives: the ICARUS data set. Islet Cell Antibody Register Users Study. Diabetes 1996; 45:1720. Baron EJ, et al. Lack of agreement among two commercial enzyme-linked immunosorbent antibody assays and a conventional immunofluorescence-based method for detecting islet cell autoantibodies. Clin Diagn Lab Immunol 1996; 3:429. Feldt-Rasmussen U. Serum thyroglobulin and thyroglobulin autoantibodies in thyroid diseases. Pathogenic and diagnostic aspects. Allergy 1983; 38:369. Wilson R, et al. Thyroid antibody titer and avidity in patients with recurrent miscarriage. Fertil Steril 1999; 71:558. Sundbeck G, et al. Prevalence of serum antithyroid peroxidase antibodies in 85-year-old women and men. Clin Chem 1995; 41:707. Ogawa T, et al. Thyroid hormone autoantibodies in patients with Graves' disease: effect of anti-thyroid drug treatment. Clin Chim Acta 1994; 228:113. Tyroler HA, et al. Apolipoprotein A-I, A-II and C-II in black and white residents of Evans County. Circulation 1980; 62:249. Luoma PV, et al. Plasma high-density lipoproteins and hepatic microsomal enzyme induction. Relation to histological changes in the liver. Eur J Clin Pharmacol 1982; 23:275. Matuchansky C, et al. Effects of cyclic (nocturnal) total parenteral nutrition and continuous enteral nutrition on circadian rhythms of blood lipids, lipo-proteins and apolipoproteins in humans. Am J Clin Nutr 1985; 41:727. Zeithofer J, et al. Changes of serum lipid patterns during long-term anticonvulsive treatment. Clin Investig 1993; 71:574. Walsh BW, et al. Effects of postmenopausal estrogen replacement on the concentrations and metabolism of plasma lipoproteins. N Engl J Med 1991; 325:1196. Piechota W, et al. Reference ranges of lipids and apolipoproteins in pregnancy. Eur J Obstet Gynecol Reprod Biol 1992; 45:27. Rabe T, et al. Lipid metabolism in Norplant-2 users—a two-year follow-up study. Total cholesterol, triglycerides, lipoproteins and apolipoproteins. Contraception 1992; 45:21. Lussier-Cacan S, et al. Cyclic fluctuations in human serum lipid and apolipo-protein levels during the normal menstrual cycle: comparison with changes occurring during oral contraceptive therapy. Metabolism 1991; 40:849. Fulton-Kehoe DL, et al. Determinants of total high density lipoprotein cholesterol and high density lipoprotein subfraction levels among Hispanic and non-Hispanic white persons with normal glucose tolerance: the San Luis Valley Diabetes Study. J Clin Epidemiol 1992; 45:1191. Naito HK. The clinical significance of apolipoprotein measurements. J Clin Immunoassay 1986; 9:120. Cernacek P, et al. Atrial natriuretic peptide: blood levels in human disease and their measurement. Clin Biochem 1988; 21:5. Richards AM, et al. Diurnal change in plasma atrial natriuretic peptide concentrations. Clin Sci 1987; 73:489. Lijnen P, et al. Effects of haemolysis and prolonged cold storage of human plasma on the alpha-atrial natriuretic peptide concentration. Clin Chim Acta 1988; 171:333. Hartter E, et al. Circadian variation and age dependence of human atrial natriuretic peptide levels in hospitalized patients. Horm Metab Res 1987; 19:490. Rascher W, et al. Atrial natriuretic peptide in infants and children. Horm Res 1987; 28:58. Saxenhofer H, et al. Corticosteroid-induced stimulation of atrial natriuretic peptide in man. Acta Endocrinol (Copenh) 1988; 118:179. Sanfield JA, et al. Epinephrine increases plasma immunoreactive atrial natriuretic hormone levels in humans. Am J Physiol 1987; 252:E740. Tan AC, et al. Atrial natriuretic peptide—the influence of various physiological and sampling conditions. Ann Clin Biochem 1987; 24:500. Mann FE, et al. Effect of angiotensin I converting enzyme inhibition on circulating atrial natriuretic peptide in humans. Klin Wochenschr 1986; 64(Suppl 5):13. Colantonio D, et al. A possible role of atrial natriuretic peptide in ethanol-induced acute diuresis. Life Sci 1991; 48:635. Harrison SP. Naproxen interference with the ion-selective electrode in the RA-1000. Clin Chem 1987; 33:421. Bernard A, et al. Cadmium in human population. Experientia 1984; 40:143. Brockhaus A, et al. Levels of cadmium and lead in blood in relation to smoking, sex, occupation, and other factors in an adult population of the FRG. Int Arch Occup Environ Health 1983; 52:167.

162. Motté P, et al. Construction and clinical validation of a sensitive and specific assay for serum mature calcitonin using monoclonal antipeptide antibodies. Clin Chim Acta 1988; 174:35. 163. Ghillani PP, et al. Identification and measurement of calcitonin precursors in serum of patients with malignant diseases. Cancer Res 1989; 49:6845. 164. Born W, et al. Diagnostic relevance of the amino-terminal cleavage peptide of procalcitonin (PAS-57), calcitonin and calcitonin gene-related peptide in medullary thyroid carcinoma patients. Regulatory Peptides 1991; 32:311. 165. Snider RH Jr, et al. Procalcitonin and its component peptides in systemic inflammation: immunochemical characterization. J Investig Med 1997; 45:552. 166. Whang K, et al. Serum calcitonin precursors in sepsis and systemic infection. J Clin Endocrinol Metab 1998; 83:3296. 167. Snider RH Jr, et al. Radioimmunoassay of calcitonin in normal human urine. Anal Chem 1978; 50:449. 168. Silva OL, et al. Urine calcitonin as a test for medullary thyroid cancer: a new screening procedure. Ann Surg 1979; 189:269. 169. Wang Y, et al. A simple high performance liquid chromatography assay for simultaneous determination of plasma norepinephrine, epinephrine, dopamine, and 3,4-dihydroxyphenyl acetic acid. J Pharm Biomed Anal 1999; 21:519. 170. Roberts WL, et al. Comparison of four commercial urinary albumin (microalbumin) methods: implications for detecting diabetic nephropathy using random urine specimens. Clin Chim Acta 1998; 273:21. 171. Khuu HM, et al. Evaluation of a fully automated high-performance liquid chromatography assay for hemoglobin A1c. Arch Pathol Lab Med 1999; 123:763. 172. Nagi DK, et al. Hyperinsulinemia in nondiabetic Asian subjects using specific assays for insulin, intact proinsulin, and des-31, 32-proinsulin. Diabetes Care 1996; 19:39. 173. Nexo E, et al. Evaluation of novel assays in clinical chemistry: quantification of plasma total homocysteine. Clin Chem 2000; 46:1150. 174. Kausz AT, et al. Screening plasma aluminum levels in relation to aluminum bone disease among asymptomatic dialysis patients. Am J Kidney Dis 1999; 34:688. 175. Bowers GN, et al. Measurement of ionized calcium in serum with ion-selective electrodes: a mature technology that can meet the daily service needs. Clin Chem 1986; 32:1437. 176. Payne RB. Clinically significant effect of protein concentration on ion-selective electrode measurements of ionised calcium. Ann Clin Biochem 1982; 19:233. 177. Boink AB, et al. Recommendation on sampling, transport, and storage for the determination of the concentration of ionized calcium in whole blood, plasma, and serum. IFC Scientific Division, Working Group on Ion-Selective Electrodes (WGSE). J Int Fed Clin Chem 1992; 4:147. 178. Zaloga GP, et al. Assessment of calcium homeostasis in the critically ill surgical patient. The diagnostic pitfalls of the McLean-Hastings nomogram. Ann Surg 1985; 202:587. 179. Packer E, et al. Effects of estrogen on daylong circulating calcium, phosphorus, 1,25-dihydroxyvitamin D, and parathyroid hormone in postmenopausal women. J Bone Miner Res 1990; 5:877. 180. Jacobson MA, et al. Foscarnet-induced hypocalcemia and effects of foscarnet on calcium metabolism. J Clin Endocrinol Metab 1991; 72:1130. 181. Fredstorp L, et al. The short and long-term effects of octreotide on calcium homeostasis in patients with acromegaly. Clin Endocrinol (Oxf) 1993; 39:331. 182. Burritt MF, et al. Comparative studies of total and ionized serum calcium values in normal subjects and patients with renal disorders. Mayo Clin Proc 1980; 55:606. 183. Wu JT. Interference of heparin in carcinoembryonic antigen radioimmunoassays. Clin Chim Acta 1983; 130:47. 184. Lokich JJ, et al. Criteria for monitoring carcinoembryonic antigen: variability of sequential assays at elevated levels. J Clin Oncol 1984; 2:181. 185. Fletcher RH. Carcinoembryonic antigen. Ann Intern Med 1986; 104:66. 186. Witter FR, et al. Folate, carotene, and smoking. Am J Obstet Gynecol 1982; 144:857. 187. Vuilleumier JP, et al. Clinical chemical methods for the routine assessment of the vitamin status in human populations. Part I: The fat-coluble vitamins A and E, and beta-carotene. Int J Vitam Nutr Res 1983; 53:265. 188. Nierenberg DW, et al. Diurnal variation in plasma levels of retinol, tocopherol, and beta-carotene. Am J Med Sci 1987; 294:187. 189. Probstfield JL, et al. Carotenoids and vitamin A: the effect of hypocholesterolemic agents on serum levels. Metabolism 1985; 34:88. 190. Leung AK, et al. Serum carotene concentrations in normal infants and children. Clin Pediatr (Phila) 1990; 29:575. 191. Grossman E, et al. Diet and weight loss: their effect on norepinephrine renin and aldosterone levels. Int J Obes 1985; 9:107. 192. Moyer TP, et al. Analysis for urinary catecholamines by liquid chromatography with amperometric detection: methodology and clinical interpretation of results. Clin Chem 1979; 25:256. 193. Hjemdahl P. Inter-laboratory comparison of plasma catecholamine determinations using several different assays. Acta Physiol Scand Suppl 1984; 527:43. 194. Dunne JW, et al. The effect of ascorbic acid on plasma sulfate conjugated catecholamines after eating bananas. Life Sci 1983; 33:1511. 195. Zuspan FP, et al. The effect of smoking and oral contraceptives on the urinary excretion of epinephrine and norepinephrine. Am J Obstet Gynecol 1979; 135:1012. 196. Liebau H, et al. Diurnal and daily variations of PRA, plasma catecholamines and blood pressure in normotensive and hypertensive man. Contrib Nephrol 1982; 30:57. 197. Hoehe M, et al. Opiates increase plasma catecholamines in humans. Psychoneuroendocrinology 1993; 18:141. 198. Fitzgibbon M, et al. Reference values for urinary HMMA, HVA, noradrenaline, adrenaline, and dopamine excretion in children using random urine samples and HPLC with electrochemical detection. Ann Clin Biochem 1992; 29:400. 199. Wolf PL. Ceruloplasmin: methods and clinical use. Crit Rev Clin Lab Sci 1982; 17:229. 200. Galdston M, et al. Ceruloplasmin. Increased serum concentration and impaired antioxidant activity in cigarette smokers, and ability to prevent suppression of elastase inhibitory capacity of alpha 1-proteinase inhibitor. Am Rev Respir Dis 1984; 129:258. 201. Hopman WPM, et al. Plasma cholecystokinin response to a liquid fat meal in vagotomized patients. Ann Surg 1984; 200:693. 202. Rehfield JF. How to measure cholecystokinin in plasma? Gastroenterology 1984; 87:434. 203. Nakano I, et al. High plasma cholecystokinin response following ingestion of test meal by patients with non-insulin dependent diabetes mellitus. Regul Pept 1986; 14:229. 204. Radberg G, et al. Cholecystokinin secretion in pregnancy. Scand J Gastroenterol 1987; 22:687. 205. Maton PN, et al. Atropine inhibits meal-stimulated release of cholecystokinin. Scand J Gastroenterol 1984; 19:831. 206. Frick G, et al. Plasma levels of cholecystokinin and gastrin during the menstrual cycle and pregnancy. Acta Obstet Gynecol Scand 1990; 69:317. 207. Douglas BR, et al. Coffee stimulation of cholecystokinin release and gallbladder contraction in humans. Am J Clin Nutr 1990; 52:553. 208. Glasbrenner B, et al. Relationship between postprandial release of CCK and PP in health and in chronic pancreatitis. Regul Pept 1994; 50:45. 209. Blank DW, et al. The method of determination must be considered in interpreting blood cholesterol levels. JAMA 1986; 256:2867. 210. Friedlander Y, et al. Variability of plasma lipids and lipoproteins: the Jerusalem Lipid Research Clinic Study. Clin Chem 1985; 31:1121. 211. Hussa RO, et al. Discordant human chorionic gonadotropin results: causes and solutions. Obstet Gynecol 1985; 65:211. 212. Ayala AR, et al. Daily rhythm of serum human chorionic gonadotropin and human chorionic somatomammotropin in normal pregnancy. Int J Gynaecol Obstet 1984; 22:173. 213. Haning RV, et al. Effects of fetal sex and dexamethasone on preterm maternal serum concentrations of human chorionic gonadotropin, progesterone, estrone, estradiol, and estriol. Am J Obstet Gynecol 1989; 161:1549. 214. Versieck J, et al. Normal levels of trace elements in human blood, plasma, or serum. Anal Chim Acta 1980; 116:217. 215. Ong CN, et al. Concentrations of heavy metals in maternal and umbilical cord blood. Biometals 1993; 6:61. 216. Versieck J. Trace elements in human body fluids and tissues. Crit Rev Clin Lab Sci 1985; 22:96. 217. Hennig FF, et al. Nickel-, chrom- and cobalt-concentrations in human tissue and body fluids of hip prosthesis patients. J Trace Elem Electrolytes Health Dis 1992; 6:239. 218. Guillard O, et al. Diurnal variations of zinc, copper and magnesium in the serum of normal fasting adults. Biomedicine 1979; 31:193. 219. Kirsten GF, et al. Serum zinc and copper levels in the 1st year of life. S Afr Med J 1985; 67:414. 220. Helgeland K, et al. Copper and zinc in human serum in Norway. Relationship to geography, sex and age. Scand J Clin Lab Invest 1982; 42:35. 221. Wu C-T, et al. Serum zinc, copper, and ceruloplasmin levels in male alcoholics. Biol Psychiatry 1984; 19:1333. 222. Ghose K, et al. Hypercupraemia induced by antiepileptic drugs. Hum Toxicol 1983; 2:519. 223. Gallagher SK, et al. Short-term and long-term variability of indices related to nutritional status. I: Ca, Cu, Fe, Mg, and Zn. Clin Chem 1989; 35:369. 224. Seckl JR, et al. Diurnal variation of plasma corticosterone in depression. Psychoneuroendocrinology 1990; 15:485. 225. Feldman D. Ketoconazole and other imidazole derivatives as inhibitors of steroidogenesis. Endocr Rev 1986; 7:409. 226. Becker DJ. The endocrine responses to protein calorie malnutrition. Annu Rev Nutr 1983; 3:187. 227. Wilkins JN, et al. Nicotine from cigarette smoking increases circulating levels of cortisol, growth hormone, and prolactin in male chronic smokers. Psychopharmacology (Berl) 1982; 78:305. 228. Epstein MT, et al. Licorice raises urinary cortisol in man. J Clin Endocrinol Metab 1978; 47:397. 229. Halbreich U, et al. Effect of age and sex on cortisol secretion in depressives and normals. Psychiatry Res 1984; 13:221. 230. Ishizuka B, et al. Pituitary hormone release in response to food ingestion: evidence for neuroendocrine signals from gut to brain. J Clin Endocrinol Metab 1983; 57:1111. 231. Gram LF, et al. Suppression of plasma cortisol after oral administration of oxazepam in man. Br J Clin Pharmacol 1984; 17:176. 232. Wilke TJ, et al. The evaluation of five commercial serum cortisol kits regarding precision and accuracy. Clin Biochem 1984; 17:311. 233. Perini GI, et al. Effects of carbamazepine on pituitary-adrenal function in healthy volunteers. J Clin Endocrinol Metab 1992; 74:406. 234. Loprinzi CL, et al. Effect of megestrol acetate on the human pituitary-adrenal axis. Mayo Clin Proc 1992; 67:1160. 235. Osei K, et al. Decreased serum C-peptide/insulin molar ratios after oral glucose ingestion in hyperthyroid patients. Diabetes Care 1984; 7:471. 236. Boyko EJ, et al. Higher insulin and C-peptide concentrations in Hispanic population at high risk for NIDDM. San Luis Valley Diabetes Study. Diabetes 1991; 40:509. 237. Meistas M, et al. Correlation of urinary excretion of C-peptide with the integrated concentration and secretion rate of insulin. Diabetes 1981; 30:639. 238. Licata AA. Acute effects of increased meat protein on urinary electrolytes and cyclic adenosine monophosphate and serum parathyroid hormone. Am J Clin Nutr 1981; 34:1779. 239. Markianos M, et al. Effects of acute cannabis use on urinary neurotransmitter metabolites and cyclic nucleotides in man. Drug Alcohol Depend 1984; 14:175. 240. Gogel E, et al. Probenecid inhibits the secretion of nephrogenous adenosine 3',5'-monophosphate in normal man. J Clin Endocrinol Metab 1983; 57:689. 241. Kruse K, Kracht U. Urinary adenosine 3',5'-monophosphate excretion in childhood. J Clin Endocrinol Metab 1981; 53:1251. 242. Colin AA, et al. Effects of theophylline on urinary excretion of cyclic AMP, calcium, and phosphorus in normal subjects. Miner Electrolyte Metab 1984; 10:359. 243. Gennari C, et al. Seasonal variation in urinary excretion of cyclic AMP in healthy people. J Endocrinol Invest 1981; 4:323. 244. Sirinathsinghji DJS, et al. Concentration patterns of plasma dehydroepiandrosterone, delta 5-androstenediol and their sulphates, testosterone and cortisol in normal healthy women and in women with anorexia nervosa. Acta Endocrinol (Copenh) 1985; 108:255. 245. Parker LN, et al. Control of adrenal androgen secretion. Endocr Rev 1980; 1:392. 246. Levesque LA, et al. The effect of phenytoin and carbamazepine on serum dehydroepiandrosterone sulfate in men and women who have partial seizures with temporal lobe involvement. J Clin Endocrinol Metab 1986; 63:243. 247. Musey VC, et al. Long term effects of a first pregnancy on the hormonal environment: estrogens and androgens. J Clin Endocrinol Metab 1987; 64:111. 248. Mitchell E, et al. Evidence for an association between dehydroepiandrosterone sulfate and nonfatal, premature myocardial infarction in males. Circulation 1994; 89:89. 249. Brown RD, et al. Plasma deoxycorticosterone in normal and abnormal pregnancy. J Clin Endocrinol Metab 1972; 35:736. 250. Antonipillai I, et al. The origin of plasma deoxycorticosterone in men and in women during the menstrual cycle. J Clin Endocrinol Metab 1983; 56:93. 251. Kuwabara Y, et al. Plasma 11-deoxycortisol in normal and abnormal human pregnancy. Am J Obstet Gynecol 1983; 147:766. 252. Hosoda H, et al. The specificity of enzyme immunoassays for plasma 11-deoxycortisol. Chem Pharm Bull (Tokyo) 1983; 31:3595. 253. Drafta D, et al. Age-related changes of plasma steroids in normal adult males. J Steroid Biochem 1982; 17:683. 254. Rittmaster RS, et al. Effect of finasteride, a 5 alpha-reductase inhibitor, on serum gonadotropins in normal men. J Clin Endocrinol Metab 1992; 75:484. 255. Genazzani AR, et al. Proopiocortin-related peptide plasma levels throughout prepuberty and puberty. J Clin Endocrinol Metab 1983; 57:56. 256. Petraglia F, et al. Simultaneous circadian variations of plasma ACTH, beta-lipotropin, beta-endorphin and cortisol. Horm Res 1983; 17:147. 257. Matsumura M, et al. Effect of a test meal, duodenal acidification, and tetragastrin on the plasma concentration of beta-endorphin-like immunoreactivity in man. Regul Pept 1982; 4:173. 258. Laatikainen T, et al. Plasma beta-endorphin and the menstrual cycle. Fertil Steril 1985; 44:206. 259. Hoffman DI, et al. Plasma beta-endorphin concentrations prior to and during pregnancy, in labor, and after delivery. Am J Obstet Gynecol 1984; 150:492. 260. Aronin N, et al. Plasma immunoreactive beta-endorphin is elevated in uraemia. Clin Endocrinol (Oxf) 1983; 18:459. 261. Panerai AE, et al. Plasma beta-endorphin, beta-lipotropin, and metenkephalin concentrations during pregnancy in normal and drug-addicted women and their newborn. J Clin Endocrinol Metab 1983; 57:537. 262. Howlett TA, et al. Release of beta endorphin and metenkephalin during exercise in normal women: response to training. BMJ 1984; 288:1950. 263. Panerai AE, et al. Plasma beta-endorphin and metenkephalin in physiological and pathological conditions. Adv Biochem Psychopharmacol 1982; 33:139. 264. Rahkila P, Laatikainen T. Effect of oral contraceptives on plasma beta-endorphin and corticotropin at rest and during exercise. Gynecol Endocrinol 1992; 6:163. 265. Govoni S, et al. Immunoreactive metenkephalin plasma concentrations in chronic alcoholics and in children born from alcoholic mothers. Life Sci 1983; 33:1581. 266. Cotes PM, et al. Changes in serum immunoreactive erythropoietin during the menstrual cycle and normal pregnancy. Br J Obstet Gynaecol 1983; 90:304. 267. Pratt MC, et al. Effect of angiotensin converting enzyme inhibitors on erythropoietin concentrations in healthy volunteers. Br J Clin Pharmacol 1992; 34:363. 268. Schwandt HJ, et al. Influence of prolonged physical exercise on the erythropoietin concentration in blood. Eur J Appl Physiol 1991; 63:463. 269. Cahan C, et al. Diurnal variations in serum erythropoietin levels in healthy subjects and sleep apnea patients. J Appl Physiol 1992; 72:2112.

270. Martikainen H, et al. Circannual concentrations of melatonin, gonadotrophins, prolactin and gonadal steroids in males in a geographical area with a large annual variation in daylight. Acta Endocrinol (Copenh) 1985; 109:446. 271. Klaiber EL, et al. Serum estradiol levels in male cigarette smokers. Am J Med 1984; 77:858. 272. Boyden TW, et al. Sex steroids and endurance running in women. Fertil Steril 1983; 39:629. 273. Siklosi G, et al. Episodic secretion of hormones and the diagnostic value of single blood estimates. II. Progesterone, oestradiol and oestrone. Acta Med Hung 1984; 41:203. 274. Buster JE, et al. Subhourly variability of circulating third trimester maternal steroid concentrations as a source of sampling error. J Clin Endocrinol Metab 1978; 46:907. 275. Andersen AN, et al. Low maternal but normal fetal prolactin levels in cigarette smoking pregnant women. Acta Obstet Gynecol Scand 1984; 63:237. 276. Haning RV Jr, et al. Effects of fetal sex and dexamethasone on preterm maternal serum concentrations of human chorionic gonadotropin, progesterone, estrone, estradiol, and estriol. Am J Obstet Gynecol 1989; 161:1549. 277. MacMahon B, et al. Cigarette smoking and urinary estrogens. N Engl J Med 1982; 307:1062. 278. Golay A, et al. Effect of obesity on ambient plasma glucose, free fatty acid, insulin, growth hormone, and glucagon concentrations. J Clin Endocrinol Metab 1986; 63:481. 279. Singer L, Ophaug R. Ionic and nonionic fluoride in plasma (or serum). Crit Rev Clin Lab Sci 1982; 18:111. 280. Rimoli C, et al. Relationship between serum concentrations of flecainide and fluoride in humans. Boll Chim Farm 1991; 130:279. 281. Nakazawa Y, et al. Serum folic acid levels and antipyrine clearance rates in smokers and non-smokers. Drug Alcohol Depend 1983; 11:201. 282. Wickham C, et al. Seasonal variation in folate nutritional status. Ir J Med Sci 1983; 152:295. 283. Strain GW, et al. Mild hypogonadotropic hypogonadism in obese men. Metabolism 1982; 31:871. 284. Csako G. Causes, consequences, and recognition of false-positive reactions for ketones. Clin Chem 1990; 36:1388. 285. Haning RV Jr, et al. Effects of fetal sex and dexamethasone on preterm maternal serum concentrations of human chorionic gonadotropin, progesterone, estrone, estradiol, and estriol. Am J Obstet Gynecol 1989; 161:1549. 286. Dart AM, et al. The effect of propranolol on luteinising hormone and prolactin plasma concentrations in hypertensive women. Br J Clin Pharmacol 1982; 14:839. 287. Sawhney RC, et al. Hormone profiles at high altitude in man. Andrologia 1985; 17:178. 288. Siklosi G, et al. Episodic secretion of hormones and the diagnostic value of single blood estimates. I. LH, FSH, prolactin. Acta Med Hung 1984; 41:195. 289. Mendelson JH, et al. Effects of marijuana or neuroendocrine hormones in human males and females. Natl Inst Drug Abuse Res Monograph Ser 1984; 44:97. 290. MacNaughton J, et al. Age related changes in follicle stimulating hormone, luteinizing hormone, oestradiol and immunoreactive inhibin in women of reproductive age. Clin Endocrinol (Oxf) 1992; 36:339. 291. Gebhart SSP, et al. Reversible impairment of gonadotropin secretion in critical illness. Observations in postmenopausal women. Arch Intern Med 1989; 149:1637. 292. Valero-Politi J, et al. Within- and between-subject biological variations of follitropin, lutropin, testosterone, and sex-hormone-binding globulin in men. Clin Chem 1993; 39:1723. 293. Medri G, et al. Pituitary glycoprotein hormones in chronic renal failure: evidence for an uncontrolled alpha-subunit release. J Endocrinol Invest 1993; 16:169. 294. Glass AR. Ketoconazole-induced stimulation of gonadotropin output in men: basis for a potential test of gonadotropin reserve. J Clin Endocrinol Metab 1986; 63:1121. 295. Gebhart SSP, et al. Reversible impairment of gonadotropin secretion in critical illness. Observations in postmenopausal women. Arch Intern Med 1989; 149:1637. 296. Frandsen EK, et al. Serum fructosamine in diabetic pregnancy. Clin Chem 1988; 34:316. 297. De Schepper J, et al. Reference values for fructosamine concentrations in children's sera: influence of protein concentration, age, and sex. Clin Chem 1988; 34:2444. 298. Rodriguez-Segade S, et al. Effects of various serum proteins on quantification of fructosamine. Clin Chem 1989; 35:134. 299. Kurishita M, et al. Glycated hemoglobin of fractionated erythrocytes, glycated albumin, and plasma fructosamine during pregnancy. Am J Obstet Gynecol 1992; 167:1372. 300. Broussolle C, et al. Evaluation of the fructosamine test in obesity: consequences for the assessment of past glycemic control in diabetes. Clin Biochem 1991; 24:203. 301. Davie SJ, et al. Biological variation in glycated proteins. Ann Clin Biochem 1993; 30:260. 302. Cho CH, et al. Effects of eight-week treatment with oral ranitidine on plasma level changes of gastrin, histamine and serotonin in duodenal ulcer patients. Pharmacol Res Commun 1985; 17:525. 303. Sherbaniuk RW, et al. Gastrin, gastric emptying, and gastroesophageal reflux after ranitidine. J Clin Gastroenterol 1983; 5:239. 304. Feldman M, et al. Sex-related differences in gastrin release and parietal cell sensitivity to gastrin in healthy human beings. J Clin Invest 1983; 71:715. 305. Uvnas-Moberg K, et al. Nocturnal variation in plasma levels of gastrin and somatostatin-like immunoreactivity in man. Acta Physiol Scand 1984; 120:517. 306. Caldara R, et al. Effect of L-dopa with and without inhibition of extra cerebral dopa decarboxylase on gastric acid secretion and gastrin release in man. Gut 1985; 26:1014. 307. Lauritsen KB, et al. Lithium inhibits basal and food-stimulated gastrin secretion. Gastroenterology 1978; 75:59. 308. Jansen JB, et al. Effect of long-term treatment with omeprazole on serum gastrin and serum group A and C pepsinogens in patients with reflux esophagitis. Gastroenterology 1990; 99:621. 309. Romeo DP, et al. Misdiagnosis of the Zollinger-Ellison syndrome due to hyperlipidemia. Gastroenterology 1990; 99:1511. 310. Lacey RJ, et al. Evidence that the presence of arginine can lead to overestimation of glucagon levels measured by radioimmunoassay. Clin Chim Acta 1992; 210:211. 311. Lang DA, et al. Pulsatile, synchronous basal insulin and glucagon secretion in man. Diabetes 1982; 31:22. 312. LeBlanc J, et al. Daily variations of plasma glucose and insulin in physically trained and sedentary subjects. Metabolism 1983; 32:552. 313. Bergstrom J, et al. Influence of protein intake on renal hemodynamics and plasma hormone concentrations in normal subjects. Acta Med Scand 1985; 217:189. 314. Gossain VV, et al. Plasma glucagon in simple obesity: effect of exercise. Am J Med Sci 1983; 286:4. 315. Starke AAR, et al. Elevated pancreatic glucagon in obesity. Diabetes 1984; 33:277. 316. Naveri H, et al. Plasma glucagon and catecholamines during exhaustive short-term exercise. Eur J Appl Physiol 1985; 53:308. 317. Holst JJ, et al. Diurnal profile of pancreatic polypeptide, pancreatic glucagon, gut glucagon, and insulin in human morbid obesity. Int J Obes 1983; 7:529. 318. Lawrence AM, et al. Chronic propranolol administration impairs glucagon release during insulin-induced hypoglycemia in normal man. J Clin Endocrinol Metab 1984; 59:622. 319. Orskov C, et al. All products of proglucagon are elevated in plasma from uremic patients. J Clin Endocrinol Metab 1992; 74:379. 320. Dolhofer R, et al. Increased glycosylation of serum albumin in diabetes mellitus. Diabetes 1980; 29:417. 321. Willey DG, et al. Glycosylated haemoglobin and plasma glycoprotein assays by affinity chromatography. Diabetologia 1984; 27:56. 322. Goldstein DE, et al. Recent advances in glycosylated hemoglobin measurements. Crit Rev Clin Lab Sci 1984; 21:187. 323. Goldstein DE, et al. Glycated hemoglobin: methodologies and clinical applications. Clin Chem 1986; 32:B64. 324. Arnetz BB, et al. The influence of aging on hemoglobin A1c (HbA1c). J Gerontol 1982; 37:648. 325. Ktorza A, et al. Effect of slight plasma glucose decrease on glycosylated hemoglobin in healthy subjects during caloric restriction. N Engl J Med 1985; 313:958. 326. Cesana G, et al. Can glycosylated hemoglobin be a job stress parameter? J Occup Med 1985; 27:357. 327. Blum M, et al. Glycohemoglobin (Hb A1) levels in oral contraceptive users. Eur J Obstet Gynecol Reprod Biol 1983; 15:97. 328. Strickland MH, et al. Haemoglobin A1c concentrations in men and women with diabetes. BMJ 1984; 289:733. 329. D'Armiento M, et al. Diazepam-stimulated GH secretion in normal subjects: relation to oestradiol plasma levels. Horm Metab Res 1984; 16:155. 330. Duursma SA, et al. Changes in serum somatomedin and growth hormone concentrations after 3 weeks oestrogen substitution in post-menopausal women; a pilot study. Acta Endocrinol (Copenh) 1984; 106:527. 331. Baumann G, et al. Molecular forms of circulating growth hormone during spontaneous secretory episodes and in the basal state. J Clin Endocrinol Metab 1985; 60:1216. 332. Zadik Z, et al. The influence of age on the 24-hour integrated concentration of growth hormone in normal individuals. J Clin Endocrinol Metab 1985; 60:513. 333. Laakmann G, et al. Stimulation of growth hormone secretion by desimipramin and chlorimipramin in man. J Clin Endocrinol Metab 1977; 44:1010. 334. Lal S, et al. Effect of clonidine on growth hormone, prolactin, luteinizing hormone, follicle-stimulating hormone, and thyroid-stimulating hormone in the serum of normal men. J Clin Endocrinol Metab 1975; 41:827. 335. Meistas MT, et al. Integrated concentrations of growth hormone, insulin, C-peptide and prolactin in human obesity. Metabolism 1982; 31:1224. 336. Elwes RDC, et al. Prolactin and growth hormone dynamics in epileptic patients receiving phenytoin. Clin Endocrinol (Oxf) 1985; 23:263. 337. Monteleone P, et al. Growth hormone response to sodium valproate in chronic schizophrenia. Biol Psychiatry 1986; 21:588. 338. Jarrett DB, et al. Recurrent depression is associated with a persistent reduction in sleep-related growth hormone secretion. Arch Gen Psychiatry 1990; 47:113. 339. Schmitt JK. Indomethacin increases plasma growth hormone levels in man. Am J Med Sci 1990; 300:144. 340. Veldhuis JD, et al. Divergent effects of short term glucocorticoid excess on the gonadotropic and somatotropic axes in normal men. J Clin Endocrinol Metab 1992; 74:96. 341. Superko HR, et al. High-density lipoprotein cholesterol measurements. A help or hindrance in practical clinical medicine? JAMA 1986; 256:2714. 342. De Marchi S, et al. Relief of pruritus and decreases in plasma histamine concentrations during erythropoietin therapy in patients with uremia. N Engl J Med 1992; 326:969. 343. Johnston CS, et al. Antihistamine effect of supplemental ascorbic acid and neutrophil chemotaxis. J Am Coll Nutr 1992; 11:172. 344. Kong AN, et al. Pharmacokinetics and pharmacodynamic modeling of direct suppression effects of methylprednisolone on serum cortisol and blood histamine in human subjects. Clin Pharmacol Ther 1989; 46:616. 345. Tuchman M, et al. Assessment of the diurnal variations in urinary homovanillic and vanillylmandelic acid excretion for the diagnosis and follow-up of patients with neuroblastoma. Clin Biochem 1985; 18:176. 346. Brown GL, et al. Urinary 3-methyoxy-4-hydroxyphenylglycol and homovanillic acid response to d-amphetamine in hyperactive children. Biol Psychiatry 1981; 16:779. 347. Martineau J, et al. Monoamines (serotonin and catecholamines) and their derivatives in infantile autism: age-related changes and drug effects. Dev Med Child Neurol 1992; 34:593. 348. Giacalone M, et al. Cefoxitin: another Porter-Silber chromogen. South Med J 1985; 78:493. 349. Williams FA, et al. Acquired primary hypoaldosteronism due to an isolated zona glomerulosa defect. N Engl J Med 1983; 309:1623. 350. Elghozi JL, et al. Pizotifen increases 5-HIAA urinary excretion in male healthy volunteers. Eur J Clin Pharmacol 1984; 27:191. 351. O'Leary PC, et al. Indomethacin therapy and the measurement of urinary 5-hydroxyindoleacetic acid. Clin Chim Acta 1982; 126:323. 352. Dzurik R, et al. Blood pressure, 5-OH indoleacetic acid, and vanilmandelic acid excretion and blood platelet aggregation in hypertensive patients treated with ketanserin. J Cardiovasc Pharmacol 1985; 7(Suppl 7):S29. 353. Ueshiba H, et al. Enzyme-linked immunosorbent assay (ELISA) method for screening of non-classical steroid 21-hydroxylase deficiency. Horm Metab Res 1994; 26:43. 354. Solyom J. Diurnal variation in blood 17-hydroxyprogesterone concentrations in untreated congenital adrenal hyperplasia. Arch Dis Child 1984; 59:743. 355. Akamatsu T, et al. Menopause related changes of adrenocortical steroid production. Asia Oceania J Obstet Gynaecol 1992; 18:271. 356. Murphy JF, et al. Plasma 17-hydroxyprogesterone concentrations in ill newborn infants. Arch Dis Child 1983; 58:532. 357. Whitworth JA, et al. Plasma 4-pregnene-17 alpha, 20 alpha-diol-3-one (17 alpha, 20 alpha-dihydroxyprogesterone) and 17 alpha-hydroxyprogesterone in man. Acta Endocrinol (Copenh) 1983; 102:271. 358. Yoneyama K, et al. The day to day variations of urinary hydroxyproline and creatinine excretions, and dietary protein intake. Nippon Eiseigaku Zasshi 1984; 39:587. 359. Minisola S, et al. Effects of age on the urinary excretion of total and non-dialyzable hydroxyproline. Ric Clin Lab 1984; 14:649. 360. Meier DE, et al. Racial differences in pre- and postmenopausal bone homeostasis: association with bone density. J Bone Miner Res 1992; 7:1181. 361. Mizunashi K, et al. Effect of omeprazole, an inhibitor of H+,K(+)-ATPase, on bone resorption in humans. Calcif Tissue Int 1993; 53:21. 362. Enberg G, et al. Immunoreactive IGF-II in serum of healthy subjects and patients with growth hormone disturbances and uraemia. Acta Endocrinol (Copenh) 1984; 107:164. 363. Hall K, et al. Somatomedin levels in childhood, adolescence and adult life. Clin Endocrinol Metab 1984; 13:91. 364. Clemmons DR, et al. Factors controlling blood concentration of somatomedin C. Clin Endocrinol Metab 1984; 13:113. 365. Kao PC, et al. Somatomedin C: an index of growth hormone activity. Mayo Clin Proc 1986; 61:908. 366. Nielsen HK, et al. Changes in biochemical markers of osteoblastic activity during the menstrual cycle. J Clin Endocrinol Metab 1990; 70:1431. 367. Miell JP, et al. Effects of dexamethasone on growth hormone (GH)-releasing hormone, arginine- and dopaminergic stimulated GH secretion, and total plasma insulin-like growth factor-I concentrations in normal male volunteers. J Clin Endocrinol Metab 1991; 72:675. 368. Daughaday WH, et al. Heterogeneity of serum peptides with immunoactivity detected by a radioimmunoassay for proinsulin-like growth factor-II E domain: description of a free E domain peptide in serum. J Clin Endocrinol Metab 1992; 75:641. 369. Weaver JU, et al. Decreased sex hormone binding globulin (SHBG) and insulin-like growth factor binding protein (IGFBP-1) in extreme obesity. Clin Endocrinol (Oxf) 1990; 33:415. 370. Lee PD, et al. Regulation and function of insulin-like growth factor-binding protein-1. Proc Soc Exp Biol Med 1993; 204:4. 371. Hall K, et al. IGFBP-1. Production and control mechanisms. Acta Endocrinol (Copenh) 1991; 124(Suppl 2):48. 372. Giudice LC, et al. Insulin-like growth factor binding proteins in maternal serum throughout gestation and in the puerperium: effects of a pregnancy-associated serum protease activity. J Clin Endocrinol Metab 1990; 71:806. 373. Argente J, et al. Normative data for insulin-like growth factors (IGFs), IGF-binding proteins, and growth hormone-binding protein in a healthy Spanish pediatric population: age- and sex-related changes. J Clin Endocrinol Metab 1993; 77:1522. 374. Miell JP, et al. The effects of dexamethasone treatment on immunoreactive and bioactive insulin-like growth factors (IGFs) and IGF-binding proteins in normal male volunteers. J Endocrinol

1993; 136:525. 375. Rutanen EM, et al. Assays for IGF binding proteins. Acta Endocrinol (Copenh) 1991; 124(Suppl 2):70. 376. Davenport ML, et al. Insulin-like growth factor-binding protein-3 proteolysis is induced after elective surgery. J Clin Endocrinol Metab 1992; 75:590. 377. Jorgensen JO, et al. Circadian patterns of serum insulin-like growth factor (IGF) II and IGF binding protein 3 in growth hormone-deficient patients and age- and sex-matched normal subjects. Acta Endocrinol (Copenh) 1990; 123:257. 378. Vorherr H, et al. Vaginal absorption of povidoneiodine. JAMA 1980; 244:2628. 379. Chanoine JC, et al. Iodine contamination of urine samples by test strips. Clin Chem 1987; 33:1935. 380. Ford HCE. Ascorbic acid interferes with an automated urinary iodide determination based on the cericarsenious acid reaction. Clin Chem 1991; 37:759. 381. Grasso L, et al. Iodine contamination in subjects admitted to a general hospital. J Endocrinol Invest 1992; 15:307. 382. Saxena S, et al. Effect of blood transfusion on serum iron and transferrin saturation. Arch Pathol Lab Med 1993; 117:622. 383. Fairbanks VF. Laboratory testing for iron status. Hosp Pract (Off Ed) 1991; 26(Suppl 3):17. 384. Touitou Y, et al. Seasonal rhythms of plasma gonadotrophins: their persistence in elderly men and women. J Endocrinol 1983; 96:15. 385. Rudd BT. Urinary 17-oxogenic and 17-oxosteroids. A case for deletion from the clinical chemistry repertoire. Ann Clin Biochem 1983; 20:65. 386. Panteghini M, Pagani F. Biological variation of lactate and pyruvate in blood. Clin Chem 1993; 39:908. 387. Vital health statistics, series 11, no 233. Rockville, MD: National Center for Health Statistics, 1984:1. 388. Panteghini M, Pagani F. Pre-analytical, analytical and biological sources of variation of lipoprotein(a). Eur J Clin Chem Clin Biochem 1993; 31:23. 389. Webb AT, et al. Does cyclosporin increase lipoprotein(a) concentrations in renal transplant recipients? Lancet 1993; 341:268. 390. de Bruin TW, et al. Lipoprotein(a) and apolipoprotein B plasma concentrations in hypothyroid, euthyroid, and hyperthyroid subjects. J Clin Endocrinol Metab 1993; 76:121. 391. Haffner SM, et al. Increased lipoprotein(a) concentrations in chronic renal failure. J Am Soc Nephrol 1992; 3:1156. 392. Huang CM, et al. The effect of alcohol withdrawal on serum concentrations of Lp(a), apolipoproteins A-1 and B, and lipids. Alcohol Clin Exp Res 1992; 16:895. 393. Cobbaert C, et al. Serum lipoprotein(a) levels in racially different populations. Am J Epidemiol 1992; 136:441. 394. Slunga L, et al. Lipoprotein (a) in a randomly selected 25-64 year old population: the Northern Sweden Monica Study. J Clin Epidemiol 1993; 46:617. 395. Marcovina SM, et al. Lipoprotein[a] concentrations and apolipoprotein[a] phenotypes in Caucasians and African Americans. The CARDIA study. Arterioscler Thromb 1993; 13:1037. 396. Soma MR, et al. The lowering of lipoprotein[a] induced by estrogen plus progesterone replacement therapy in postmenopausal women. Arch Intern Med 1993; 153:1462. 397. Crook D, et al. Lipoprotein Lp(a) levels are reduced by danazol, an anabolic steroid. Atherosclerosis 1992; 92:41. 398. Van Biervliet JP, et al. Lipoprotein(a) profiles and evolution in newborns. Atherosclerosis 1991; 86:173. 399. Kervinen K, et al. A rapid increase in lipoprotein (a) levels after ethanol withdrawal in alcoholic men. Life Sci 1991; 48:2183. 400. Maeda S, et al. Transient changes of serum lipoprotein(a) as an acute phase protein. Atherosclerosis 1989; 78:145. 401. Hollifield JW. Potassium and magnesium abnormalities: diuretics and arrhythmias in hypertension. Am J Med 1984; 77:28. 402. Schilsky RL, et al. Hypomagnesemia and renal magnesium wasting in patients receiving cisplatin. Ann Intern Med 1979; 90:929. 403. Whang R, et al. Frequency of hypomagnesemia in hospitalized patients receiving digitalis. Arch Intern Med 1985; 145:655. 404. Mather HM, et al. Diurnal profiles of plasma magnesium and blood glucose in diabetes. Diabetologia 1982; 22:180. 405. Thompson CB, et al. Association between cyclosporin neurotoxicity and hypomagnesaemia. Lancet 1984; 2:1116. 406. Fischer JA, Dambacher MA. Differential diagnosis of hypercalcitoninemia. Biomed Pharmacother 1984; 38:255. 407. Hatano S, et al. Erythrocyte manganese concentration in healthy Japanese children, adults, and the elderly, and in cord blood. Am J Clin Nutr 1983; 37:457. 408. Pleban PA, Pearson KH. Determination of manganese in whole blood and serum. Clin Chem 1979; 25:1915. 409. Theron JJ, et al. Effect of physical exercise on plasma melatonin levels in normal volunteers. S Afr Med J 1984; 66:838. 410. Attanasio A, et al. Circadian rhythms in serum melatonin from infancy to adolescence. J Clin Endocrinol Metab 1985; 61:388. 411. Cowen PJ, et al. Treatment with beta-adrenoceptor blockers reduces plasma melatonin concentration. Br J Clin Pharmacol 1985; 19:258. 412. Touitou Y, et al. Age- and sex-associated modification of plasma melatonin concentrations in man. Relationship to pathology, malignant or not, and autopsy findings. Acta Endocrinol (Copenh) 1985; 108:135. 413. Webley GE, et al. The circadian pattern of melatonin and its positive relationship with progesterone in women. J Clin Endocrinol Metab 1986; 63:323. 414. Rojdmark S, Wetterberg L. Short-term fasting inhibits the nocturnal melatonin secretion in healthy man. Clin Endocrinol (Oxf) 1989; 30:451. 415. Kivela A. Serum melatonin during human pregnancy. Acta Endocrinol (Copenh) 1991; 124:233. 416. Viljoen M, et al. Melatonin in chronic renal failure. Nephron 1992; 60:138. 417. Juan D. Pheochromocytoma: clinical manifestations and diagnostic tests. Urology 1981; 17:1. 418. Elia J, et al. Stimulant drug treatment of hyperactivity: biochemical correlates. Clin Pharmacol Ther 1990; 48:57. 419. Wills MR, et al. Serum and lymphocyte, aluminum and nickel in chronic renal failure. Clin Chim Acta 1985; 145:193. 420. Gundberg CM, et al. Measurements of gamma-carboxyglutamate and circulating osteocalcin in normal children and adults. Clin Chim Acta 1983; 128:1. 421. Epstein S, et al. Differences in serum bone GLA protein with age and sex. Lancet 1984; 1:307. 422. Gundberg CM, et al. Osteocalcin in human serum: a circadian rhythm. J Clin Endocrinol Metab 1985; 60:736. 423. Reid IR, et al. Low serum osteocalcin levels in glucocorticoid-treated asthmatics. J Clin Endocrinol Metab 1986; 62:379. 424. Menon RK, et al. Impaired carboxylation of osteocalcin in warfarin-treated patients. J Clin Endocrinol Metab 1987; 64:59. 425. Cole DEC, et al. Changing osteocalcin concentrations during pregnancy and lactation: implications for maternal mineral metabolism. J Clin Endocrinol Metab 1987; 65:290. 426. Napal J, et al. Stress decreases the serum level of osteocalcin. Bone Miner 1993; 21:113. 427. Perry HM 3d, et al. The effects of thiazide diuretics on calcium metabolism in the aged. J Am Geriatr Soc 1993; 41:818. 428. Karlsson R, et al. Osteocalcin 24-hour profiles during normal pregnancy. Gynecol Obstet Invest 1992; 34:197. 429. Kakonen SM, et al. Development and evaluation of three immunofluorometric assays that measure different forms of osteocalcin in serum. Clin Chem 2000; 46:332. 430. Power MJ, Fottrell PF. Osteocalcin: diagnostic methods and clinical applications. Crit Rev Clin Lab Sci 1991; 28:287. 431. Karlsson R, et al. Oral contraception affects osteocalcin serum profiles in young women. Osteoporos Int 1992; 2:118. 432. Elomaa I, et al. Seasonal variation of urinary calcium and oxalate excretion, serum 25(OH)D 3 and albumin level in relation to renal stone formation. Scand J Urol Nephrol 1982; 16:155. 433. Zerwekh JE, et al. Assay of urinary oxalate: six methodologies compared. Clin Chem 1983; 29:1977. 434. Balcke P, et al. Ascorbic acid aggravates secondary hyperoxalemia in patients on chronic hemodialysis. Ann Intern Med 1984; 101:344. 435. France NC, et al. In vitro oxalogenesis, and measurement of oxalate in serum. Clin Chem 1985; 31:335. 436. Leumann EP, et al. Urinary oxalate and glycolate excretion in healthy infants and children. Pediatr Nephrol 1990; 4:493. 437. Balchin ZE, et al. Biological variation of urinary oxalate in different specimen types. Ann Clin Biochem 1991; 28:622. 438. Otsuki Y, et al. Serial plasma oxytocin levels during pregnancy and labor. Acta Obstet Gynecol Scand 1983; 62:15. 439. Amico JA, et al. Oxytocin in human plasma: correlation with neurophysin and stimulation with estrogen. J Clin Endocrinol Metab 1981; 52:988. 440. Kumaresan P, et al. Human ovulation and plasma oxytocin. Int J Gynaecol Obstet 1983; 21:413. 441. Zaccaria M, et al. Effects of pirenzepine on plasma insulin, glucagon and pancreatic polypeptide levels in normal man. Eur J Clin Pharmacol 1985; 27:701. 442. Lugari R, et al. Human pancreatic polypeptide and somatostatin in chronic renal failure. Proc Eur Dial Transplant Assoc Eur Ren Assoc 1985; 21:614. 443. Hansen BC, et al. Fluctuations in basal plasma levels of pancreatic polypeptide in monkeys and humans. Am J Physiol 1985; 248:R739. 444. O'Hare MMT, et al. Radioimmunoassay for pancreatic polypeptide, and its age-related changes in concentration. Clin Chem 1983; 29:1923. 445. Cho CH, et al. Serum concentration of pancreatic polypeptide in normal subjects and in peptic ulcer patients. Horm Metab Res 1985; 17:215. 446. Jorde R, et al. Fasting and postprandial plasma pancreatic polypeptide (PP) levels in obesity. Int J Obes 1984; 8:393. 447. Hanukoglu A, et al. Human pancreatic polypeptide in children and young adults. Horm Metab Res 1990; 22:41. 448. Hackeng WHL, et al. Clinical implications of estimation of intact parathyroid hormone (PTH) versus total immunoreactive PTH in normal subjects and hyperparathyroid patients. J Clin Endocrinol Metab 1986; 63:447. 449. Bell NH, et al. Evidence for alteration of the vitamin D-endocrine system in obese subjects. J Clin Invest 1985; 76:370. 450. Bell NH, et al. Evidence for alteration of the vitamin D-endocrine system in blacks. J Clin Invest 1985; 76:470. 451. Petersen MM, et al. Parathyroid hormone and 25-hydroxyvitamin D concentrations in sick and normal elderly people. BMJ 1983; 287:521. 452. Specker BL, et al. Calcium-regulating hormones and minerals from birth to 18 months of age: a cross-sectional study. II. Effects of sex, race, age, season, and diet on serum minerals, parathyroid hormone, and calcitonin. Pediatrics 1986; 77:891. 453. Marcus R, et al. Age-related changes in parathyroid hormone and parathyroid hormone action in normal humans. J Clin Endocrinol Metab 1984; 58:223. 454. Kitamura N, et al. Episodic fluctuation in serum intact parathyroid hormone concentration in men. J Clin Endocrinol Metab 1990; 70:252. 455. Rudnicki M, et al. Effects of age, sex, season and diet on serum ionized calcium, parathyroid hormone and vitamin D in a random population. J Intern Med 1993; 234:195. 456. Frolich A, et al. Serum concentrations of intact parathyroid hormone during late human pregnancy: a longitudinal study. Eur J Obstet Gynecol Reprod Biol 1991; 42:85. 457. Saggese G, et al. Determination of intact parathyrin by immunoradiometric assay evaluated in normal children and in patients with various disorders of calcium metabolism. Clin Chem 1991; 37:1999. 458. Specker BL, et al. Changes in calcium homeostasis over the first year postpartum: effect of lactation and weaning. Obstet Gynecol 1991; 78:56. 459. Saggese G, et al. Intact parathyroid hormone levels during pregnancy, in healthy term neonates and in hypocalcemic preterm infants. Acta Paediatr Scand 1991; 80:36. 460. Prince RL, et al. Effects of transdermal estrogen replacement on parathyroid hormone secretion. J Clin Endocrinol Metab 1990; 71:1284. 461. Armitage EK. Parathyrin (parathyroid hormone): metabolism and methods for assay. Clin Chem 1986; 32:418. 462. Grill V, et al. Parathyroid hormone-related protein: a possible endocrine function in lactation. Clin Endocrinol (Oxf) 1992; 37:405. 463. Bucht E, et al. Parathyroid hormone-related peptide, measured by a mid-molecule radioimmunoassay, in various hypercalcaemic and normocalcaemic conditions. Acta Endocrinol (Copenh) 1992; 127:294. 464. Mallette LE. The parathyroid polyhormones: new concepts in the spectrum of peptide hormone action. Endocr Rev 1991; 12:110. 465. Sonnenblick M, et al. Paraprotein interference with colorimetry of phosphate in serum of some patients with multiple myeloma. Clin Chem 1986; 32:1537. 466. Uza G. Effect of clonidine on serum inorganic phosphorus in patients with essential arterial hypertension. J Hum Hypertens 1990; 4:687. 467. Licata AA, et al. The urinary excretion of phosphoethanolamine in diseases other than hypophosphatasia. Am J Med 1978; 64:133. 468. Eastman JR, Bixler D. Urinary phosphoethanolamine: normal values by age. Clin Chem 1980; 26:1757. 469. Houghton DJ, et al. Relationship of maternal and fetal levels of human placental lactogen to the weight and sex of the fetus. Placenta 1984; 5:455. 470. Shaw MA, et al. Adsorption of pregnanediol-3 alpha-glucuronide by plastics and glass. J Steroid Biochem 1983; 19:1681. 471. Metcalf MG, et al. Indices of ovulation: comparison of plasma and salivary levels of progesterone with urinary pregnanediol. J Endocrinol 1984; 100:75. 472. Junkermann H, et al. Circadian rhythm of serum progesterone levels in human pregnancy and its relation to the rhythm of cortisol. Acta Endocrinol (Copenh) 1982; 101:98. 473. Smith RL. Effect of serum-separating gels on progesterone assays. Clin Chem 1985; 31:1239. 474. Healy DL, et al. Pulsatile progesterone secretion: its relevance to clinical evaluation of corpus luteum function. Fertil Steril 1984; 41:114. 475. Spitz IM, et al. Response to intermittent RU486 in women. Fertil Steril 1993; 59:971. 476. Cohen RM, et al. A radioimmunoassay for circulating human proinsulin. Diabetes 1985; 34:84. 477. Deacon CF, et al. Measurement of circulating human proinsulin concentrations using a proinsulin-specific antiserum. Diabetes 1985; 34:491. 478. Beer SF, et al. The effect of thyroid disease on proinsulin and C-peptide levels. Clin Endocrinol (Oxf) 1989; 30:379. 479. Kahn SE, et al. Effect of glucocorticoid and growth hormone treatment on proinsulin levels in humans. Diabetes 1993; 42:1082. 480. Davies MJ, et al. Insulin deficiency and increased plasma concentration of intact and 32/33 split proinsulin in subjects with impaired glucose tolerance. Diabet Med 1993; 10:313. 481. Christensen SE, et al. Body temperature elevation, exercise and serum prolactin concentrations. Acta Endocrinol (Copenh) 1985; 109:458. 482. Jackson RD, et al. Macroprolactinemia presenting like a pituitary tumor. Am J Med 1985; 78:346. 483. Djahanbakhch O, et al. Changes in plasma levels of prolactin, in relation to those of FSH, oestradiol, androstenedione and progesterone around the preovulatory surge of LH in women. Clin Endocrinol (Oxf) 1984; 20:463. 484. Govoni S, et al. Plasma prolactin concentrations in a large population of healthy old people. BMJ 1983; 287:1107. 485. Mendelson J, et al. Acute effects of marihuana smoking on prolactin levels in human females. J Pharmacol Exp Ther 1985; 232:220.

486. 487. 488. 489. 490. 491. 492. 493. 494. 495. 496. 497. 498. 499. 500. 501. 502. 503. 504. 505. 506. 507. 508. 509. 510. 511. 512. 513. 514. 515. 516. 517. 518. 519. 520. 521. 522. 523. 524. 525. 526. 527. 528. 529. 530. 531. 532. 533. 534. 535. 536. 537. 538. 539. 540. 541. 542. 543. 544. 545. 546. 547. 548. 549. 550. 551. 552. 553. 554. 555. 556. 557. 558. 559. 560. 561. 562. 563. 564. 565. 566. 567. 568. 569. 570. 571. 572. 573. 574. 575. 576. 577. 578. 579. 580. 581. 582. 583. 584. 585. 586. 587. 588. 589. 590. 591.

Verbeelen D, et al. Effect of 1,25-dihydroxyvitamin D3 and nifedipine on prolactin release in normal man. J Endocrinol Invest 1985; 8:103. Lindstedt F, et al. “Hyperprolactinemia” in healthy women. Clin Chem 1984; 30:165. Musey VC, et al. Long-term effect of a first pregnancy on the secretion of prolactin. N Engl J Med 1987; 316:229. Maestri E, et al. Effects of five days verapamil administration on serum GH and PRL levels. Horm Metab Res 1985; 17:482. Mendelson JH, et al. Cocaine effects on pulsatile secretion of anterior pituitary, gonadal, and adrenal hormones. J Clin Endocrinol Metab 1989; 69:1256. Yeksan M, et al. Effect of recombinant human erythropoietin (r-HuEPO) therapy on plasma FT3, FT4, TSH, FSH, LH, free testosterone and prolactin levels in hemodialysis patients. Int J Artif Organs 1992; 15:585. Willenbring ML, et al. Psychoneuroendocrine effects of methadone maintenance. Psychoneuroendocrinology 1989; 14:371. Maestri E, et al. Effects of five days verapamil administration on serum GH and PRL levels. Horm Metab Res 1985; 17:482. Bettica P, et al. Bone-resorption markers galactosyl hydroxylysine, pyridinium crosslinks, and hydroxyproline compared. Clin Chem 1992; 38:2313. Schlemmer AE. Marked diurnal variation in urinary excretion of pyridinium cross-links in premenopausal women. J Clin Endocrinol Metab 1992; 74:476. Oesterling JE. Prostate specific antigen: a critical assessment of the most useful tumor marker for adenocarcinoma of the prostate. J Urol 1991; 145:907. Pummer K, et al. False positive prostate specific antigen values in the sera of women with renal cell carcinoma. J Urol 1992; 148:21. Oesterling JE, et al. Serum prostate-specific antigen in a community-based population of healthy men. Establishment of age-specific reference ranges. JAMA 1993; 270:860. Ruckle HC, et al. Prostate-specific antigen: critical issues for the practicing physician. Mayo Clin Proc 1994; 69:59. Glenski WJ, et al. Prostate-specific antigen: establishment of the reference range for the clinically normal prostate gland and the effect of digital rectal examination, ejaculation, and time on serum concentrations. Prostate 1992; 21:99. Zhou AME. Multiple forms of prostate-specific antigen in serum: differences in immunorecognition by monoclonal and polyclonal assays. Clin Chem 1993; 39:2483. Price A. Abrupt changes in prostate-specific antigen concentration in acute renal failure. Clin Chem 1993; 39:161. Lipton A, et al. Increased urinary excretion of pyridinium cross-links in cancer patients. Clin Chem 1993; 39:614. Peel N, et al. Sulfasalazine may interfere with HPLC assay of urinary pyridinium crosslinks. Clin Chem 1994; 40:167. Uebelhart D, et al. Effect of menopause and hormone replacement therapy on the urinary excretion of pyridinium cross-links. J Clin Endocrinol Metab 1991; 72:367. Eastell R, et al. Nyctohemeral changes in bone turnover assessed by serum bone Glaprotein concentration and urinary deoxypyridinoline excretion: effects of growth and ageing. Clin Sci 1992; 83:375. Hetland ML, et al. Low bone mass and high bone turnover in male long distance runners. J Clin Endocrinol Metab 1993; 77:770. Harris ST, et al. The effect of short term treatment with alendronate on vertebral density and biochemical markers of bone remodeling in early post-menopausal women. J Clin Endocrinol Metab 1993; 76:1399. Lipton A, et al. Increased urinary excretion of pyridinium cross-links in cancer patients. Clin Chem 1993; 39:614. McLaren AM, et al. Physiological variations in the urinary excretion of pyridinium crosslinks of collagen. Br J Rheumatol 1993; 32:307. Goldhaber SZ, et al. Plasma renin substrate, renin activity, and aldosterone levels in a sample of oral contraceptive users from a community survey. Am Heart J 1984; 107:119. Sealey JE, et al. Changes in active and inactive renin throughout normal pregnancy. Clin Exp Hypertens [A] 1982; 4:2373. Fiselier J, et al. The basal levels of active and inactive plasma renin concentration in infancy and childhood. Clin Sci 1984; 67:383. Higaki J, et al. A new sensitive direct radioimmunoassay for human plasma renin and its clinical application. J Lab Clin Med 1984; 104:947. Pessina AC, et al. Digitalis glycosides and plasma renin activity. Pharmacol Res Commun 1982; 14:629. Sealey JE. Plasma renin activity and plasma prorenin assays. Clin Chem 1991; 37:1811. Oektedalen O, et al. Basal hyperchlorhydria and its relation to the plasma concentrations of secretin, vasoactive intestinal peptide, and gastrin during prolonged strain. Regul Pept (Netherlands) 1983; 5:235. Oektedalen O, et al. Secretin—a new stress hormone? Regul Pept 1982; 4:213. Grekas DM, et al. Plasma secretin, pancreozymin, and somatostatin-like hormone in chronic renal failure patients. Uremia Invest 1984; 8:117. Beglinger C, et al. Pancreatic enzyme response to a liquid meal and to hormonal stimulation. Correlation with plasma secretin and cholecystokinin levels. J Clin Invest 1985; 75:1471. Subotzky EF, et al. Plasma zinc, copper, selenium, ferritin and whole blood manganese concentrations in children with kwashiorkor in the acute stage and during refeeding. Ann Trop Paediatr 1992; 12:13. Lloyd B, et al. Effect of smoking, alcohol, and other factors on the selenium status of a healthy population. J Epidemiol Community Health 1983; 37:213. Hurd RW, et al. Selenium, zinc, and copper changes with valproic acid: possible relation to drug side effects. Neurology 1985; 34:1393. Dworkin B, et al. Diminished blood selenium levels in renal failure patients on dialysis: correlations with nutritional status. Am J Med Sci 1987; 293:6. Marano G, et al. Increased serum selenium levels in patients under corticosteroid treatment. Pharmacol Toxicol 1990; 67:120. Hawker FH, et al. Effects of acute illness on selenium homeostasis. Crit Care Med 1990; 18:442. Marano G, et al. Changes of serum selenium and serum cholesterol in children during sexual maturation. J Trace Elem Electrolytes Health Dis 1991; 5:59. Anderson GM, et al. Effect of a meal on human whole blood serotonin. Gastroenterology 1985; 88:86. Davis DD, et al. Biological stress responses in high and low trait anxious students. Biol Psychiatry 1985; 20:843. DeLisi LE, et al. Increased whole blood serotonin concentrations in chronic schizophrenic patients. Arch Gen Psychiatry 1981; 38:647. Sanz Alaejos M, Diaz Romero C. Urinary selenium concentrations. Clin Chem 1993; 39:2040. Flachaire E, et al. Determination of reference values for serotonin concentration in platelets of healthy newborns, children, adults, and elderly subjects by HPLC with electrochemical detection. Clin Chem 1990; 36:2117. Hindberg I, et al. Serotonin concentrations in plasma and variations during the menstrual cycle. Clin Chem 1992; 38:2087. Eynard N, et al. Platelet serotonin content and free and total plasma tryptophan in healthy volunteers during 24 hours. Clin Chem 1993; 39:2337. Pietraszek MH, et al. Alcohol-induced depression: involvement of serotonin. Alcohol Alcohol 1991; 26:155. Steyn ME, et al. Whole blood serotonin levels in chronic renal failure. Life Sci 1992; 51:359. Gonzales GF, Carrillo C. Blood serotonin levels in postmenopausal women: effects of age and serum oestradiol levels. Maturitas 1993; 17:23. Anderson DC. Sex-hormone-binding globulin. Clin Endocrinol (Oxf) 1974; 3:69. Uriel J, et al. Maternal serum levels of sex steroid-binding protein during pregnancy. Br J Obstet Gynaecol 1981; 88:1229. Maruyama Y, et al. Variation with age in the levels of sex-steroid-binding plasma protein as determined by radioimmunoassay. Acta Endocrinol (Copenh) 1984; 106:428. Beastall GH, et al. Hormone binding globulins and anticonvulsant therapy. Scott Med J 1985; 30:101. Belgorosky A, Rivarola MA. Progressive decrease in serum sex hormone-binding globulin from infancy to late prepuberty in boys. J Clin Endocrinol Metab 1986; 63:510. Grenman S, et al. Sex steroid, gonadotropin, cortisol, and prolactin levels in healthy, massively obese women: correlation with abdominal fat cell size and effect of weight reduction. J Clin Endocrinol Metab 1986; 63:1257. Lonning PE, et al. Plasma levels of estradiol, estrone, estrone sulfate and sex hormone binding globulin in patients receiving rifampicin. J Steroid Biochem 1989; 33:631. Caballero MJ, et al. Effects of physical exercise on sex hormone binding globulin, high density lipoprotein cholesterol, total cholesterol and triglycerides in postmenopausal women. Endocr Res 1992; 18:261. Dibbelt L, et al. Group comparison of serum ethinyl estradiol, SHBG and CBG levels in 83 women using two low-dose combination oral contraceptives for three months. Contraception 1991; 43:1. Isojarvi JI. Serum steroid hormones and pituitary function in female epileptic patients during carbamazepine therapy. Epilepsia 1990; 31:438. Berlyne GM, Caruso C. Measurement of silicon in biological fluids in man using flameless furnace atomic absorption spectrophotometry. Clin Chim Acta 1983; 129:239. Mauras Y, et al. Increase in blood silicon concentration in patients with renal failure. Biomedicine 1980; 33:228. Saito H, et al. Plasma somatostatin level during natural and interrupted nocturnal sleep in man. Acta Endocrinol (Copenh) 1983; 104:129. Burhol PG, et al. Diurnal profile of plasma somatostatin in man. Acta Physiol Scand 1984; 120:67. Conlon JM, et al. The nature of big plasma somatostatin: implications for the measurement of somatostatin-like immunoreactivity in human plasma. Anal Biochem 1982; 125:243. Uvnas-Moberg K, et al. Plasma levels of gastrin, somatostatin, VIP, insulin and oxytocin during the menstrual cycle in women (with and without oral contraceptives). Acta Obstet Gynecol Scand 1989; 68:165. Holst N, et al. Plasma concentrations of motilin and somatostatin are increased in late pregnancy and postpartum. Br J Obstet Gynaecol 1992; 99:338. Bertello P, et al. Circadian patterns of plasma cortisol and testosterone in chronic male alcoholics. Alcohol Clin Exp Res 1982; 6:475. Sharp AM, et al. Further studies on danazol interference in testosterone radioimmunoassays. Clin Chem 1983; 29:141. Guglielmini C, et al. Variations of serum testosterone concentrations after physical exercises of different duration. Int J Sports Med 1984; 5:246. Wheeler GD, et al. Reduced serum testosterone and prolactin levels in male distance runners. JAMA 1984; 252:514. Buergi U, et al. Serum thyroglobulin before and after palpation of the thyroid. N Engl J Med 1983; 308:777. Penny R, et al. Thyroid-stimulating hormone and thyroglobulin levels decrease with chronological age in children and adolescents. J Clin Endocrinol Metab 1983; 56:177. Nakamura S, et al. Serum thyroglobulin (Tg) concentration in healthy subjects: absence of age- and sex-related differences. Endocrinol Jpn 1984; 31:93. Nakamura S, et al. Serum thyroglobulin concentration in normal pregnancy. Endocrinol Jpn 1984; 31:675. Catania A, et al. Circulating thyroglobulin and antithyroglobulin antibodies after fine needle aspiration of thyroid nodules. Horm Metab Res 1985; 17:49. Aiello DP, Manni A. Thyroglobulin measurement vs iodine 131 total-body scan for follow-up of well-differentiated thyroid cancer. Arch Intern Med 1990; 150:437. Ross HA, et al. Is free thyroxine accurately measurable at room temperature? Clin Chem 1992; 38:880. Kojima N, et al. Age- and sex-related differences of serum thyroxine binding globulin (TBG) in healthy subjects. Acta Endocrinol (Copenh) 1983; 104:303. Franklyn JA, et al. The effect of propranolol on circulating thyroid hormone measurements in thyrotoxic and euthyroid subjects. Acta Endocrinol (Copenh) 1985; 108:351. Bentsen KD, et al. Serum thyroid hormones and blood folic acid during monotherapy with carbamazepine or valproate. A controlled study. Acta Neurol Scand 1983; 67:235. Harrup JS, et al. Circannual and within-individual variation of thyroid function tests in normal subjects. Ann Clin Biochem 1985; 22:371. Ramsay I. Drug and non-thyroid induced changes in thyroid function tests. Postgrad Med J 1985; 61:375. Cashin-Hemphill L, et al. Alterations in serum thyroid hormonal indices with colestipolniacin therapy. Ann Intern Med 1987; 107:324. Benvenga S, et al. Effect of free fatty acids and nonlipid inhibitors of thyroid hormone binding in the immunoradiometric assay of thyroxin-binding globulin. Clin Chem 1987; 33:1752. Spencer CA, et al. Dynamics of serum thyrotropin and thyroid hormone changes in fasting. J Clin Endocrinol Metab 1983; 56:883. O'Malley BP, et al. Circadian rhythms of serum thyrotrophin and body temperature in euthyroid individuals and their responses to warming. Clin Sci 1984; 67:433. Weeke J, et al. A longitudinal study of serum TSH, and total and free iodothyronines during normal pregnancy. Acta Endocrinol (Copenh) 1982; 101:531. Greenspan SL, et al. Pulsatile secretion of thyrotropin in man. J Clin Endocrinol Metab 1986; 63:661. Ain KB, et al. Thyroid hormone levels affected by time of blood sampling in thyroxine-treated patients. Thyroid 1993; 3:81. Schedlowski M, et al. Acute psychological stress increases plasma levels of cortisol, prolactin and TSH. Life Sci 1992; 50:1201. Jiang N-S, et al. Assay for thyroid stimulating immunoglobulin. Mayo Clin Proc 1986; 61:753. Mielke AW, et al. Effects of a fat-containing meal on sex hormones in men. Metabolism 1990; 39:943. Hugues JN, et al. Effects of cimetidine on thyroid hormones. Clin Endocrinol (Oxf) 1982; 17:297. Guillaume J, et al. Components of the total serum thyroid hormone concentrations during pregnancy: high free thyroxine and blunted thyrotropin (TSH) response to TSH-releasing hormone in the first trimester. J Clin Endocrinol Metab 1985; 60:678. Bellastella A, et al. Circannual rhythms of plasma growth hormone, thyrotropin and thyroid hormones in prepuberty. Clin Endocrinol (Oxf) 1984; 20:531. Borowski GD, et al. Effect of long-term amiodarone therapy on thyroid hormone levels and thyroid function. Am J Med 1985; 78:443. Sepkovic DW, et al. Thyroid activity in cigarette smokers. Arch Intern Med 1983; 144:501. Franklyn JA, et al. Measurement of free thyroid hormones in patients on long-term phenytoin therapy. Eur J Clin Pharmacol 1984; 26:633. Franklyn JA, et al. The influence of age and sex on tests of thyroid function. Ann Clin Biochem 1985; 22:502. Ohnhaus EE, et al. The effect of antipyrine, phenobarbitol and rifampicin on thyroid hormone metabolism in man. Eur J Clin Invest 1981; 11:381. McConnell RJ. Abnormal thyroid function test results in patients taking salsalate. JAMA 1992; 267:1242. Ramaker J, Wood WG. Transthyretin—an explanation of “anomalous” serum thyroid hormone values in severe illness? J Clin Chem Clin Biochem 1990; 28:155. Docter R, et al. Free thyroxine assessed with three assays in sera of patients with nonthyroidal illness and of subjects with abnormal concentrations of thyroxine-binding proteins. Clin Chem 1993; 39:1668.

592. 593. 594. 595. 596. 597. 598. 599. 600. 601. 602. 603. 604. 605. 606. 607. 608. 609. 610. 611. 612. 613. 614. 615. 616. 617. 618. 619. 620. 621. 622. 623. 624. 625. 626. 627. 628. 629. 630. 631. 632. 633. 634. 635. 636. 637. 638. 639. 640. 641. 642. 643. 644. 645. 646. 647. 648. 649. 650. 651. 652. 653. 654. 655. 656. 657. 658. 659. 660. 661. 662. 663. 664. 665. 666. 667. 668. 669. 670. 671. 672. 673. 674. 675. 676. 677. 678. 679. 680. 681. 682. 683. 684. 685. 686. 687. 688. 689. 690. 691. 692. 693. 694. 695. 696. 697. 698. 699. 700. 701.

Wang Y-S, et al. Comparison of a new ultrafiltration method for serum free T 4 and free T3 with two RIA kits in eight groups of patients. J Endocrinolxs Invest 1985; 8:495. Ng RH, et al. Falsely high results for triglycerides in patients receiving intravenous nitroglycerin. Clin Chem 1986; 32:2098. Price A, et al. Comparison of methods for the determination of unbound triiodothyronine in pregnancy. Clin Endocrinol (Oxf) 1992; 37:41. Nanji AA, et al. Increase in serum uric acid level associated with cisplatin therapy. Correlation with liver but not kidney platinum concentrations. Arch Intern Med 1985; 145:2013. Burritt MF, et al. Pediatric reference intervals for 19 biologic variables in healthy children. Mayo Clin Proc 1990; 65:329. Simonoff M, et al. Vanadium in human serum, as determined by neutron activation analysis. Clin Chem 1984; 30:1700. Naylor GJ. Vanadium and affective disorders. Biol Psychiatry 1983; 18:103. Robins AH, Barron JL. The effect of disulfiram on the urinary excretion of catecholamine metabolites in alcoholic males. S Afr Med J 1983; 63:41. Hill P, et al. VIP and prolactin release in response to meals. Scand J Gastroenterol 1986; 21:958. Bloom SR, et al. Vasoactive intestinal peptide secreting tumors. Pathophysiological and clinical correlations. Ann N Y Acad Sci 1988; 527:518. Henriksen JH, et al. Circulating endogenous vasoactive intestinal polypeptide (VIP) in patients with uraemia and liver cirrhosis. Eur J Clin Invest 1986; 16:211. Woie L, et al. Increase in plasma vasoactive intestinal polypeptide (VIP) in muscular exercise in humans. Gen Pharmacol 1986; 17:321. Barreca T, et al. Plasma somatostatin and vasoactive intestinal polypeptide responses to an oral mixed test meal in obese patients. Horm Res 1989; 31:234. Stromberg P, et al. Vasopressin and prostaglandins in premenstrual pain and primary dysmenorrhea. Acta Obstet Gynecol Scand 1984; 63:533. Soelberg Sorensen P, Hammer M. Effects of long-term carbamazepine treatment on water metabolism and plasma vasopressin concentration. Eur J Clin Pharmacol 1984; 26:719. Dietz R, et al. Vasopressor systems during smoking in humans. Klin Wochenschr 1984; 62(Suppl 2):11. De Vane GW. Vasopressin levels during pregnancy and labor. J Reprod Med 1985; 30:324. Punnonen R, et al. Plasma vasopressin during normal menstrual cycle. Horm Res 1983; 17:90. Stewart WK, et al. Plasma retinol and retinol binding protein concentrations in patients on maintenance haemodialysis with and without vitamin A supplements. Nephron 1982; 30:15. Mejia LA, Arroyave G. Determination of vitamin A in blood. Some practical considerations on the time of collection of the specimens and the stability of the vitamin. Am J Clin Nutr 1983; 37:147. Kozlowski BW. Anticonvulsant medication use and circulating levels of total thyroxine, retinol binding protein, and vitamin A in children with delayed cognitive development. Am J Clin Nutr 1987; 46:360. Ito Y, et al. Effects of the consumption of cigarettes, alcohol and foods on serum concentrations of carotenoids, retinol and tocopherols in healthy inhabitants living in a rural area of Hokkaido. Nippon Eiseigaku Zasshi 1991; 46:874. Lewis CJ, et al. Relationship between age and serum vitamin A in children aged 4-11 y. Am J Clin Nutr 1990; 52:353. Leigh D, et al. Erythrocyte transketolase activity in the Wernicke-Korsakoff syndrome. Br J Psychiatry 1981; 139:153. Wyatt DT, et al. Age-dependent changes in thiamin concentrations in whole blood and cerebrospinal fluid in infants and children. Am J Clin Nutr 1991; 53:530. Leinert J, et al. Methods and their evaluation for the estimation of vitamin B6 levels in man. 5. Serum pyridoxal-5-phosphate determination: method and comparison of methods. [In German]. Int J Vitam Nutr Res 1983; 53:156. Shideler CE. Vitamin B6: an overview. Am J Med Technol 1983; 49:17. Ubbink JB, et al. Relationship between vitamin B-6 status and elevated pyridoxal kinase levels induced by theophylline therapy in humans. J Nutr 1990; 120:1352. Fahmy NR. Consumption of vitamin B12 during sodium nitroprusside administration in humans. Anesthesiology 1981; 54:305. Fairbanks VF, Elveback LR. Tests for pernicious anemia: serum vitamin B12 assay. Mayo Clin Proc 1983; 58:135. Hjelt K, et al. Oral contraceptives and the cobalamin (vitamin B12) metabolism. Acta Obstet Gynecol Scand 1985; 64:59. Fernandes-Costa F, et al. A sex difference in serum cobalamin and transcobalamin levels. Am J Clin Nutr 1985; 41:784. Saxena S, Carmel R. Racial differences in vitamin B12 levels in the United States. Am J Clin Pathol 1987; 88:95. Sasaki R, et al. Influences of sex and age on serum ascorbic acid. Tohoku J Exp Med 1983; 140:97. Sinha R, et al. Determinants of plasma ascorbic acid in a healthy male population. Cancer Epidemiol Biomarkers Prev 1992; 1:297. MacRury SM, et al. Seasonal and climatic variation in cholesterol and vitamin C: effect of vitamin C supplementation. Scott Med J 1992; 37:49. Rosen JF, Chesney RW. Circulating calcitriol concentrations in health and disease. J Pediatr 1983; 103:1. Halloran BP, et al. Serum concentration of 1,25-dihydroxyvitamin D in the human: diurnal variation. J Clin Endocrinol Metab 1985; 60:1104. Bikle DD, et al. Free 1,25-dihydroxyvitamin D levels in serum from normal subjects, pregnant subjects, and subjects with liver disease. J Clin Invest 1984; 74:1966. Jongen MJM, et al. An international comparison of vitamin D metabolite measurements. Clin Chem 1984; 30:399. Lichtenstein P, et al. Calcium-regulating hormones and minerals from birth to 18 months of age: a cross-sectional study. I. Effects of sex, race, age, season, and diet on vitamin D status. Pediatrics 1986; 77:883. Glass AR, Eil C. Ketoconazole-induced reduction in serum 1,25-dihydroxyvitamin D. J Clin Endocrinol Metab 1986; 63:766. Van der Wiel HE, et al. Biochemical parameters of bone turnover during ten days of bed rest and subsequent mobilization. Bone Miner 1991; 13:123. Villa ML, et al. Effects of aluminum hydroxide on the parathyroid-vitamin D axis of postmenopausal women. J Clin Endocrinol Metab 1991; 73:1256. Rodland O, et al. Serum levels of vitamin D metabolites in isotretinointreated acne patients. Acta Derm Venereol (Stockh) 1992; 72:217. Reasner CA 2nd, et al. Alteration of vitamin D metabolism in Mexican-Americans. J Bone Miner Res 1990; 5:13. Ala-Houhala M, et al. Serum 25-hydroxyvitamin D levels in Finnish children aged 2 to 17 years. Acta Paediatr Scand 1984; 73:232. Tjellesen L, et al. Serum vitamin D metabolites in epileptic patients treated with 2 different anti-convulsants. Neurol Scand 1982; 66:335. Goldray D, et al. Vitamin D deficiency in elderly patients in a general hospital. J Am Geriatr Soc 1989; 37:589. Scragg R, et al. Plasma 25-hydroxyvitamin D3 and its relation to physical activity and other heart disease risk factors in the general population. Ann Epidemiol 1992; 2:697. Haga P, et al. Plasma tocopherol levels and vitamin E/beta-lipoprotein relationships during pregnancy and in cord blood. Am J Clin Nutr 1982; 36:1200. Stein G, et al. Serum vitamin E levels in patients with chronic renal failure. Int J Artif Organs 1983; 6:285. Ogunmekan AO. Plasma vitamin E (alpha tocopherol) levels in normal children and in epileptic children with and without anticonvulsant drug therapy. Trop Geogr Med 1985; 37:175. Markowitz ME, et al. Circadian variations in serum zinc (Zn) concentrations: correlation with blood ionized calcium, serum total calcium and phosphate in humans. Am J Clin Nutr 1985; 41:689. Walker AM, et al. Carbonic anhydrase inhibitors induce elevations in human whole-blood zinc levels. Arch Ophthalmol 1984; 102:1785. Zumkley H, et al. Zinc metabolism during captopril treatment. Horm Metab Res 1985; 17:256. Ohno H, et al. Exercise-induced changes in blood zinc and related proteins in humans. J Appl Physiol 1985; 58:1453. Swanson CA, King JC. Reduced serum zinc concentration during pregnancy. Obstet Gynecol 1983; 62:313. Aggett PJ. Zinc metabolism in chronic renal insufficiency with or without dialysis therapy. Contrib Nephrol 1984; 38:95. Young DS. SI units for clinical laboratory data. JAMA 1978; 240:1618. Soldin SJ, et al. Biochemical basis of pediatric disease, 3rd ed. Washington: AACC Press, 1998. Soldin SJ, et al. Pediatric reference ranges, 2nd ed. Washington: AACC Press, 1997. Ladenson JH. Nonanalytical sources of variation in clinical chemistry results. In: Sonnenwirth AC, Jarrett L, eds. Gradwohl's clinical laboratory methods and diagnosis, 8th ed. St. Louis: CV Mosby, 1980:149. Teitz NW, ed. Clinical guide to laboratory tests, 3rd ed. Philadelphia: WB Saunders, 1995. Dufour DR. Causes and control of preanalytic variation. In: Kaplan LA, Pesce AJ, eds. Clinical chemistry: theory, analysis, and correlation, 3rd ed. St. Louis: CV Mosby, 1996:65. Young DS. Effects of preanalytical variables on clinical laboratory tests, 2nd ed. Washington: AACC Press, 1997. Siest G, et al., eds. Interpretation of clinical laboratory tests. Foster City, CA: Biomedical Publications, 1985. Tryding N, et al., eds. Drug effects in clinical chemistry 1992, 6th ed. Stockholm: Apoteksbolaget, Lakemedelsverket and Svensk Forening for Klinisk Kemi, 1992. Yamanishi H, et al. Stability of acetoacetate after venesection. Clin Chem 1993; 39:920. Pratt JH, et al. Racial differences in aldosterone excretion and plasma aldosterone concentrations in children. N Engl J Med 1989; 321:1152. Prummel MF, et al. The course of biochemical parameters of bone turnover during treatment with corticosteroids. J Clin Endocrinol Metab 1991; 72:382. Beneteau-Burnat B, et al. Serum angiotensin-converting enzyme in healthy and sarcoidotic children: comparison with the reference interval for adults. Clin Chem 1990; 36:344. Colantonio D, et al. A possible role of atrial natriuretic peptide in ethanol-induced acute diuresis. Life Sci 1991; 48:635. Fibiger W, et al. Cortisol and catecholamines changes as functions of time-of-day and self-reported mood. Neurosci Biobehav Rev 1984; 8:523. Weeke J, et al. The effect of heating and central cooling on serum TSH, GH, and norepinephrine in resting normal man. Acta Physiol Scand 1983; 117:33. Chen M, et al. Plasma catecholamines, dietary carbohydrate, and glucose intolerance: a comparison between young and old men. J Clin Endocrinol Metab 1986; 62:1193. Cryer PE. Phaeochromocytoma. Clin Endocrinol Metab 1985; 14:203. Tietz NW, et al. Laboratory values in fit aging individuals—sexagenarians through centenarians. Clin Chem 1992; 38:1167. Gallagher SK, et al. Short-term and long-term variability of indices related to nutritional status. I: Ca, Cu, Fe, Mg, and Zn. Clin Chem 1989; 35:369. Perrone RD, et al. Serum creatinine as an index of renal function: new insights into old concepts. Clin Chem 1992; 38:1933. Lew SQ, Bosch JP. Effect of diet on creatinine clearance and excretion in young and elderly healthy subjects and in patients with renal disease. J Am Soc Nephrol 1991; 2:856. Saad EA, et al. Peripheral plasma metenkephalin levels in ovulatory and anovulatory human menstrual cycles. Fertil Steril 1992; 58:307. Mozzanica N, et al. Association between circadian rhythms of endogenous hypothalamic opioid peptides and of natural killer cell activity. Int J Immunopharmacol 1991; 13:317. Lamb E, et al. Serum glycated albumin and fructosamine in renal dialysis patients. Nephron 1993; 64:82. Atkin SH, et al. Fingerstick glucose determination in shock. Ann Intern Med 1991; 114:1020. Weykamp CW, et al. Interference of carbamylated and acetylated hemoglobins in assays of glycohemoglobin by HPLC electrophoresis, affinity chromatography, and enzyme immunoassay. Clin Chem 1993; 39:138 Phillipou G, Phillips PJ. Intraindividual variation of glycohemoglobin: implications for interpretation and analytical goals. Clin Chem 1993; 39:2305. Weykamp CW, et al. Influence of hemoglobin variants and derivatives on glycohemoglobin determinations, as investigated by 102 laboratories using 16 methods. Clin Chem 1993; 39:1717. Kilpatrick ES, et al. Increased fetal hemoglobin in insulin-treated diabetes mellitus contributes to the imprecision of glycohemoglobin measurements. Clin Chem 1993; 39:833. Panteghini M, Pagani F. Biological variation of lactate and pyruvate in blood. Clin Chem 1993; 39:908. Isojarvi JI, et al. Thyroid function with antiepileptic drugs. Epilepsia 1992; 33:142. Cugini P, et al. Vasoactive intestinal peptide fluctuates in human blood with a circadian rhythm. Regul Pept 1991; 34:141. Cugini P, et al. The circadian rhythm of atrial natriuretic peptide, vasoactive i ntestinal peptide, betaendorphin and cortisol in healthy young and elderly subjects. Clin Auton Res 1992; 2:113. May RB, Sunder TR. Hematologic manifestations of long-term valproate therapy. Epilepsia 1993; 34:1098. Massoudi MS, et al. Prevalence of thyroid antibodies among healthy middle-aged women. Findings from the thyroid study in healthy women. Ann Epidemiol 1995; 5:229. Engler H, et al. Anti-thyroid peroxidase (anti-TPO) antibodies in thyroid diseases, non-thyroidal illness and controls. Clinical validity of a new commercial method for detection of anti-TPO (thyroid microsomal) autoantibodies. Clin Chim Acta 1994; 225:123. Romaldini JH, et al. Effect of L-thyroxine administration on antithyroid antibody levels, lipid profile, and thyroid volume in patients with Hashimoto's thyroiditis. Thyroid 1996; 6:183. Iles RK, et al. Cross-reaction with luteinizing hormone beta-core is responsible for the age-dependent increase of immunoreactive beta-core fragment of human chorionic gonadotropin in women with nonmalignant conditions. Clin Chem 1999; 45:532. Cole LA. Immunoassay of human chorionic gonadotropin, its free subunits, and metabolites. Clin Chem 1997; 43:2233. Jensen TB, et al. Library of sequence-specific radioimmunoassays for human chromogranin A. Clin Chem 1999; 45:549. Boomsma F, et al. Sensitivity and specificity of a new ELISA method for determination of chromogranin A in the diagnosis of pheochromocytoma and neuroblastoma. Clin Chim Acta 1995; 239:57. Takiyyuddin MA, et al. Chromogranin A in human hypertension. Influence of heredity. Hypertension 1995; 26:213. Canale MP, Bravo EL. Diagnostic specificity of serum chromogranin-A for pheochromocytoma in patients with renal dysfunction. J Clin Endocrinol Metab 1994; 78:1139. Inder WJ, et al. Prolonged exercise increases peripheral plasma ACTH, CRH, and AVP in male athletes. J Appl Physiol 1998; 85:835. Sorem KA, et al. Circulating maternal corticotropin-releasing hormone and gonadotropin-releasing hormone in normal and abnormal pregnancies. Am J Obstet Gynecol 1996; 175:912. Papanicolaou DA, et al. A single midnight serum cortisol measurement distinguishes Cushing's syndrome from pseudo-Cushing states. J Clin Endocrinol Metab 1998; 83:1163. Zaninotto M, et al. A proposal for standardizing urine collections for bone resorption markers measurement. J Clin Lab Anal 1998; 12:145. Aoshima H, et al. Circadian variation of urinary type I collagen crosslinked C-telopeptide and free and peptide-bound forms of pyridinium crosslinks. Bone 1998; 22:73. Ju HS, et al. Comparison of analytical performance and biological variability of three bone resorption assays. Clin Chem 1997; 43:1570. Bollen AM, et al. Circadian variation in urinary excretion of bone collagen cross-links. J Bone Miner Res 1995; 10:1885.

702. 703. 704. 705. 706. 707. 708. 709. 710. 711. 712. 713. 714. 715. 716. 717. 718. 719. 720. 721. 722. 723. 724. 725. 726. 727. 728. 729. 730. 731. 732. 733. 734. 735. 736. 737. 738. 739. 740. 741. 742. 743. 744. 745. 746. 747. 748. 749. 750. 751. 752. 753. 754. 755. 756. 757. 758. 759. 760. 761. 762. 763. 764. 765. 766. 767. 768. 769. 770. 771.

Mora S, et al. Urinary markers of bone turnover in healthy children and adolescents: age-related changes and effect of puberty. Calcif Tissue Int 1998; 63:369. Rotblatt MD, Koda-Kimble MA. Review of drug interference with urine glucose tests. Diabetes Care 1987; 10:103. Chance JJ, et al. Technical evaluation of five glucose meters with data management capabilities. Am J Clin Pathol 1999; 111:547. Ubbink JB, et al. Comparison of three different plasma homocysteine assays with gas chromatography-mass spectrometry. Clin Chem 1999; 45:670. Schwaninger M, et al. Elevated plasma concentrations of homocysteine in antiepileptic drug treatment. Epilepsia 1999; 40:345. Mijatovic V, et al. Randomized, double-blind, placebocontrolled study of the effects of raloxifene and conjugated equine estrogen on plasma homocysteine levels in healthy postmenopausal women. Fertil Steril 1998; 70:1085. Wright M, et al. Effect of acute exercise on plasma homocysteine. J Sports Med Phys Fitness 1998; 38:262. Lakshmi AV, Ramalakshmi BA. Effect of pyridoxine or riboflavin supplementation on plasma homocysteine levels in women with oral lesions. Natl Med J India 1998; 11:171. Bonnette RE, et al. Plasma homocyst(e)ine concentrations in pregnant and nonpregnant women with controlled folate intake. Obstet Gynecol 1998; 92:167. Cattaneo M, et al. Tamoxifen reduces plasma homocysteine levels in healthy women. Br J Cancer 1998; 77:2264. Clarke R, et al. Variability and determinants of total homocysteine concentrations in plasma in an elderly population. Clin Chem 1998; 44:102. Nygard O, et al. Coffee consumption and plasma total homocysteine: The Horda-land Homocysteine Study. Am J Clin Nutr 1997; 65:136. Arnadottir M, et al. Hyperhomocysteinemia in cyclosporine-treated renal transplant recipients. Transplantation 1996; 61:509. Cravo ML, et al. Hyperhomocysteinemia in chronic alcoholism: correlation with folate, vitamin B-12, and vitamin B-6 status. Am J Clin Nutr 1996; 63:220. Wouters MG, et al. Plasma homocysteine and menopausal status. Eur J Clin Invest 1995; 25:801. Guttormsen AB, et al. Plasma concentrations of homocysteine and other aminothiol compounds are related to food intake in healthy human subjects. J Nutr 1994; 124:1934. Ubbink JB, et al. The effect of blood sample aging and food consumption on plasma total homocysteine levels. Clin Chim Acta 1992; 207:119. Widjaja A, et al. Within- and between-subject variation in commonly measured anthropometric and biochemical variables. Clin Chem 1999; 45:561. Watt HC, et al. The pattern of maternal serum inhibin-A concentrations in the second trimester of pregnancy. Prenat Diagn 1998; 18:846. Petraglia F, et al. Changes of dimeric inhibin B levels in maternal serum throughout healthy gestation and in women with gestational diseases. J Clin Endocrinol Metab 1997; 82:2991. Yamoto M, et al. Serum levels of inhibin A and inhibin B in women with normal and abnormal luteal function. Obstet Gynecol 1997; 89:773. Watt HC, et al. Maternal serum inhibin-A levels in twin pregnancies: implications for screening for Down's syndrome. Prenat Diagn 1996; 16:927. Wald NJ, et al. Maternal serum inhibin-A in pregnancies with insulin-dependent diabetes mellitus: implications for screening for Down's syndrome. Prenat Diagn 1996; 16:923. Tenover JS, Bremner WJ. Circadian rhythm of serum immunoreactive inhibin in young and elderly men. J Gerontol 1991; 46:M181. Byrd W, et al. Regulation of biologically active dimeric inhibin A and B from infancy to adulthood in the male. J Clin Endocrinol Metab 1998; 83:2849. Welt CK, et al. Female reproductive aging is marked by decreased secretion of dimeric inhibin. J Clin Endocrinol Metab 1999; 84:105. Liberman CS, et al. Circulating iodide concentrations during and after pregnancy. J Clin Endocrinol Metab 1998; 83:3545. Hollowell JG, et al. Iodine nutrition in the United States. Trends and public health implications: iodine excretion data from National Health and Nutrition Examination Surveys I and III (1971–1974 and 1988–1994). J Clin Endocrinol Metab 1998; 83:3401. Siff JE, et al. Usefulness of the total iron binding capacity in the evaluation and treatment of acute iron overdose. Ann Emerg Med 1999; 33:73. Ertl T, et al. Postnatal changes of leptin levels in full-term and preterm neonates: their relation to intrauterine growth, gender and testosterone. Biol Neonate 1999; 75:167. Castracane VD, et al. Serum leptin concentration in women: effect of age, obesity, and estrogen administration. Fertil Steril 1998; 70:472. Kristrom B, et al. Short-term changes in serum leptin levels provide a strong metabolic marker for the growth response to growth hormone treatment in children. Swedish Study Group for Growth Hormone Treatment. J Clin Endocrinol Metab 1998; 83:2735. Johansen KL, et al. Leptin, body composition, and indices of malnutrition in patients on dialysis. J Am Soc Nephrol 1998; 9:1080. Mantzoros CS, et al. Circulating insulin concentrations, smoking, and alcohol intake are important independent predictors of leptin in young healthy men. Obes Res 1998; 6:179. Dubuc GR, et al. Changes of serum leptin and endocrine and metabolic parameters after 7 days of energy restriction in men and women. Metabolism 1998; 47:429. Saad MF, et al. Diurnal and ultradian rhythmicity of plasma leptin: effects of gender and adiposity. J Clin Endocrinol Metab 1998; 83:453. Lenders JW, et al. Plasma metanephrines in the diagnosis of pheochromocytoma. Ann Intern Med 1995; 123:101. Reinartz JJ, et al. Plastic vs glass SST evacuated serumseparator blood-drawing tubes for endocrinologic analytes. Clin Chem 1993; 39:2535. Metcalf PA, et al. Albuminuria in people at least 40 years old: effect of alcohol consumption, regular exercise, and cigarette smoking. Clin Chem 1993; 39:1793. Metcalf PA, et al. Dietary nutrient intakes and slight albuminuria in people at least 40 years old. Clin Chem 1993; 39:2191. Roberts WL, et al. Comparison of four commercial urinary albumin (microalbumin) methods: implications for detecting diabetic nephropathy using random urine specimens. Clin Chim Acta 1998; 273:21. James MA, et al. Microalbuminuria in elderly hypertensives: reproducibility and relation to clinic and ambulatory blood pressure. J Hypertens 1994; 12:309. Taskiran M, et al. Urinary albumin excretion in hospitalized patients with acute myocardial infarction. Prevalence of microalbuminuria and correlation to left ventricle wall thickness. Scand Cardiovasc J 1998; 32:163. Jensen HV, et al. Urinary excretion of albumin and transferrin in lithium maintenance treatment: daily versus alternate-day lithium dosing schedule. Psychopharmacology 1995; 122:317. Myrup B, et al. Effect of low-dose heparin on urinary albumin excretion in insulindependent diabetes mellitus. Lancet 1995; 345:421. Collins AC, et al. Storage temperature and differing methods of sample preparation in the measurement of urinary albumin. Diabetologia 1993; 36:993. Yokoyama H, et al. Changes of albumin concentrations in the first morning urine according to age and sex in 2990 healthy children and adults. Diabetes Res Clin Pract 1993; 21:167. Gerber LM, et al. Differences in urinary albumin excretion rate between normotensive and hypertensive, white and nonwhite subjects. Arch Intern Med 1992; 152:373. Shin SJ, et al. The variability of 24-hour, overnight 12-hour, and morning one-hour, urinary albumin excretion in normoalbuminuric and microalbuminuric hospitalized diabetics. Kao Hsiung I Hsueh Ko Hsueh Tsa Chih 1990; 6:511. Slag MF, et al. Free thyroxine levels in critically ill patients. A comparison of currently available assays. JAMA 1981; 246:2702. Oremek GM, Seiffert UB. Physical activity releases prostate-specific antigen (PSA) from the prostate gland into blood and increases serum PSA concentrations. Clin Chem 1996; 42:691. Sasagawa I, et al. Serum levels of total and free prostate specific antigen in men on hemodialysis. J Urol 1998; 160:83. Jung K, et al. Ratio of free-to-total prostate specific antigen in serum cannot distinguish patients with prostate cancer from those with chronic inflammation of the prostate. J Urol 1998; 159:1595. Ornstein DK, et al. The effect of prostate volume, age, total prostate specific antigen level and acute inflammation on the percentage of free serum prostate specific antigen levels in men without clinically detectable prostate cancer. J Urol 1998; 159:1234. Ornstein DK, et al. Biological variation of total, free and percent free serum prostate specific antigen levels in screening volunteers. J Urol 1997; 157:2179. Nixon RG, et al. Variation of free and total prostate-specific antigen levels: the effect on the percent free/total prostate-specific antigen. Arch Pathol Lab Med 1997; 121:385. Woodrum D, et al. Stability of free prostate-specific antigen in serum samples under a variety of sample collection and sample storage conditions. Urology 1996; 48(6A Suppl):33. Deinum J, et al. Improved immunoradiometric assay for plasma renin. Clin Chem 1999; 45:847. Orwoll ES, et al. Collagen N-telopeptide excretion in men: the effects of age and intrasubject variability. J Clin Endocrinol Metab 1998; 83:3930. Gertz BJ, et al. Application of a new serum assay for type I collagen cross-linked N-telopeptides: assessment of diurnal changes in bone turnover with and without alendronate treatment. Calcif Tissue Int 1998; 63:102. Sone T, et al. Urinary excretion of type I collagen crosslinked N-telopeptides in healthy Japanese adults: age- and sex-related changes and reference limits. Bone 1995; 17:335. Aloia JF, et al. Biochemical and hormonal variables in black and white women matched for age and weight. J Lab Clin Med 1998; 132:383. Nahoul K, Roger M. Age-related decline of plasma bioavailable testosterone in adult men. J Steroid Biochem 1990; 35:293. Belgorosky A, et al. Progressive increase in non-sex-hormone-binding globulinbound testosterone from infancy to late prepuberty in boys. J Clin Endocrinol Metab 1987; 64:482. Winters SJ, et al. The analog free testosterone assay: are the results in men clinically useful? Clin Chem 1998; 44:2178. Rosner W. Errors in the measurement of plasma free testosterone. J Clin Endocrinol Metab 1997; 82:2014. Spencer CA, et al. Serum thyroglobulin autoantibodies: prevalence, influence on serum thyroglobulin measurement, and prognostic significance in patients with differentiated thyroid carcinoma. J Clin Endocrinol Metab 1998; 83:1121. Hawkins RC. Furosemide interference in newer free thyroxine assays. Clin Chem 1998; 44:2550. Wilke TJ. Estimation of free thyroid hormone concentrations in the clinical laboratory. Clin Chem 1986; 32:585. Chopra IJ. Thyroid function in nonthyroidal illnesses. Ann Intern Med 1983; 98:946.

CHAPTER 238 TECHNIQUES OF LABORATORY TESTING Principles and Practice of Endocrinology and Metabolism

CHAPTER 238 TECHNIQUES OF LABORATORY TESTING D. ROBERT DUFOUR Spectrophotometry Enzymatic Methods Atomic Absorption Spectrophotometry Electrochemistry Immunologic Methods Immunoassays Immunometric Assays Methods of Developing Antibody Antibody Specificity Antigenicity Versus Bioactivity Competitive Protein Binding and Radioreceptor Assays Assays for Free Hormone Chromatography Electrophoresis Bioassay Common Interferences in Endocrine Testing Glucose Hemoglobin A1c Free Thyroxine Chapter References

The endocrinologist must be cognizant of the common methods used to measure the concentration of endocrine compounds in blood and/or body fluids. A single substance often can be measured by more than one method. In some laboratories, two or three different methods may be used to measure the same analyte. The choice of method may depend on the time of day at which the specimen is received, because some methods may be easier to perform and, thus, may be used for emergency or near-patient testing. Occasionally, this may lead to differing results, because interfering substances commonly do not react identically to the compound of interest and, thus, do not have the same effect on apparent concentrations. For example, one method for measuring bicarbonate shows a positive interference from salicylate, but salicylate does not affect results obtained with most other methods. If bicarbonate were determined using the former method, either no anion gap or, possibly, a negative anion gap, would be present, which would lead to markedly different interpretations of the patient's condition. No method is completely free of all interferences or is absolutely specific for the compound of interest. This chapter reviews the most common techniques used in laboratories and their limitations. Methods using nucleic acid measurement are not discussed in this chapter but are covered in Chapter 240. In this chapter, in contrast to their use in Chapter 237, the terms sensitivity and specificity refer to analytic sensitivity (ability to detect a small amount of a substance) and analytic specificity (ability to detect a substance in the presence of interfering compounds).

SPECTROPHOTOMETRY Spectrophotometry and its variants are used for the majority of simple analyses in the clinical laboratory. A spectrophotometer is an instrument that allows light of a specific wavelength to pass through a solution to a detector. Spectrophotometric techniques rely on a direct linear relationship between the concentration of a substance and the amount of light absorbed when passing through the solution. This relationship, called Beer's law, holds true when substances are present in pure form in an aqueous solution. In biologic fluids, the multitude of substances present poses a difficult problem for the analyst. If a substance is strongly colored and is present in high concentration, the amount of light absorbed can be directly measured; this is the technique used for hemoglobin determinations and for bilirubin measurement on some instruments. For most substances, however, this simple approach does not work, because many compounds in addition to the substance of interest absorb light at the same wavelength. The most common approach to eliminate these interferences is to combine the substance of interest with another compound to give a highly colored reaction product that absorbs light at an order of magnitude greater than that of any interfering substances. For many analytes, such as glucose, calcium, and phosphate, this step is enough to give the measurement adequate specificity. For example, glucose combines with o-toluidine to give a product with an intense blue color. The apparent concentration of glucose measured by this reaction is within 10% of the value obtained using the accepted reference method.1 For other substances, the method is specific enough to use for routine analyses, but some commonly encountered compounds also react. An example of such a reaction is that of creatinine with picrate, the Jaffe reaction, a method used by the great majority of laboratories to measure creatinine.2 In urine and in most sera, the method provides an accurate approximation of the creatinine concentration. Ketone bodies cross-react in the most commonly used application of this method3; in patients with alcoholic or diabetic ketoacidosis, it is not uncommon for the apparent concentration of creatinine to be markedly elevated, leading to a false perception of renal insufficiency (Fig. 238-1). For other substances (e.g., urinary steroid and catecholamine metabolites), a large number of drugs and, for vanillylmandelic acid, food substances, cross-react, so that the interpretation of simple spectrophotometric methods is difficult at best. A number of modifications can increase the specificity of spectrophotometry, allowing its use for measurement of a larger number of compounds.

FIGURE 238-1. Changes in blood urea nitrogen (BUN) and creatinine in a patient with diabetic ketoacidosis. Point A represents the time at which serum ketones began to rise, and point B represents the last time that serum ketones were positive at a dilution of 1:32 or greater. At point C, ketones were no longer detectable. Changes in serum creatinine parallel changes in ketones, but BUN remains virtually unchanged during the entire episode. Because the chemical reaction used in most creatinine assays is nonspecific, ketone bodies produce the same colored product.

ENZYMATIC METHODS Enzymatic reactions can be used to confer added specificity to chemical reactions. Enzymes are similar to antibodies, in that they have a reactive site which recognizes a specific structural or chemical sequence. Occasionally, other compounds with a similar structure may be cleaved by the enzyme. An example is the cross-reaction of isopropanol, commonly used to prepare the skin for venipuncture, in the measurement of ethanol by alcohol dehydrogenase.4 In an enzymatic assay, the rate of appearance of a product of the enzyme-catalyzed reaction is directly related to the concentration of the compound of interest, provided all other components of the reaction are present in standardized amounts. The assumption that only the compound of interest is in variable concentration is often not correct, causing erroneous results for some enzymatic measurements. As is discussed later, oxygen is an important variable in the measurement of glucose by glucose oxidase, and changes in the oxygen content of blood can affect measurement of glucose.5 Most enzyme measurements involve a sequence of reactions, with only the final product being measured. If another substance reacts in one of the steps, or interferes with one of the reactions, the result will be inaccurate. This is an occasional problem with urine containing large amounts of ascorbic acid, an antioxidant that prevents production of the colored final product used for detecting glucosuria.6

ATOMIC ABSORPTION SPECTROPHOTOMETRY

The most specific spectrophotometric method is atomic absorption spectrophotometry, which uses a lamp containing a strip of the metal being measured. Atomic absorption is both extremely sensitive and specific for the elements that are measured, and is a reference technique for the assay of metals such as calcium and magnesium. It can be used only for metals, limiting its applicability, and the lengthy preparation time limits the number of specimens that can be analyzed. The extremely low concentrations of some trace metals make their measurements very dependent on careful technique to prevent contamination, which can occur from such diverse sources as collection tubes, wooden applicator sticks, and even the glass used in the collection of blood or the preparation of the specimens.7

ELECTROCHEMISTRY Another common technique is electrochemistry, in which molecules or ions either produce a chemical reaction that generates electrons, or interact with a selective membrane and create a difference in electrical potential. Electrochemical methods are commonly used to measure blood gases and electrolytes. They are highly specific for the substance being measured. To avoid interferences, most sodium electrochemical methods dilute specimens before measurement; thus, they measure the sodium concentration in plasma. The body regulates osmolality, which is related to the sodium concentration in water rather than to the total plasma concentration. In normal individuals, plasma is composed of 93% water and 7% solids (e.g., proteins and lipoproteins). In situations in which protein or lipid comprises a larger percentage of plasma (e.g., in multiple myeloma or severe hypertriglyceridemia), methods that measure the plasma sodium concentration produce falsely low results, termed pseudohyponatremia. (This phenomenon is also observed with older flame photometric techniques, which also required specimen dilution.) Some newer instruments, especially those using whole blood measurements (as are commonly found in critical care settings) measure sodium activity (related to the sodium concentration in water) directly without dilution; pseudohyponatremia is not observed with such methods.8

IMMUNOLOGIC METHODS The use of antibodies as reagents has allowed measurement of many compounds that cannot be assayed by other methods. Antibodies have high specificity in their reactions, minimizing (but not eliminating) cross-reactions. Antibody techniques can detect extremely small concentrations of substances, far below what can be measured with other methods. Detection of a reaction between antigen and antibody is based on labels that are placed on either the antibody or the antigen. Among the more common labels are radioactive isotopes, enzymes, fluorescent compounds, and chemiluminescent molecules. Two basic forms of assay use antibodies to measure compounds. Immunoassays use a small amount of antibody that can bind either labeled antigen (from the reagents) or unlabeled antigen in the specimen. This technique is useful for measuring small compounds (such as thyroid and steroid hormones). Immunometric assays use antibodies to form “sandwiches” with antigen in the middle. Typically, immunometric assays use two different antibodies that recognize different parts of the molecule, markedly improving specificity. Label is placed on the second antibody; thus, the amount of label bound is directly proportional to the amount of antigen in the sample. This type of assay is used with larger antigens that have more than one antibody-binding site (e.g., peptide hormones). An illustration of these two different principles of immunoassay is given in Figure 238-2.

FIGURE 238-2. The common feature of all immunoassays is that a limited amount of labeled antigen competes for antibody binding sites with unlabeled antigen present in the standards or in a patient's serum. A, In homogeneous immunoassays, the label behaves differently if bound to antibody or if free within solution; therefore, no separation step is necessary, simplifying the measurement. This is the principle of enzyme-multiplied immunoassays and fluorescence polarization inhibition assays, which are the simplest immunoassay techniques to perform. B, In heterogeneous immunoassays, the remaining free labeled antigen must be separated from that bound to antibody before measurement. This is most often accomplished by precipitating unbound antigen by binding it to a nonspecific absorbent, such as charcoal or a resin. Alternative methods include precipitating antigen–antibody complexes and fixing antigen or antibody to a solid medium, such as the sides of a test tube or small beads. This methodology is typical of radioimmunoassay and fluorescence immunoassay. C, Antibody may be labeled instead of labeling the antigen. This is the principle of immunometric methods, such as enzyme-linked immunosorbent assay and immunoradiometry. The major advantages are that antibodies are generally easier to label, and sensitivity and specificity can be improved by using two antibodies.

IMMUNOASSAYS The basic principle underlying immunoassays is that a labeled compound competes with unlabeled compound present in the patient's serum for a limited number of binding sites on the antibody molecules. In radioimmunoassay (RIA), the label is a radioactive isotope, usually iodine or tritium; however, the label can also be a fluorescent compound (as in fluorescence immunoassays [FIA] or fluorescence polarization inhibition assays [FPIA]), or an enzyme (as in enzyme-multiplied immune technique [EMIT] or enzyme-linked immunosorbent [ELISA] assays), or a chemiluminescent compound. Generally, chemiluminescent and radioisotopic methods have the highest sensitivity. The higher the concentration of unlabeled compound in the patient's serum, the less labeled compound will bind to the antibody. After an incubation period to allow for equilibrium to be reached, the amount of label attached to the antibody is measured. In many methods, a separation step is required to remove any residual free labeled compound; such methods are called heterogeneous, because they measure label in only one of (at least) two separate specimens. With some labels, the bound and free labeled compounds behave differently, so that no separation step is required; such methods are termed homogeneous assays. IMMUNOMETRIC ASSAYS In immunometric assays, typically the first antibody is bound to a solid support. The sample is incubated with the bound antibody; after equilibrium is reached, the solid support is washed. A second, labeled antibody is then added and becomes bound to the solid support only if antigen is bound to the first antibody. After washing, the amount of labeled antibody bound to the solid is measured. Typically, this technique is used for large molecules that have more than one different antigenic site (epitope) that can be recognized by antibodies. By using antibodies against two different epitopes, immunometric assays are highly specific and typically show few interferences. METHODS OF DEVELOPING ANTIBODY Antibodies are typically obtained by immunizing animals with the substance of interest. For small molecules, such as drugs, steroid hormones, and thyroid hormones, the molecule often must be attached to another substance (hapten) to make it more immunogenic. For larger molecules, immunization with the compound itself is usually adequate to provoke synthesis of antibodies. If serum from the animal injected with the compound is absorbed with human specimens lacking the compound, the antibody can be obtained in relatively pure form. With large molecules, the antibody typically consists of a mixture of antibodies (produced by different plasma cells) against different epitopes on the molecule; such a mixture is termed polyclonal. Different animals from the same species may make antibody of differing mixtures that react differently with the molecule of interest. This may cause significant variation in measurements from one lot of immunoassay reagent to the next. This is particularly problematic for molecules that have multiple antigenic sites, such as peptide hormones.9 An alternative method to develop antibodies is typically used with mice. After immunizing the mouse, the animal is sacrificed and the spleen is taken to isolate plasma cells. Single plasma cells are fused with myeloma cells to make a new cell line, termed a hybridoma, which produces antibody against a single epitope in a reproducible fashion. The hybridomas are grown in tissue culture, and the monoclonal antibody produced by each is tested for reactivity against the molecule of interest. Once a hybridoma producing a monoclonal antibody of suitable specificity is identified, a virtually limitless supply can be produced with identical reactivity against the molecule of interest. Although antibody methods are usually quite accurate, they are prone to several types of interferences and inaccuracies that may cause difficulty in interpreting their

results. ANTIBODY SPECIFICITY Although some antibodies are highly specific and can differentiate compounds that vary only in the presence of a single constituent atom (e.g., triiodothyronine [T3] and thyroxine [T4]), many antibodies lack specificity in reactivity. Three common examples are the antibody for cortisol, which cross-reacts with endogenous and exogenous glucocorticoids10; the antibody used to detect amphetamine in urine, which cross-reacts with any number of prescription and over-the-counter drugs having a structure similar to amphetamine11; and the antibody to digoxin, which cross-reacts with inactive digoxin metabolites and endogenous digoxin-like substances.12 Such cross-reactivity is not typically seen with immunometric assays. Some patients' sera contain antibodies that react with immunoglobulins from other species. These are termed heterophile antibodies. This is most commonly seen in patients who have been treated with monoclonal antibodies (for example, in tumor localization or for treatment of malignancy), such that antibodies develop against mouse proteins.13,14 These human antibodies against mouse antibodies (HAMA) may also be found in those working with laboratory mice and in a small percentage of the general population. HAMAs are a major cause of interference in immunometric assays if they are present in high titers and may cause falsely elevated results by binding the labeled indicator antibody to the solid support. This may affect any substance measured by immunometric assays, including most peptide hormones, tumor markers, and other substances. Less commonly, rheumatoid factor (an autoantibody that reacts with the Fc portion of antibody molecules) may be able to react with immunoglobulins from other species, producing falsely elevated results in immunometric assays. Finally, autoantibodies may be present in serum that compete with the antibody in the reagent for binding to the molecule of interest. Common examples of autoantibodies are antibodies to thyroglobulin, a major source of interference in the use of thyroglobulin as a tumor marker15; antiinsulin antibodies (present in many diabetics) that interfere in insulin assays; and anti–thyroid hormone antibodies, which are found in a small percentage of patients with autoimmune thyroid disease.16 Such autoantibodies typically produce falsely elevated results in immunoassays and falsely low results in immunometric assays. ANTIGENICITY VERSUS BIOACTIVITY Although antibodies detect the presence of an antigen, they indicate nothing about the bioactivity of the antigen being measured. Although this is not likely to pose a problem for small hormones such as T4, it is of considerable importance in the measurement of peptide hormones, which may circulate as large-molecular-weight, and often inactive, prohormones, or which may be cleaved to metabolites that have little or no bioactivity. This problem became apparent early in the development of immunoassays, when it was noted that most patients with lung carcinomas had elevated serum adrenocorticotropic hormone levels but usually had no evidence of adrenal hyperfunction. This is due to the presence of a larger, inactive form of the hormone that cross-reacts in the assay.17 Other hormones (e.g., calcitonin, insulin, and glucagon) also have inactive precursors in the circulation. The now seldom-used midmolecule assays for parathyroid hormone (PTH) predominantly measure an inactive fragment that accumulates whenever renal function is impaired.18 Other hormones also may have inactive metabolites that lead to overestimation of the amount of active hormone present. Occasionally, with immunometric assays, such circulating inactive fragments may cause falsely low results for a hormone. For example, the immunometric assay for intact parathyroid hormone uses a bound antibody that recognizes epitopes in the C-terminal portion of the hormone, whereas the labeled second antibody recognizes epitopes in the N-terminal end of the molecule. In renal failure, high circulating levels of inactive fragments may compete for binding sites with parathyroid hormone, reducing its binding. Because the inactive fragments lack the N-terminal epitope, falsely low PTH values may be seen.19,20

COMPETITIVE PROTEIN BINDING AND RADIORECEPTOR ASSAYS Two other techniques are similar in principle to radioimmunoassay. Competitive protein binding uses a natural serum carrier protein such as corticosteroid-binding globulin or thyroid-binding globulin, and a labeled hormone that competes for the same binding sites as hormone in the patient's serum. An example is the formerly used Murphy-Patee test for T 4. This technique is virtually limited to steroid and thyroid hormones, because most peptide hormones do not have carrier proteins. Competitive protein-binding assays are subject to much the same interferences as are immunoassays; in addition, many are less specific, showing cross-reactions with a larger number of related compounds than is true for immunoassay. Competitive protein-binding techniques are still widely used for measurement of vitamin B12 and folic acid, and in preparation for measurement of vitamin D metabolites. In the case of vitamin B12, the use of intrinsic factor is necessary to assure that the assay measures only the active vitamin and not a number of inactive related cobalamins. In radioreceptor assays, the binder is a cellular receptor, either in intact cells or purified in some subcellular fraction. Radioreceptor assays have the theoretical advantage of recognizing only the active part of the hormone, thereby eliminating one of the limitations of immunoassays. Radioreceptor assays can be used for steroid, peptide, and catecholamine classes of hormones. Thus, they virtually have the same applicability as immunoassays. Furthermore, both naturally occurring and synthetic hormones can be measured with radioreceptor techniques. The basic principle is the same as in immunoassay, namely, competition for binding sites between radioactively labeled hormone and unlabeled hormone in the patient's serum. Radioreceptor assays have not found widespread applicability because of their lack of specificity. For example, thyroid-stimulating hormone (TSH), luteinizing hormone, and human chorionic gonadotropin all combine with TSH receptors, and prolactin, growth hormone, and human placental lactogen all bind to cellular prolactin receptors. In addition, the small number of binding sites often reduces sensitivity below what can be achieved with immunoassays.

ASSAYS FOR FREE HORMONE Because the free hormone often is the bioactive form and, commonly, protein-bound hormones are an inactive reservoir, a great deal of attention has been paid to the development of assays for free hormone. This has led to a search for other body fluids that might accurately reflect free hormone levels. Salivary hormone levels have been noted to correlate extremely well with free plasma levels for most steroid hormones analyzed. Salivary hormone concentrations are independent of the rate of production of saliva and show the same circadian and other rhythmic variations as do plasma hormone concentrations. To date, these assays have not gained wide acceptance, but they likely will become more important in the future. (For more details, see the excellent review in ref. 21.)

CHROMATOGRAPHY Chromatographic methods use differences in the solubility of compounds to achieve their separation (Fig. 238-3; Table 238-1). In typical chromatography, a solid support medium is used, usually made of a polar material. Relatively nonpolar liquid is passed over or through the solid support, and compounds separate based on their relative solubility in the stationary or mobile phases. The time for a compound to be removed from the solid support (eluted) is characteristic for each compound. Variants of chromatography include reversed-phase chromatography, in which the stationary phase is nonpolar and the liquid phase is polar; affinity chromatography, in which the stationary phase contains chemical sequences that bind only to specific molecules (for example, boronate binds glucose-protein linkages and is used for total glycohemoglobin measurement); and gas-liquid chromatography, in which the stationary phase is a liquid and the mobile phase is an inert gas, so that compounds are separated on the basis of their volatility. Chromatographs can be connected to mass spectrometers, allowing separated compounds to be identified based on their masses and pattern of fragmentation by electrons.

FIGURE 238-3. In all chromatographic methods, substances are separated based on whether they have a greater affinity for the stationary or the mobile phase. Compound A is similar to the stationary phase in one physical property (“straight sides”) and is separated from compound B as the mobile phase passes over the stationary phase.

TABLE 238-1. Types of Chromatographic Separation

Chromatography can both separate and quantify structurally related substances. Although such methods are reliable, accurate, specific, and highly sensitive (orders of magnitude more sensitive than immunoassays when coupled to mass spectrographic detectors), they are not commonly used in clinical laboratories because of their relatively high cost per test, the often extensive sample preparation required before analysis, and the small number of specimens that can be analyzed per day. They are commonly used in the measurement of catecholamine and serotonin and their metabolites, steroid hormone metabolites, vitamin D metabolites, and glycated hemoglobins.

ELECTROPHORESIS Electrophoretic techniques separate compounds on the basis of differences in their charge density and, particularly with polyacrylamide gel, also on the basis of molecular weight. The basic technique of electrophoresis involves a usually inert supporting gel on which the sample is applied in the presence of a buffer to control pH. When the sample is placed in an electrical field, any charged compounds migrate, with their distance of migration primarily determined by charge density. After separation of the compounds is complete, a variety of reactive chemicals can be applied to the gel to detect the presence of the compounds of interest. For example, in protein electrophoresis, a protein stain is used; for lipoprotein electrophoresis, an enzymatic assay measuring cholesterol is used; and in alkaline phosphatase isoenzyme electrophoresis, the chemical reaction for measuring alkaline phosphatase in serum is used. Quantification of the amount of each of the separated forms uses a spectrophotometric technique termed densitometry, in which the intensity of each separate band of reactivity is measured. Quantification by electrophoresis is not highly accurate or reproducible; when available, more specific methods are typically used.

BIOASSAY Although bioassay was one of the first modes of testing used in endocrinology, more easily performed techniques such as immunoassay have almost totally replaced this method. Interest has been renewed, however, in new and highly sensitive in vitro bioassays for hormones. In a typical in vitro bioassay for a hormone-activating adenylate cyclase, the change in cyclic adenosine monophosphate (cAMP) in the cell culture medium is used as an indicator of active hormone concentration. The advantage of bioassays is that they directly measure active hormone. In vitro bioassays have been developed for most hormones. An example is a bioassay for PTH using cells from an osteosarcoma. The test for thyroid-stimulating immunoglobulins is, in effect, a bioassay, because the final product measured is cAMP induced by the antibody–TSH receptor interaction. Such assays are not yet commercially available for most hormones but, conceivably, could replace immunoassays after they have been validated by wider use. Their main theoretical advantage is that they should detect only a specific active hormone; however, at least in the case of PTH, PTH-related peptide also binds to the PTH receptor and activates adenylate cyclase.22,23

COMMON INTERFERENCES IN ENDOCRINE TESTING Although a few specific methodologic interferences have been discussed, for several common tests the endocrinologist must have a clear understanding of the differences between methods of analysis or the reliability of results obtained by specific methods. Although summaries of interferences with different methods are given in Table 237-1, a few of the more common assays that may produce misleading results are given below. GLUCOSE As discussed earlier, most methods for measuring glucose use enzymatic assays. The most widely used of these is the glucose oxidase assay. The chemical reactions for glucose oxidase are shown below.

In serum, the amount of oxygen is trivial and primarily based on the oxygen content of air at the time of measurement; as such, it is constant from one specimen to another. In whole blood, as commonly used in fingerstick glucose measurement, however, the amount of oxygen in the specimen is dependent on the PO2 of the specimen and its hematocrit. With spectrophotometric methods, the amount of the colored reaction product is directly related to both the amount of glucose and the amount of oxygen, whereas with electrochemical methods the amount of the electrical signal is directly related to glucose and inversely related to the oxygen content of blood. In patients tested at home, usually little day-to-day variation exists in either hematocrit or oxygen content of blood, and results are fairly reliable. In hospitalized patients, particularly those critically ill, significant abnormalities of both oxygen content and hematocrit are often present; this frequently causes significant differences in measured glucose between serum specimens and whole blood specimens. Many manufacturers of glucose-testing strips have moved away from using glucose oxidase for this reason. A remaining limitation of whole blood assays for glucose is the difference between water content of whole blood and plasma. As mentioned earlier, plasma is 93% water, whereas whole blood is only ~80% water. The glucose content of whole blood is, thus, lower than that of serum. Manufacturers have calibrated their assays to give results similar to those for serum specimens by putting in a correction factor that is based on presence of a normal hematocrit. In patients with an abnormal hematocrit, the correction factor is erroneous, and glucose results diverge from those of serum. When the hematocrit is high, results are falsely low, whereas anemia causes falsely high results.24 HEMOGLOBIN A1C A number of different methods are used in laboratories to measure glucose attachment to hemoglobin. When glucose attaches to the N-terminal valine on the b chain of hemoglobin A, it produces hemoglobin A1c. Other substitutions at the same position produce other variants of hemoglobin A (collectively termed total hemoglobin A1). Glucose can also attach to other amino groups on hemoglobin, particularly lysine residues; all such substitutions produce glycohemoglobin molecules. Normally, hemoglobin A1c represents ~80% of total glycohemoglobin. Different assays for “hemoglobin A1c” measure different modifications of hemoglobin; formerly, results from different assays were not easily compared to each other. The Glycohemoglobin Standardization Program has significantly improved agreement between laboratories, making differences in methods less obvious for most diabetic patients.25 Because some situations exist in which differences between methods do occur, endocrinologists still must be aware of these differences and know the method used by their laboratory. The most commonly used methods are high-performance, ion-exchange chromatography (HPLC) techniques. Although HPLC methods are generally the most reproducible and were used in the Diabetes Control and Complications Trial, rare situations exist in which results are not reliable. The most common is the lack of

hemoglobin A, as occurs most commonly in persons with SS, SC, or CC hemoglobinopathies. Less commonly, congenital variants of hemoglobin (such as hemoglobin Raleigh26) have substitutions on the N-terminal valine that lead to a pattern of elution identical to that of hemoglobin A1c. Assays for total glycohemoglobin are used by many laboratories, particularly those in smaller hospitals, because of their ease of automation. A mathematical conversion is used to approximate the true hemoglobin A1c concentration from total glycohemoglobin results. The conversion formula appears to underestimate hemoglobin A1c values in patients with poorly controlled diabetes. In situations in which HPLC methods are not useful, total glycohemoglobin assays may still provide clinically useful information. Immunoassays for hemoglobin A1c do not produce inaccurate results in the rare hemoglobin variants with substitutions at the N-terminal valine.26 Data are inadequate to assess their utility in patients with homozygous hemoglobin variants, but they appear to produce reliable results in patients with heterozygous hemoglobin variants. A few laboratories still use simple ion-exchange chromatography or electrophoretic techniques; these typically measure other hemoglobin A1s, and often give falsely elevated results in patients with high levels of hemoglobin F. FREE THYROXINE Several different methods are used to measure free T4. The accepted reference method is equilibrium dialysis, in which a sample of serum is placed inside a dialysis bag, free T4 is allowed to diffuse out of the bag and establish an equilibrium with the amount inside the bag, and the free T4 concentration on the outside is measured. This technique assumes an equilibrium constant for the binding of thyroid hormone to its binding proteins. If the affinity of binding proteins were reduced, then this technique could overestimate free T4. Free T4 values measured by this technique are significantly higher in patients with acute, nonthyroidal illness than in normal individuals, perhaps reflecting this mechanism. The most widely used free T4 assay involves use of an analog of T4 in an immunoassay. The proprietary analog is said not to bind to thyroid-binding globulin, although evidence suggests that it does bind to albumin. In acute illness, a high percentage of patients have low free T4 results by this assay. An increasing number of laboratories use a technique termed immune extraction, in which serum is briefly incubated with an antibody. After washing, labeled T4 is added, and the amount of label bound to the antibody is inversely related to free T4 in the sample. Such methods agree better with equilibrium dialysis results, although elevated levels are virtually never seen in acute illness.27 The foregoing are three of the more common examples of how methodology may cause significant variations in results in patients with different clinical findings. A more complete listing of such methodologic considerations can be found in Table 237-1. CHAPTER REFERENCES 1. Passey RB, Gillum RL, Fuller JB, et al. Evaluation and comparison of 10 glucose methods and the reference method recommended in the proposed product class standard (1974). Clin Chem 1977; 23:131. 2. College of American Pathologists' Comprehensive Chemistry Survey. College of American Pathologist: Northfield, IL 1986. 3. Watkins PJ. The effect of ketone bodies on the determination of creatinine. Clin Chim Acta 1967; 18:191. 4. Vasiliades J. Emergency toxicology: the evaluation of three analytical methods for the determination of misused alcohols. Clin Toxicol 1977; 10:399. 5. Maser RE, Butler MA, DeCherney GS. Use of arterial blood with bedside glucose reflectance meters in an intensive care unit: are they accurate? Crit Care Med 1994; 22:595. 6. Rotblatt MD, Koda-Kimble MA. Review of drug interference with urine glucose tests. Diabetes Care 1987; 10:103. 7. Ericson SP, McHalsky ML, Rabinow BE, et al. Sampling and analysis techniques for monitoring serum for trace elements. Clin Chem 1986; 32:1350. 8. Ladenson JH, Apple FS, Koch DD. Misleading hyponatremia due to hyperlipemia: a method-dependent error. Ann Intern Med 1981; 95:707. 9. Fuentes-Arderiu J, Navarro-Moreno MA. Errors in interpreting hormonal assays. Ann Intern Med 1986; 105:798. 10. Nolan GE, Smith JB, Chavre VJ, Jubiz W. Spurious overestimation of plasma cortisol in patients with chronic renal failure. J Clin Endocrinol Metab 1981; 52:1242. 11. Budd RD. Amphetamine EMIT—structure versus reactivity. Clin Toxicol 1981; 18:91. 12. Soldin SJ. Digoxin—issues and controversies. Clin Chem 1986; 32:5. 13. Boscato LM, Stuart MC. Incidence and specificity of interference in two-site immunoassays. Clin Chem 1986; 32:1491. 14. Kricka LJ. Human anti-animal antibody interferences in immunological assays. Clin Chem 1999; 45:942. 15. Spencer CA, Takeuchi M, Kazarosyan M, et al. Serum thyroglobulin autoantibodies: prevalence, influence on serum thyroglobulin measurement, and prognostic significance in patients with differentiated thyroid carcinoma. J Clin Endocrinol Metab 1998; 83:1121. 16. Despres N, Grant AM. Antibody interference in thyroid assays: a potential for clinical misinformation. Clin Chem 1998; 44:440. 17. Ayvazian LF, Schneider B, Gewirtz G, Yalow RS. Ectopic production of big ACTH in carcinoma of the lung. Its clinical usefulness as a biologic marker. Am Rev Respir Dis 1975; 111:279. 18. Freitag J, Martin KJ, Hruska KA, et al. Impaired parathyroid hormone metabolism in patients with chronic renal failure. N Engl J Med 1978; 298:29. 19. Dilena BA, White GH. Interference with measurement of intact parathyrin in serum from renal dialysis patients. Clin Chem 1989; 35:1543. 20. Blind E, Raue F, Reichel H, Schmidt-Gayk H. Validity of intact plasma parathyrin measurements in chronic renal failure as determined by two-site immunoradiometric assays with N- or C-terminal capture antibodies. Clin Chem 1992; 38:2345. 21. Riad-Fahmy D, Read GF, Walker RF, Griffiths K. Steroids in saliva for assessing endocrine function. Endocr Rev 1982; 3:367. 22. Mundy GR, Ibbotson KJ, D'Souza SM, et al. The hypercalcemia of cancer. Clinical implications and pathogenic mechanisms. N Engl J Med 1984; 310:1718. 23. Merendino JJ Jr, Insogna KL, Milstone LM, et al. A parathyroid hormone-like protein from cultured human keratinocytes. Science 1986; 231:388. 24. Chance JJ, Li DJ, Jones KA, et al. Technical evaluation of five glucose meters with data management capabilities. Am J Clin Pathol 1999; 111:547. 25. Weykamp CW. Effect of calibration on dispersion of glycohemoglobin values determined by 111 laboratories using 21 methods. Clin Chem 1994; 40:136. 26. Chen D, et al. Hemoglobin Raleigh as the cause of a falsely increased hemoglobin A 1c in an automated ion exchange HPLC method. Clin Chem 1998; 44:1296. 27. Wong TK, Pekary AE, Hoo GS, et al. Comparison of methods for measuring free thyroxin in nonthyroidal illness. Clin Chem 1992; 38:720.

CHAPTER 239 EFFECTS OF DRUGS ON ENDOCRINE FUNCTION AND VALUES Principles and Practice of Endocrinology and Metabolism

CHAPTER 239 EFFECTS OF DRUGS ON ENDOCRINE FUNCTION AND VALUES MEETA SHARMA

TABLE 239-1.

Alphabetical List of Drugs with Effects on Endocrine Function and Values*

A great number of drugs may suggest, mimic, or cause disorders of endocrine or metabolic function. Some of these are negligible and relatively well tolerated; others may be severe or even life-threatening. With some drugs, the disorder occurs rarely; with others, it occurs very frequently. Moreover, a great number of drugs, although they do not affect endocrine function, influence the analysis of one or several endocrine-metabolic tests. This can cause an increase or decrease in the apparent value or may render the test difficult or impossible to perform. Such effects may lead to the erroneous diagnosis of an endocrine-metabolic disorder when none is present or may mask the bona fide presence of such a disorder. This chapter consists of an extensive alphabetical list of many of the drugs that exert one or several of the previously described effects. Most of the drugs are not hormonal. However, if some hormonal agents exert effects on endocrine function or values other than the purpose for which they are prescribed, they may be listed as well. Note that some of the apparent drug effects are based on clinical or laboratory information that may have been reported in the literature only once or twice, and data may be insufficient to prove causality. Table 239-1 includes findings that suggest or indicate endocrine-metabolic dysfunction (e.g., gynecomastia caused by spironolactone), effects on laboratory values (e.g., hyperkalemia caused by spironolactone), and/or interferences with a test on an analytical basis (e.g., an increase of urinary 17-hydroxycorticosteroids by a metabolite of spironolactone). Most of the drugs listed have been approved by the Food and Drug Administration. A few, however, are in use only in nations other than the United States. In addition, several agents are included that are only investigational. Although most of the signs and symptoms are endocrine-metabolic in nature, other clinically relevant findings are included to assist the clinician in making the correct diagnosis. Several of the drugs are grouped into chemical categories, but some of the constituent drugs within these categories are listed separately if they have been individually associated with certain reported effects. Unless otherwise specified, tests are of the blood. Urine tests are specified as “ur,” analytical effects are specified as “an.” The analytical effect often depends on the method used to perform the test. The term “exp” indicates an effect in experimental animals. The term “inc” refers to an increase of a laboratory value, and “dec” indicates a decrease of a laboratory value. For references or further information, contact Dr. Meeta Sharma, Washington Hospital Center, 110 Irving St. NW, Washington, DC 20010. TABLE 239-1.

Alphabetical List of Drugs with Effects on Endocrine Function and Values*

CHAPTER 240 DNA DIAGNOSIS OF ENDOCRINE DISEASE Principles and Practice of Endocrinology and Metabolism

CHAPTER 240 DNA DIAGNOSIS OF ENDOCRINE DISEASE J. FIELDING HEJTMANCIK AND HARRY OSTRER Situations Appropriate for DNA Diagnosis Direct Detection of Mutation Types of Mutations Methods of Detecting Mutations Screening for Mutations Diagnosis Using Linked Markers Principles of Linkage Types of Polymorphic Markers Requirements for Linkage-Based Diagnosis Practical Considerations Chapter References

The dramatic progress in molecular genetics over the last two decades has made diagnosis possible for a variety of inherited diseases by direct analysis of DNA. Patients with many of these diseases are seen routinely by endocrinologists and primary care physicians, who are likely to have use for these diagnostic tests. Although these tests are extremely powerful diagnostic tools, their use and interpretation often are not straightforward. A confusing and rapidly changing array of laboratory techniques, each with particular strengths and weaknesses, confront the practitioner. In this chapter, the nature of these tests, the genetic principles on which they are based, and the types of information they provide are discussed. Finally, a list of diseases for which DNA diagnostic tests are available and suggestions for their practical use are provided.

SITUATIONS APPROPRIATE FOR DNA DIAGNOSIS In general, DNA analysis is used when more classic diagnostic tests cannot be applied. Although DNA diagnosis is extremely powerful and highly accurate, it tends to be more labor intensive and expensive than the diagnostic tests used routinely in most hospitals. This may change with the advent of powerful new technology to screen for DNA sequence changes using oligonucleotide probes arrayed on silicon chips (see later). Presently, however, cases are found in which the more routine diagnostic tests cannot be used or have marked limitations, often when the aberrant gene product is unknown or cannot be detected in readily accessible tissues. One particularly important example likely to come to the attention of an endocrinologist is multiple endocrine neoplasia, in which presymptomatic DNA diagnosis allows persons at risk to be observed closely for the development of tumors or undergo presymptomatic intervention.1 Both multiple endocrine neoplasia type 1, the gene for which is located on chromosome band 11q13,2 and multiple endocrine neoplasia type 2, which includes medullary thyroid carcinoma and is caused by mutations in the c-ret protooncogene on chromosome band 10q11, can be identified by direct screening for causative mutations or, if necessary, diagnosed by linkage (see later and Chap. 188). Other examples include prenatal and postmortem testing. In these cases, DNA analysis is preferred because it relies on detection of changes in the DNA complement of the patient, which is (mostly) identical in all cells and at all stages of development and disease progression. Thus, a blood sample or even a saline mouth rinse might be collected rather than a biopsy sample from difficult-to-obtain tissues, such as the liver or those of the central nervous system. Prenatal diagnosis, although it provides prospective parents with information about the disease status of a pregnancy at risk, can also confront them with the difficult decision of whether to continue with a pregnancy in the face of a potentially devastating illness or undergo pregnancy termination. This decision can be avoided now that preimplantation diagnosis of genetic disease is possible. In this procedure, which is available at a small number of specialized centers, women use fertility drugs to produce hyperovulation; the oocytes are collected by vaginal ultra-sonography and then inseminated in vitro. One or two blastomeres are harvested, usually at the eight-cell stage, and analyzed for genetic defects. Whereas initial cases were limited to sexing for X-linked diseases, the polymerase chain reaction technique has allowed reliable detection of single gene defects, including those for cystic fibrosis, Lesch-Nyhan disease, Tay-Sachs disease, and Duchenne muscular dystrophy.3 DNA analysis falls into two categories: direct detection of mutations and diagnosis by linkage to genetic markers. In some situations, such as with familial male precocious puberty, specific mutations produce the disease phenotype.4 These can be identified most efficiently by testing for the presence of the specific mutation in the affected gene. In other cases, the disease-causing gene and its product have not been identified, or disease-causing mutations cannot be identified readily. For example, maturityonset diabetes of youth (MODY) is known to be a genetic condition with an autosomal dominant pattern of transmission. In some families, mutations of the glucokinase gene on chromosome 7 have been demonstrated.5,6 In other families with MODY, mutations in the hepatic nuclear factor 1 gene on chromosome 12 or the hepatic nuclear factor 4 gene on chromosome 20 have been demonstrated.7,8 When MODY is first identified in a family, which gene harbors the causative mutation is not obvious. By studying the transmission of anonymous DNA markers linked to the causative genes on chromosomes 7, 12, and 20 in MODY families, linkage to one of these three genes can be established and the likelihood that any individual has inherited a MODY-causing gene can be determined with a high degree of reliability. DNA diagnostic techniques also have been used to determine a person's risk for developing tumors (as with familial medullary carcinoma of the thyroid) and to type tumors once they have occurred. Analysis of tumor tissue can provide information about both the stage and the probable natural history of the tumor. For example, marked increases in the copy number of the N-myc gene in pheochromocytoma is an indication of poor prognosis.9 Because of its broad applicability, DNA diagnosis is useful throughout the human life cycle. Persons who contemplate having children can be tested for carrier recessive gene status to determine their risk. Pregnant women can have prenatal (or preimplantation) diagnosis to determine whether the fetus is affected with a genetic condition. Persons with a family history of disease can undergo testing to determine their risk of developing the disease. DNA diagnostic methods can be applied to postmortem tissue samples and, thus, can be used to determine whether other family members are at risk for the development of a given disorder.10

DIRECT DETECTION OF MUTATION TYPES OF MUTATIONS Inherited disease results from mutations of many different types. Some mutational events involve whole chromosomes. These include deletions, duplications, translocations, and insertions. Deletions and duplications occur from unequal crossing over. Normally, two chromosomes pair during meiosis, and crossing over occurs between the two chromosomes at homologous sites. Sometimes, two or more members of a gene family occur as a tandem array, and pairing occurs between nonidentical members of the gene family. As a result of crossing over, there is a net gain of information on one chromosome, and a net loss of information on the other chromosome. Both can be disruptive. Translocations are exchanges between nonidentical chromosomes. The chromosomes may pair at short regions that have similar sequences. After breaks at these sites, a region from one chromosome is joined to the other. Translocations can disrupt genes, cause novel chimeric genes to be formed, or put a gene under the control of a novel regulatory mechanism. Other mutations involve individual genes. Point mutations result from the substitution of one base for another. Insertion is a process whereby pieces of extraneous DNA are integrated into a gene, disrupting the gene or adding novel coding sequences. In gene conversion, a certain stretch of one gene is lost and is replaced by a region from another gene. One means by which gene conversion may occur is from two unequal recombinational events. A novel form of mutation takes place when a trinucleotide sequence (e.g., CTG) is present as a tandem repeat in a gene.11 DNA polymerase can slip during replication of the repeat region. As a result, the number of repeats either increases or decreases. Alternatively, the polymerase may cycle many times through the trinucleotide repeat, resulting in a marked increase in the number of repeats to hundreds or even thousands of copies. Diseases caused by unstable trinucleotide repeats are often associated with anticipation, in which increasing severity or decreasing age of onset occurs in affected individuals in subsequent generations of a family.12 The effect of these mutations is to alter the sequence, processing, or quantity of RNA transcripts and, thus, to alter the resulting protein products. In some cases, the stability or activity of the protein product is diminished, whereas in other cases, the quantity of the product may be increased or it may gain novel functions. The effect of a mutation on the protein product determines the number of copies of a gene that must carry the mutation to cause disease and, hence, the inheritance pattern for that disease. If inheritance of a single copy of the mutated gene is sufficient to produce disease, it is dominant. If inheritance of two mutated copies of the gene is required to produce disease, it is recessive. Finally, deletions of DNA stretches including several genes may produce phenotypes that are composites of the genetic

conditions caused by the individual genes. These are called contiguous gene syndromes.13 When large, these deletions can be detected through cytogenetic analysis as chromosomal deletions. Otherwise, molecular analysis is required. METHODS OF DETECTING MUTATIONS When the specific mutation causing an inherited disease is known, the disease may be diagnosed by detecting the mutation directly. The choice of test is determined by the relative frequency of the mutation. Some methods are efficient for detecting single prevalent mutations, whereas others are more efficient for identifying multiple mutations that may cause a genetic disease. In some cases, the disease is caused by a wide variety of mutations, no single one of which is prevalent. Some mutations may not be detectable using the commonly available techniques. These mutations must be detected by direct sequencing of the affected person's DNA or by techniques designed to screen for unknown mutations. If the disease-causing mutation cannot be identified but nevertheless is known to be present in a particular gene or genomic region, individuals of unknown genetic status within a family can be tested by using the technique of genetic linkage analysis. Most contemporary DNA diagnostic techniques rely on the polymerase chain reaction (PCR)14 (Fig. 240-1). Primers that are specific for a region of the genome are used to amplify short stretches of DNA. Two primers are used, priming the synthesis of opposite strands of the DNA molecule and “pointing” toward each other. DNA synthesis is initiated from these primers in rounds, with each round of DNA synthesis doubling the number of copies of the template DNA included between the primers. After the first few rounds of synthesis, most of the amplified fragments are products of earlier DNA synthesis reactions rather than of the sample DNA. Hence they are a single discrete size, having the two primers as common ends. The length of the amplified fragment is determined by the distance between the two primers. Amplification of as little as a single DNA molecule, such as from a single blastomere, can be carried out.15 PCR amplification may be applied simultaneously to many regions of a single gene or to multiple different genes. The presence of a PCR product itself is diagnostic for an intact target region on at least one chromosome. Alteration in the size of a PCR product may indicate the presence of insertions or deletions.

FIGURE 240-1. The polymerase chain reaction. The reaction is carried out in cycles, with several different temperatures used in each cycle. First, the double-stranded DNA template is denatured or melted into single strands. Next, the oligonucleotide primers are annealed to the melted DNA. Third, a new DNA strand is synthesized as the polymerase elongates from the primer and copies the template strand. Initially, each cycle doubles the amount of the target region, although the reaction loses efficiency at higher cycle numbers. Usually, 25 to 40 cycles are sufficient to synthesize enough DNA for easy visualization of the product, which has a discrete allele size, by staining with ethidium bromide.

One way the PCR product can be analyzed is by restriction endonuclease analysis (Fig. 240-2). Restriction enzymes cleave DNA at sites with specific recognition sequences. For example, the restriction enzyme EcoRI cleaves DNA at the six-base sequence GAATTC, referred to as the restriction site for that enzyme. When this sequence is altered by a point mutation, the enzyme no longer cleaves at the site, resulting in a longer DNA fragment. Sometimes, point mutations create new restriction sites. Gain or loss of a restriction enzyme recognition site in a DNA sequence is indicative of the presence of a mutation.

FIGURE 240-2. Analysis of known mutations or polymorphisms (diagnosis of Arg201 His mutation in McCune-Albright syndrome). A single base change from G to A in codon 9 of the normal Gsa gene is one of the causes of the McCune-Albright syndrome. This base change also creates a recognition site for NlaIII (CATG), so that the enzyme now cuts at this site. The region encompassing the mutation is amplified using the polymerase chain reaction (PCR). A, The PCR product is digested with NlaIII and electrophoresed. The His201 allele gives two fragments, whereas the Arg201 allele is not cut and gives a single, larger band. B, The PCR product is spotted on a membrane and used as a target for hybridization with two labeled allele-specific oligonucleotide (ASO) probes. The Arg201 and His201 oligonucleotides are each 19 bases long and differ by a single base. Stringent washing that will melt a hybrid with a single mismatch is carried out. A hybridization signal results when at least one copy of the allele homologous to the labeled oligonucleotide probe is present.

The fragments that are generated by restriction enzyme digestion of PCR products or other DNA can be separated on the basis of size by electrophoresis in an agarose gel. Small fragments migrate more rapidly than larger ones. The fragments of DNA in the gel are identified by staining with a specific dye, such as ethidium bromide or silver. The sizes of the fragments depend on the location of the restriction sites. The occurrence of a restriction site is indicated by the presence of two fragments whose sizes sum to that of the uncut DNA molecule. Homozygotes for the presence of the restriction site have only the smaller DNA fragments, whereas heterozygotes have both smaller and larger fragments. A potential pitfall in restriction endonuclease analysis is incomplete cutting by a restriction enzyme. As a result, a homozygote for the presence of a restriction site may be inappropriately labeled as a heterozygote. To correct for such possible inaccuracies, controls are built into the restriction endonuclease digestions. These may include (a) choosing a restriction enzyme that cuts at more than one site in the PCR product, with one site being constant and the other being at a location where mutations are sometimes found; (b) cutting with more than one restriction enzyme, where on